HOMEWORK

EconomicsResources.zip

Home >Business & Finance homework help >Economics homework help >HOMEWORK

Economics Resources/Taxes on Savings - GRUBER File.pdf

__MACOSX/Economics Resources/._Taxes on Savings - GRUBER File.pdf

Economics Resources/Laffer - Ihori.pdf

Laffer paradox, Leviathan, and political contest Author(s): Toshihiro Ihori and C.C. Yang Source: Public Choice, Vol. 151, No. 1/2 (April 2012), pp. 137-148 Published by: Springer Stable URL: https://www.jstor.org/stable/41406919 Accessed: 18-11-2018 17:29 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Public Choice

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Public Choice (2012) 151:137-148 DO! 10. 1007/sl 1 127-010-9737-z

Laffer paradox, Leviathan, and political contest

Toshihiro Ihori • C.C. Yang

Received: 17 January 2010 / Accepted: 20 October 2010 / Published online: 3 November 2010 €> Springer Science+Business Media« LLC 2010

Abstract This paper considers a political contest model wherein self-interested politicians seek rents from the public budget, while general voters make political efforts to protest against politicians* rent seeking directly (for example, through voting in referendums such as the passage of Proposition 13) or indirectly (for example, through donating money to organized groups such as the National Taxpayer Union). We show that the political contest may ironically lead to the Laffer paradox; that is, rent-seeking politicians may intend to set the tax rate higher than the revenue-maximizing rate. For taming Leviathans, political protests may not be as effective as competition among governments.

Keywords Laffer paradox • Leviathan • Political contest • Revenue-maximizing rate

JEL Classification D72 F20 H41 H71

1 Introduction

Leviathan-type governments without constitutional constraints impose taxes at a rate that maximizes the tax revenue. This rate is higher than the rate that maximizes social welfare in the standard framework wherein social welfare depends on useful public goods but not

T. Ihori (El) Department of Economics, University of Tokyo, Hongo, Tokyo 1 13-0033, Japan e-mail: ihori @e.u-tokyo.ac.jp

C.C. Yang Institute of Economics, Academia Sinica, Nankang, Taipei 1 15, Taiwan e-mail: [email protected]

C.C. Yang Department of Public Finance, National Chengchi University, Taipei, Taiwan

C.C. Yang Department of Public Finance, Feng Chia University, Taichung, Taiwan

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

138 Public Choice (2012) 151:137-148

on wasteful spending or rent. It is widely known that a strong constraint can be set by the constitution to limit the tax rate. However, other constraints are required when constitutional

constraints are not available. Plausible candidates include competition among governments (as an exit) and pressure by voters (as a voice). Institutional competition among governments may take the form of an "exit" such

as tax competition, which is popular in theory and practice.1 Brennan and Buchanan (1980), among others, showed that because of institutional competition among governments, revenue-seeking governments in a federation will end up on the upward-sloping part of the Laffer curve. On the contrary, Apolte (2001) indicated that such a taming effect can only be expected if a certain rule of competition among several decentralized governments is ap- plied. He suggested that federalism is not necessarily a substitute for constitutional limits to Leviathans.

In addition to institutional competition, it is important to examine the role of political pressure by general voters because the amount of rent seeking is usually affected by the voters' "voice," as pointed out by Hoyt (1999).2 See also Cheikobossian (2008), Edwards and Keen (1996), and Besley and Smart (2007).

Suppose that there are two types of public spending: wasteful spending and useful spend- ing. A rent-seeking government would prefer to increase the share of wasteful spending by conducting its political activities. On the other hand, the voters also have an incentive to perform their political activities or make efforts to reduce the share of wasteful spending and increase that of useful spending. The actual distribution of tax revenue between use- ful and wasteful spending is determined as the outcome of political contests between the rent-seeking government and the voters.

In this paper, we consider a simple formulation of a political contest. In our approach, the rent-seeking politicians and the general voters engage in a political contest in terms of resources. The greater the amount of political effort by the voters (rent seekers), the greater is the share of useful spending (wasteful spending) at the given level of total tax revenue. This political contest can result in a compromise. In reality, voters make some political efforts to influence budgetary outcomes through voting, writing articles, lobbies, and protests, while politicians make such efforts through campaigns, logrolling, bribery, and corruption.

Buchanan (1980) suggested a property right perspective on rent seeking wherein rent- seeking activities may be viewed as attempts to redefine property rights. Our political con- test model adopts this approach. More specifically, the voters may have property rights over the tax revenue collected nominally. However, these rights are not secure, since they can be altered or reallocated as a result of theft or rent seeking by the politicians. Offense creates a demand for defense, and hence, as first pointed out by Wenders (1987), rent seeking self- generates rent defending. Instead of remaining idle and awaiting the outcome of politicians' rent seeking, the voters may intend to protest against such activities. As a parallel to cam- paigns, logrolling, bribery, and corruption by politicians aiming to exploit budgetary rents, voting, writing articles, lobbies, and protests by voters against exploitation by politicians can be observed in the real world.

A natural conjecture about the outcome of a political contest is that the equilibrium tax rate will be set on the upward slope of the Laffer curve since the political effort by general voters imposes some degree of political constraints on rent-seeking behavior. Contrary to

1 Tiebout (1956) and Hirsschman (1970) are two classical works on the "exit" issue.

2 For research on rent-seeking, see Congleton et al. (2008).

Ф Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Public Choice (2012) 151:137-148 139

this conjecture, we show that the rent-seeking politicians may intend to set the tax rate higher than the revenue-maximizing rate. This is mainly because an increase in the tax rate will engender a negative income effect on the political efforts of voters. The inclusion of a political contest leads to two main effects if the tax rate is raised. First,

the corresponding increase in the tax revenue, if any, will stimulate both the rent-seeking be-

havior of politicians and the rent-reducing behavior of voters in the political contest. How- ever, this tax revenue effect is nil at the revenue-maximizing tax rate because the tax revenue

will not marginally change at this rate. Second, an increase in the tax rate at the revenue- maximizing point will undermine the political efforts of voters by reducing their disposable income. This negative income effect is beneficial to the rent seeker in the political contest as it, other things being equal, raises the relative share of tax revenue allocated to him/her. The second effect dominates the first effect at the top of the Laffer curve, thereby leading to the

Laffer paradox. The main message conveyed by our paper is that the "voice" of the general public may not be as effective as competition among governments at curbing politicians' rent seeking. We also consider an extended model in which politicians exhibit neither completely self-

interested nor completely benevolent behavior. We show that if the degree of a politician's rent seeking is not very high, the Laffer paradox does not occur. It occurs only if the degree

of politicians' rent seeking exceeds some threshold. Shughart and Tollison (1991) and Wrede (1996, 1999), among others, showed that in the

case of tax source sharing, revenue-seeking governments in a federation will end up on the downward-sloping part of the Laffer curve.3 In the present framework, we assume away tax source sharing but incorporate the cost of obtaining rent. Interestingly, politicians still intend

to set the tax rate at a level higher than the revenue-maximizing rate.

The rest of the paper is organized as follows. Section 2 presents the analytical framework. Section 3 considers the political contest model where the politician is a rent seeker, while Sect. 4 examines a more general version of the model in which the politician maximizes the weighted sum of his/her rent and the welfare of voters. Finally, Sect. 5 concludes the paper.

2 Basic model

2. 1 Analytical framework

We develop a simple budgetary model in which the rent-seeking politicians (RPs) and the general voters (VTs) interact in a small open economy.

The government not only provides useful public goods G but also engages in wasteful spending S. Public good G is beneficial to the voters, whereas wasteful spending S is ben- eficial to the rent-seeking politicians. Following the tradition of Leviathan models of gov- ernment, as in Brennan and Buchanan (1980) and others, politicians prefer wasteful public

spending (5), which provides them with opportunities to enhance their personal welfare. The relative price of public and private goods is set to unity for simplicity. Let r denote

the tax rate, У, the total income, and г У, the total tax revenue. The government budget constraint is given as follows:

С+5=гУ, (1)

3See also Anderson et al. (1989).

Ф Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

140 Public Choice (2012) 151:137-148

where S denotes the gross wasteful spending or gross rent of RP. With regard to the budget constraint of politicians, we have

5=S + a, (2)

where S represents the net wasteful spending or net rent of RP, and a political spending or efforts by RP. The objective of the representative RP is to maximize S. On the other hand, the social welfare, W, which reflects VTs preferences over public goods G and private consumption c, is given by

W = u(G) + h(cl (3)

where u(G) denotes utility from public consumption G, with u! > 0 and u" < 0, and h(c) refers to utility from private consumption с with b! > 0 and h" < 0. VTs are consumers and investors in the economic sphere. They engage in private in-

vestment &, which has the productive effect of raising income, and thereby, tax revenue. Moreover, к may be regarded as the various efforts made to increase private income, such as physical investment, human investment, or labor supply. We assume that Y is dependent on private investment by the private sector with Y = £ f(k) = nf(k ), where / is the per capita income and n is the number of general voters. The function / is assumed to satisfy the

standard condition: /' > 0, and f" < 0. Henceforth, we assume n = 1 for simplicity; this implies that the free-rider problem does not exist among VTs, which provides them with the best scenario for dealing with RPs. However, the main result of our paper will qualitatively hold even if we allow for the case wherein n > 1 (see Sect. 3.4). VTs also make political efforts e . These political efforts may be direct, for example,

through voting in referendums such as the passage of Proposition 13,4 or indirect, for ex- ample, through donating money to organized groups such as the National Taxpayer Union.5 The budget constraint of each voter is given as

c + e + k = ('-r)f(k). (4)

For simplicity, investment is assumed to produce output instantaneously. Therefore, we may use the static model.

2.2 Pure rent-seeking model

We first consider the pure rent-seeking model as a benchmark. Without any political contest,

a = e = 0 and G = 0. RP is assumed to maximize S simply by choosing r. The timing of the game is as follows. First, RP chooses r to maximize S . Then, VT determines к and c. The first-order condition with respect to к for VT is

(l-r)/'(*)=l. (5)

4 Proposition 13 endorsed by California voters to limit property tax burdens is a renowned example.

5The National Taxpayer Union in the USA is "a nonprofit, nonpartisan citizen group whose members work every day for lower taxes and smaller government at all levels." There are many other similar oiganized groups, including the California Taxpayers' Association ("a watchdog group founded in 1926 to protect tax- payers from unnecessary taxes and to promote efficient, quality government services'*) and World Taxpayers Associations ("working together for lower taxes, less waste, accountable government and taxpayer rights all over the world"). The quotations here appear in the websites of the respective organizations.

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Public Choice (2012) 151:137-148 141

VT's responses to r can be summarized by the functions

k = k( x) and

c = c( r).

It is clear that the total tax revenue, xY , also becomes a function of r.

The optimal condition with respect to r for RP is given as

dS d(xY) -j- dx = dx - - = °. (6) -j- dx dx

Since the total tax revenue is used solely for rent (or wasteful spending), it is optimal for RP to choose the tax rate that maximizes the total tax revenue. We denote the revenue-

maximizing tax rate by Хм .

2.3 Pure benevolent model

We consider the pure benevolent model as the other benchmark. A benevolent RP chooses G and X to maximize

W = M(G) + /i[(l-r)/-*].

The timing of the game is as follows. First, RP chooses G and r to maximize W. Then VT chooses к to maximize W at the given G and r.

In the second stage of the game, the first-order condition with respect to к is the same as that in Sect. 2.2; that is,

(l-r)/'=l. (5)

As a result, VT's response functions for к and с are the same as those in Sect. 2.2. In the first stage of the game, the first-order condition with respect to r (and hence G) is

= u'd-^- dx + *'[(1 - r )/' - 1]^ dx - tí f = 0. (7) dx dx dx

The second term reduces to zero owing to the first-order condition with respect to k' that is, (5). The third term represents the (negative) income effect of raising tax on consumption. Thus, at the optimum level,

- - > 0 since h f J > 0. dx > since J In other words, the optimal tax rate set by the benevolent RP is less than the revenue- maximizing tax rate, г м. This is the standard result since the benevolent RP considers the marginal cost (negative income effect) of raising r on private consumption с as well as the marginal benefit from raising r on the provision of public good G.

Э Political contest approach

Section 2 considers two extreme governments, namely, pure rent seeking and pure benevo- lent. These two extremes correspond to two broad types of governments that are based on the doctrine of self-interest and the doctrine of the common good, respectively. In both the

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

6See Nitzan (1994), Garfìnkel and Skaperdas (2006), and Konrad (2007) for surveys of relevant literature.

7This form of the contest success function is widely employed in studies on conflict/contest. See Konrad (2007, Sect. 2.3) for its justifications. We discuss a more general formulation in Sect. 3.4.

8 As noted by Garfìnkel and Skaperdas (2006), this property is analytically convenient like the Cobb-Douglas form in the case of production functions in neoclassical economics, and this may be a reason for its popularity among applications.

â Springer

142 Public Choice (2012) 151:137-148

models, RP is only allowed to choose tax rate r, and VT is only allowed to choose invest- ment k. In particular, RP's rent seeking is constrained only by his/her ability to extract tax revenues from VT through taxation. However, it is obvious that VT makes political efforts to influence budgetary outcomes through voting, writing articles, lobbies, and protests. Like- wise, it is obvious that RP makes such efforts through campaigns, logrolling, bribery, and corruption.

We now incorporate political efforts by RP and VT into the pure model. The timing of the game is as follows:

Stage I: RP determines the tax rate and his/her political effort. Stage П: VT decides his/her investment, private consumption, and his/her political effort. Stage III: The political contest determines the actual distribution of tax revenue between

useful and wasteful projects.

This formulation is a natural extension of the pure rent-seeking model (Sect. 2.2) and the pure benevolent model (Sect. 2.3) wherein RP is allowed to choose a apart from tax rate r, while VT is allowed to choose e apart from investment k. The variable a represents RP's political efforts to seek rents from the government budget, while the variable e represents VT's political efforts to oppose RP's rent seeking.

3.1 Stage III

RP's political efforts to exploit budget rents and VT's political efforts to oppose RP's ex- ploitation trigger a conflict or contest between RP and VT. The conflict/contest involved

is presumably complicated, but a key factor used to determine the "output" of the con- flict/contest is the "inputs" expended by players. Following the seminal work of Tullock (1980) and the ensuing literature,6 we adopt the "production function" approach to the con- flict/contest and assume that the outcome of the political conflict/contest is a function of the

relative share of the political spending of players. Specifically, RP's gross gain S is deter- mined by

5=-^-гУ, a -te (8.1) a -te whereas VT's gross gain G is determined by

G = -?-tY. (8.2) a + e

The outcome of the political conflict/contest between RP and VT is summarized by contest success functions (8.1) and (8.2).7 These functions show that an increase in a at the given e results in an increased distribution of the "pie" r Y in favor of RP but against VT and vice versa. Moreover, the functions exhibit the property of homogeneity of degree zero such that the same proportional increase or decrease in a and e leaves the conflict/contest outcome unchanged.8

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Public Choice (2012) 151:137-148 143

The contest success function usually yields the probability of winning or losing. This for-

mulation may then be justified if both RP and VT are concerned with the expected division of tax revenue. Alternatively, following Long and Vousden (1987) and others, the contest success function may be given a nonprobabilistic interpretation: players expend resources competing for a share of divisible rent rather than the entire indivisible rent and, therefore, the relative share of tax revenue is allocated according to the relative share of players' polit- ical efforts.

Note that (8.1) can be rewritten as

S = - - tY - a. (8.1)' a- he

Thus, RP's net gain or rent S is given by the difference between S and a. If a = 0, then

5= 5 = 0 according to (8. 1 ) and (8. 1 )'. Moreover, note that an increase in e reduces с at a given level of k. The gain in G for VT

is at the expense of private consumption. The central trade-off faced by agents in the conflict literature is between producing goods and exploiting what others have produced (Garfinkel and Skaperdas 2006). The main trade-off faced by VT in our model is between investments for producing goods, which can be used in the private or public sector, and protests against RP's rent seeking in the public sector.

3.2 Stage II

Next, the representative household (VT) maximizes W by choosing his/her investment, con- sumption, and political effort, taking RP's political effort and the tax rate as given, and antic- ipating the political contest constraint (8.2). Then, for the first-order conditions with respect to e and к , we have

u'Ge = tí and (9)

u'GxYTf' + - r) - 1] = 0, (10)

where G< = ä and g'y - From the optimizing behavior of voters, we obtain the response functions for e, к , and c.

In general, e, к, and с are formulated as functions of r and a. However, since we are mainly

interested in the effect of г on e , that is, |^, at r = zM> it is appropriate to separate the effect of r Y on e from that of т on e. At r = z^, xY is fixed with respect to r in the first-order effect sense, such that the value of is the same between the two formulations. Then, we have

е = е(г,гУ,д), (11.1)

£ = £(г,гГ,я), and (11.2)

c = c(r, г Y, a). (11.3)

With regard to the partial derivatives of (1 1.1), (1 1.2), (1 1.3) with respect to r, we have

= '[-u'Gxyf' A + h'f']h"(f - 1), (12.1) 3r A Як 1

= ±[u"(Ge)2 1 + u'Gee+h"K-u'GtYf' + h'f ), and (12.2) Зт A

& Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

144 Public Choice (2012) 151:137-148

^- = (/' - ')kx - eX9 respectively, (12.3) ox

where

A m [u"(Ge)2 + m'G„ + h")[u'GrYTf" + h'f"( 1 - r) + A"[/'(l - г) - 1 ](/' - 1)]

+ h"(f'-ì)[u"GeGtrTf'-h"[f'ì-T)-ì]],

Gee rr = - ,2аг л , and Л > 0 if the second-order condition is satisfied. Note that these deriva- rr ( a+e)¿

tives are meaningful only at the point of г = Тд/ . At this point, = / + r/'fcT = 0. Using (10), (12.1) reduces to

= !-/,"(/' -I)2. (12.1)' ox Л r

Since h" < 0, we have ex = < 0 (unless /' = 1). In addition, (12.2) reduces to

^ = 7[«"(G,)2 A + M'G„ + h")-(f- 1). (12.2)' ox A T

The sign of kT = II depends on the sign of /' - 1 . On the other hand, since = / + r/'fcT = 0 at r = tm, кт is negative at this point. Considering (12.2)', it follows that /' > 1, and hence, ex < 0 at that point. Note that the sign of ex is generally ambiguous at т Ф xM. In other words, an increase in r normally reduces the disposable income, which is the negative income effect. On the other hand, an increase in r raises the total tax revenue if r < rM, which stimulates political effort e; this may be called the tax revenue effect. If the positive tax revenue effect dominates the negative income effect, an increase in r would simulate political effort e. However, the total tax revenue effect is absent at the revenue-maximized point гд/, and hence, an increase in г undermines political effort e .

3.3 Stage I

Here, the rent-seeking RP maximizes 5 by choosing his/her political effort a and tax rate r, anticipating the political contest outcome (8.1) and VT's response functions. The first-order condition with respect to a reduces to

( e-aea)xY = (a- he)2, (13)

The left-hand side of (13) represents the marginal benefit of increasing a and the right-hand side indicates the marginal cost of increasing a for RP.

At the same time, the effect of the tax rate on S at r = хм can be written as

3S _dxY[ a v aexY 1 v aeT Эх _dxY[ dz [ a + a e v (û aexY + e)2 J 1 v (a- aeT be)2' ( }

where denotes the derivative of total tax revenue with respect to г and exY = 377. By definition, = 0 at r = xM* and hence, the first term of (14) reduces to zero at this point.

On the other hand, since ex < 0, the second term of (14) is positive, and hence, > 0 at

X = хм- In other words, ~¡r < 0 at RP's optimal choice of r, which implies that the optimal level of X set by RP is higher than xM. We can call this the Laffer paradox.

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Public Choice (2012) 151:137-148 145

3.4 Some remarks

First, let us consider the case of n > 1 and compare the cooperative and non-cooperative solutions. Suppose that VTs behave noncooperatively in choosing e at Stage II. (8.2) is now replaced by

G = ei+e-' r Y, a +e,+ <?_,

where e_, = e¡. Each VT chooses e¡ to maximize

Wi = u { a+* + e-T [/,Л) + fjikj^' j + Л(0 " t)f(ki) ~ ki ~ e,)

by taking e-i as given. Then the first-order condition with respect to e¡ gives

u'Ge = h' (9)'

where Ge = with e = e¡ for all i's under the symmetric assumption. Note that Y = f(k) if n = 1 but Y = nf(k) if n > 1. In the cooperative case where VTs maximize £ W¡, the first-order condition becomes

nu'Ge = h' . (9)"

Since VTs internalize the spillover effect of each member's political effort at the coopera-

tive solution, the total marginal benefit of e is the sum of each member's marginal benefit,

which is expressed in the left-hand side of (9)". Comparing the two first-order conditions (9)'

and (9)", it is clear that the equilibrium level of e (VT's political effort) at the noncooperative

solution is less than that of the cooperative solution.

Nevertheless, we can still show that in the noncooperative case, Cx < 0 at the revenue-

maximizing point. This is because the first-order condition for each VT in the noncoopera-

tive case is qualitatively the same as that in the cooperative case as long as n < oo. If n goes

to infinity, then the non-cooperative solution implies that e = 0 and the equilibrium reduces

to the pure rent-seeking model of Sect. 2.2. Second, our seemingly paradoxical outcome holds in more general formulations of the

political contest as long as r is set before VT determines e' therefore, an increase in г may reduce e at r = xM. For example, consider the following setting, which is more general than (8.2):

3G 3G л G = G(a,e,rY), Ga = - < 0, Ge = - > 0, л

da oe

3G л 3Ge л rY=dxÝ> " = "э7 <

It can be shown that the Laffer paradox still occurs under this formulation.

Ф Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

146 Public Choice (2012) 151:137-148

4 Degree of rent seeking

4.1 Analytical framework

In the basic framework presented in Sect. 2, we follow the Weberian tradition and suppose that politicians are distinct from voters.9 Specifically, it has been assumed that politicians adopt politics as a vocation and strive to make it their sole source of income. In this section, we relax this assumption. We now consider that politicians themselves are identical to voters, except that politi-

cians use their political influence to seek rents once they are in power. Alternatively, to be elected or reelected, rent-seeking politicians must also pay attention to voter welfare. In any case, politicians may exhibit neither completely self-interested nor completely benevolent behavior.

Suppose that many types of politicians or governments exist. The types of governments or politicians may be represented by their degree of rent seeking, L. If the politicians are only concerned with rent seeking, as in the pure rent-seeking model developed in Sect. 2.2, the degree of rent seeking is the highest and it is normalized as unity. On the other hand, if the politicians are purely benevolent and seek to maximize the social welfare of voters, as in Sect. 2.3, rent seeking is absent and its degree is normalized as zero. In general, the degree of rent seeking, L, is given between 0 and 1. This formulation of

0 < L < 1 is an interesting combination of pure Leviathan and pure benevolent models. We allow politicians to choose a besides r and allow voters to choose e besides k. We consider

L to be exogenously given in our model. It can be perceived that the actual L in a society emerges from the electoral systems or political institutions of the society; evidently, different resulting Ls reflect the different qualities of these systems and institutions.

Specifically, once a type of politician, L, is selected, the objective of RP, E, is given as

E = LS + (1-L)W. (15)

Keen (1995) and Edwards and Keen (1996) use a similar formulation.

4.2 Analytical result

Suppose that a type of politician, L, is exogenously given. The objective of RP, Z, is given as (15). Then the effect of the tax rate on E is given by

ЭЕ dS „ rdW ã7 = iS7 + (1-"ã7' „ (l<*

where

3 S ЭгКГ a aety "I aet 3 дт~ S Эт [a + a e T (a aety + e)*'~ "I (a+<?)2' aet (17) dW 'dxY' a aeTy "1 aex 1 ( dxY' dW эг " c|"ãr[^T7 'dxY' a + ry(^TTj2J+ aeTy "1 (ï+ê)i' aex 1 + Ас(Сг+СгУ"9Г) ( dxY' (18)

9For a discussion on the Weberian tradition of modeling politicians, see Merlo (2006).

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Public Choice (2012) 151:137-148 147

ст = 1^, and cXY = ^7. Hence, at r = Тм , we have

ЭЕ ae г Г aer Л

ЭЕ э7 = -Lry bŽ? ae г + 0 - L)rcrK Г aer + hcCt[ Л (19)

Since ex < 0, the first term of (19) is positive. Examining the sign of cT = f£ at r = xM yields

cr = (/'-D*r-*r <0. (20)

With cr < 0, the second term of (19) is negative. If the second term dominates the first term,

an increase in r at a given level of r Y would reduce E. In this case, the Laffer paradox does not occur.

From (19), it is clear that is increasing with L. Let us define L, which satisfies = 0 at r = хм . Then > 0 at r = тм if and only if L > L. In other words, the Laffer paradox will occur if L > L and not occur otherwise. Note that if L = 0, then < 0 at r = тм must be true. It then follows that 0 < L < 1 .

In this general version of the rent-seeking model, the Laffer paradox does not necessarily occur, since the paradoxical possibility also depends on the level of L. If the degree of a politician's rent-seeking is higher, and his/her L is greater than L, the Laffer paradox is more likely to occur, and vice versa.

5 Concluding comments

Pure benevolent governments impose a tax rate at a level lower than the revenue-maximizing tax rate. On the other hand, pure Leviathan-type governments impose taxes at the level that maximizes the tax revenue. It is now widely recognized that competition among govern- ments can serve as an appropriate substitute for constitutional constraints on the power of politicians.

Instead of institutional competition, we have examined the role of political protests as limits to Leviathans. More specifically, we consider a political contest model wherein self- interested politicians seek rents from public budgets, while general voters make political efforts to protest against politicians' rent seeking directly (for example, through voting in referendums such as the passage of Proposition 13) or indirectly (for example, through do- nating money to organized groups such as the National Taxpayer Union). It is shown that ironically, the Laffer paradox can occur in the political contest between rent-seeking politi- cians and general voters. Therefore, we provide an example where "voice" can increase, rather than decrease, the tax rate.

We have explored the possibility that political protests may not limit the power of politi- cians. We do not claim that the Laffer paradox always occurs in a political contest model. If the degree of a politician rent seeking is low, the Laffer paradox is less likely to occur. Our model is admittedly highly stylized, and it abstracts from several possible complications in the real world. In particular, we focus on the conflict/contest between voters and politicians but ignore their heterogeneity. This excludes possible conflicts among voters (for exam- ple, various individuals or interest groups competing for budgets, as addressed in Becker 1983) and among politicians themselves (for example, politicians pursuing their own career and personal interests and disagreeing over the distribution of budgets as revealed in Baron and Ferejohn 1989). Nevertheless, we hope that this paper has highlighted the limitation of "voice" in constraining the power of politicians and served as a meaningful attempt toward attaining a relatively complete solution for containing Leviathans.

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

148 Public Choice (2012) 151:137-148

Acknowledgements Earlier versions of the paper were presented at the Public Choice Society meetings in Las Vegas, European Public Choice Society meetings in Athens, and Australasian Public Choice Confer- ence in Melbourne in 2009. We would like to thank the participants, referees, and the editor for their useful comments.

References

Anderson, G. M., Shughart, W. F. II, Robert, D., & Tollison, R. D. (1989). Political entry barriers and tax incidence: The political economy of sales and excise taxes. Public Finance/Finances Publiques , 44, 8-18.

Apolte, T. (2001). How tame will Leviathan become in institutional competition? Public Choice , /07, 359- 381.

Baron, D., & Ferejohn, J. (1989). Bargaining in legislatures. American Political Science Review , 83, 1181- 1206.

Becker, G. (1983). A theory of competition among pressure groups for political influence. Quarterly Journal of Economics y 98 , 371-400.

Besley, T., & Smart, M. (2007). Fiscal restraints and voter welfare. Journal of Public Economics , 9/, 755- 773.

Brennan, G., & Buchanan, J. M. (1980). The power to tax: analytical foundations of a fiscal constitution. Cambridge: Cambridge University Press.

Buchanan, J. M. (1980). Rent seeking and profit seeking. In J. M. Buchanan, R. D. Tollison, & G. Tullock (Eds.), Toward a theory of the rent-seeking society (pp. 3-15). College Station: Texas A&M University Press.

Cheikobossian, G. (2008). Heterogeneous groups and rent-seeking for public goods. European Journal of Political Economy , 24, 133-150.

Congleton, R. D., Hillman, A. L., & Konrad, K. A. (Eds.) (2008). Forty years of rent-seeking research. Heidelberg: Springer.

Edwards, J., & Keen, M. (1996). Tax competition and Leviathan. European Economic Review , 40 , 1 13-134. Garfìnkel, M. R., & Skaperdas, S. (2006). Economics of conflict: an overview. In T. Sandler & K. Hartley

(Eds.), Handbook of defense economics (Vol. 2). Amsterdam: Elsevier. Hirsschman, A. O. (1970). Exit, Voice and Loyalty: Responses to decline in firms, organization, and states.

Cambridge: Harvard University Press.

Hoyt, W. H. (1999). Leviathan, local government expenditures, and capitalization. Regional Science and Urban Economics , 29, 155-171.

Keen, M. (1995). Pursuing Leviathan: fiscal federalism and international tax competition. In IIPF 51st Congress , Lisbon.

Konrad, K. A. (2007). Strategy in contests - an introduction, book manuscript. Berlin: WZB. Long, N. V., & Vousden, N. (1987). Risk-averse rent seeking with shared rents. Economic Journal , 97, 971-

985.

Merlo, A. (2006). Whither political economy? theories, facts and issues. In R. Blundell, W. K. Newey, & T. Persson (Eds.), Advances in economics and econometrics (pp. 381-421). Cambridge: Cambridge University Press.

Nitzan, S. (1994). Modelling rent-seeking contests. European Journal of Political Economy , /0, 41-60. Shughart, W. F. II, & Tollison, R. D. (1991). Fiscal federalism and the Laffer curve. Economia Delle Scelte

Pubbliche , /, 21-28.

Tiebout, С. M. (1956). A pure theory of local expenditure. Journal of Political Economy , 54, 416-424. Tullock, G. (1980). Efficient rent-seeking. In J. M. Buchanan, R. D. Tollison, & G. Tullock (Eds.), Toward a

theory of the rent-seeking society (pp. 97-1 12). College Station: Texas A&M University Press. Wenders, J. T. (1987). On perfect rent dissipation. American Economic Review , 77, 456-459.

Wrede, M. (1996). Vertical and horizontal tax competition: will uncoordinated Leviathan end up on the wrong side of the Laffer curve? Finanzarchiv, 53 , 461-479.

Wrede, M. (1999). Tragedy of the fiscal common: fiscal stock externalities in a Leviathan model of federalism. Public Choice, 101 , 177-193.

Ф Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:29:19 UTC All use subject to https://about.jstor.org/terms

Contents

p. [137]
p. 138
p. 139
p. 140
p. 141
p. 142
p. 143
p. 144
p. 145
p. 146
p. 147
p. 148

Issue Table of Contents

Public Choice, Vol. 151, No. 1/2 (April 2012) pp. 1-407

Front Matter
An evaluation of EU regional policy. Do structural actions crowd out public spending? [pp. 1-21]
Bargaining unexplained [pp. 23-41]
Incumbent positioning, ideological heterogeneity and mobilization in U.S. House elections [pp. 43-61]
Voter uncertainty and failure of Duverger's law: an empirical analysis [pp. 63-90]
Inequity and risk aversion in sequential public good games [pp. 91-119]
Coalition incentives for political budget cycles [pp. 121-136]
Laffer paradox, Leviathan, and political contest [pp. 137-148]
Bureaucrats and short-term politics [pp. 149-163]
Fiscal decentralization and natural hazard risks [pp. 165-183]
Islam and democracy [pp. 185-192]
Social identity and voting behavior [pp. 193-214]
Do ideological and political motives really matter in the public choice of local services management? Evidence from urban water services in Spain [pp. 215-228]
þÿ�þ�ÿ��T��h��e�� e��c��o��n��o��m��i��c�� e��f��f��e��c��t��s�� o��f�� f��e��d��e��r��a��l��i��s��m�� a��n��d�� d��e��c��e��n��t��r��a��l��i��z��a��t��i��o��n��a�� c��r��o��s��s��-��c��o��u��n��t��r��y�� a��s��s��e��s��s��m��e��n��t�� [��p��p��.�� 2��2��9��-��2��5��4��]
Determinants of government size: evidence from China [pp. 255-270]
þÿ�þ�ÿ��C��h��i��n��a��'��s�� e��v��o��l��u��t��i��o��n�� t��o��w��a��r��d�� a��n�� a��u��t��h��o��r��i��t��a��r��i��a��n�� m��a��r��k��e��t�� e��c��o��n��o��m��y��a�� p��r��e��d��a��t��o��r��-��p��r��e��y�� e��v��o��l��u��t��i��o��n��a��r��y�� m��o��d��e��l�� w��i��t��h�� i��n��t��e��l��l��i��g��e��n��t�� d��e��s��i��g��n�� [��p��p��.�� 2��7��1��-��2��8��7��]
European monetary policy and the ECB rotation model: Voting power of the core versus the periphery [pp. 289-323]
Do elections affect the composition of fiscal policy in developed, established democracies? [pp. 325-362]
Do the IMF and the World Bank influence voting in the UN General Assembly? [pp. 363-397]
BOOK REVIEWS

Review: untitled [pp. 399-401]
Review: untitled [pp. 403-404]
Review: untitled [pp. 405-407]

Back Matter

__MACOSX/Economics Resources/._Laffer - Ihori.pdf

Economics Resources/Structure and Coherence in the Political Economy of Public Finance - Winer File.pdf

__MACOSX/Economics Resources/._Structure and Coherence in the Political Economy of Public Finance - Winer File.pdf

Economics Resources/Fiscal Competition - Wildasin File.pdf

__MACOSX/Economics Resources/._Fiscal Competition - Wildasin File.pdf

Economics Resources/Positive Principles of Taxation - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Positive Principles of Taxation - HOLCOMBE File.pdf

Economics Resources/Federalism - Kessler.pdf

Communication in Federal Politics: Universalism, Policy Uniformity, and the Optimal Allocation of Fiscal Authority

Author(s): Anke S. Kessler

Source: Journal of Political Economy , Vol. 122, No. 4 (August 2014), pp. 766-805

Published by: The University of Chicago Press

Stable URL: https://www.jstor.org/stable/10.1086/676404

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of Political Economy

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

https://www.jstor.org/stable/10.1086/676404

Communication in Federal Politics: Universalism, Policy Uniformity, and

the Optimal Allocation of Fiscal Authority

Anke S. Kessler

Simon Fraser University

The paper presents a positive model of policy formation in federal leg-

islatures when delegates engage in the strategic exchange of policy- relevant information. Depending on the type of policy under consider-

I. I

The

Daro Torst Instit and s with dian Coun usual

ation,communicationbetweendelegatesgenerallysuffersfromabiasthat makes truthful communication difficult and sometimes impossible. This generates inefficient federal policy choices that are often endogenously characterized by overspending, universalism, and uniformity. Building on these findings, I develop a theory of fiscal ðde-Þcentralization, which revisits the work of Oates in a world of incomplete information and strategic communication. Empirical results from a cross section of US municipalities are consistent with the predicted pattern of spending.

ntroduction

majority of countries are federally organized; multiple tiers of gov-

ernment fulfill a variety of functions in revenue raising, taxation, and public expenditure. Among those, there has been a recent trend toward

I am indebted to Philip Reny ðthe editorÞ for his detailed and thoughtful comments on an earlier draft and two anonymous referees for valuable suggestions. I also wish to thank

n Acemoglu, Philippe Aghion, Elhanan Helpman, Kevin Milligan, Krishna Pendakur, en Persson, Guido Tabellini, as well as various participants of the 2012 International ute of Public Finance congress, the 2008 Canadian Public Economic Group Meeting, eminars at the Universities of Basel, Cologne, and Munich for helpful discussions, special thanks belonging to Tim Besley and Ken Shepsle. I am grateful to the Cana- Institute for Advanced Research and the Social Science and Humanities Research cil for financial support and to Ross Hickey for invaluable research assistance. The disclaimer applies.

766

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

reallocating fiscal responsibilities from central governments to regional or local governments. Examples of countries that decentralized include the United States, the United Kingdom with its newly created regional

communication in federal politics 767

legislatures in Scotland and Wales, a number of Latin American coun- tries ðFaguet 2004Þ, as well as Indonesia and Pakistan.1 Decentralization is also one of the World Bank’s stated policy objectives for developing countries.2 At the same time, the debate among member countries of the European Union has evolved around “subsidiarity” versus centralization, that is, which functions should remain with the regions and which can sensibly be assumed by the European Parliament and the Council. In popular discussions decentralization is often seen as preferable, fos-

tering efficiency through intergovernmental competition and account- ability through local say over service provision. The economic literature building on Oates’s ð1972Þ famous decentralization theorem promotes a more balanced view by emphasizing a trade-off between local policy de- cisions that are better tailored to the needs of the local population ð“closer to the people”Þ and the obvious advantages of centralization if policies exhibit large economies of scale or spillover effects across jurisdictions. While Oates’s perspective has shaped the theory of fiscal federalism for decades, it has recently come under criticism on the grounds that an es- sential aspect of policy formation is missing, namely, an understanding of how political actors on various levels of government are incentivized and interact. This paper addresses fiscal federalism from a political economy per-

spective. It has two purposes. First, I develop a model of legislative be- havior that allows for communication as a key factor in legislative deci- sion making on the federal ðcentralÞ level.3 In a second step, I am then

1 In the United States, the most prominent explicit decentralization initiative was to re- turn responsibility over major welfare programs to the individual states through the Per-

sonal Responsibility and Work Opportunity Reconciliation Act of 1996, under which fund- ing for state-run welfare programs switched from open-ended matching grants to fixed block grants and simultaneously increased the discretion of states to make decisions re- garding their own welfare expenditures. Although the Obama administration has moved toward more centralized provision of some aspects of government like health insurance, federalism is still important when it comes to policies such as the stimulus package and aid to states, which has totaled $2.4 trillion over the past 3 years ðsource: Congressional Budget OfficeÞ.

2 In 2002, around 30 World Bank projects had decentralization components. A total of US$500 million was given in the form of various loans to countries such as Mexico, Ar- gentina, Brazil, India, Mexico. and Pakistan. See http://www1.worldbank.org/publicsector /decentralization/http://www1.worldbank.org/publicsector/decentralization/.

3 Communication in legislatures has received surprisingly little attention in the literature. Austen-Smith ð1990Þ analyzes the informational role of debate in federal legislatures in a series of examples with majority voting over fixed or endogenous agendas. The author shows that communication allows legislators who would otherwise reveal their information through proposals to share this information prior to the agenda-setting stage. Gilligan and Krehbiel ð1989Þ, Epstein ð1998Þ, and Krishna and Morgan ð2001Þ study legislatures in which informed

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

able to compare federal decisions to local decisions in order to gain in- sights into the optimal degree of ðde-Þcentralization and the determi- nants of fiscal authority.

768 journal of political economy

The analytical framework I employ conceptually formalizes the process of federal policy formation through consultation with interested parties ðlocal officialsÞ or, alternatively, a policy debate among local delegates in a federal assembly. The aim is to capture the stylized elements of a coun- try with multiple layers of government and to study which factors deter- mine the incentives of local representatives to truthfully reveal locally dis- persed information on their constituents’ preferences over public policy decisions. The economy is divided into a given number of ðgeographicalÞ districts, which are defined by common tastes over a local public project ðgoodÞ with spillover effects on other districts. Local preferences are local knowledge. Under federal fiscal authority, districts ðthrough a represen- tativeÞ can communicate their preferences to the central government, which implements whichever policy it considers best given the transmit- ted information.4 Federally controlled public policies are funded through general taxation, which implies that costs are underestimated at the local level ceteris paribus. As I show, this implies that local representatives have an incentive to overstate the local benefit, on average, in order to seek federal ratification. Moreover, this tendency unambiguously increases in the number of districts and is reinforced for projects whose benefits are locally concentrated ð“distributivepolitics”Þ. Forpolicies that havea public good character, however, this tendency is mitigated. Intuitively, delegates will have an incentive to understate the local benefit ðrespectively, over- state the local costÞ for projects that are costly for their constituents but— because of their positive and large spillovers—are nevertheless likely to receive federal approval. Irrespective of the type of policy under consid- eration, the federal legislature recognizes the resulting communication bias, implying that meaningful transmission of information becomes dif- ficult or even impossible. In the former case, the equilibrium is charac- terized by universalism: every interested district is assured a project. In the latter case, federal policies suffer from uniformity of provision: although it is commonly known that local preferences ðgenericallyÞ differ, those differ- ences fail to be taken into account when the policy decision is made. The

committees can communicate information through policy proposals ðsubmitting billsÞ to an uninformed floor. The objective of these papers is to understand how different rules for adding amendments affect equilibrium informational efficiency. In an abstract model out-

4 In other words, the central government cannot commit to ðmessage-contingentÞ pol- icies that are different from what maximizes its own objective. Communication therefore takes the form of cheap talk ðCrawford and Sobel 1982Þ.

side the political economy context, Dessein ð2002Þ studies communication vs. delegation in a model that is similar to mine but has only two actors, a principal and an agent. The agent has private information that is relevant to the decision, which the principal can either make herself ðafter communicating with the agentÞ or delegate.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

failure is rooted in the strategic considerations between the federal gov- ernment and local officials ðbetween local representatives in a federal as- sembly, respectivelyÞ, which entirely prevents the credible transmission of

communication in federal politics 769

any information. Thus, my results provide a theoretical foundation for two stylized facts about federal politics that have been frequently invoked as assumptions in the theoretical literature, namely, that the legislative process is characterized by universalism and that federal policies are uni- form, that is, do not take local circumstances into account. On the basis of the above findings, I next compare the equilibrium

policy under centralization with the policies that would have been chosen under decentralized decision making. The results give some understand- ing of what determines which level of government should have author- ity over a policy. The reasoning here, although conceptually different from Oates’s original argument, is similar in its conclusions: centraliza- tion is better at internalizing externalities but worse at accommodating the needs of the local population. In this sense, the model provides a foundation for Oates’s celebrated decentralization theorem that is de- rived from an explicit model of government behavior: policy uniformity is not assumed from the outset but is a direct consequence of the impossi- bility of credibly communicating information about local tastes to higher levels of government in equilibrium. Because the problem is more severe the larger the number of districts, however, I also find a congestion effect, which takes the form of diseconomies of scale in communication. As a result, centralization is better only if there are not too many districts, ce- teris paribus. The theoretical model has several implications regarding the rela-

tionship between the size of the legislature and both the level and the composition of public spending. In the final part of the paper, I show that the data are consistent with these predictions. In particular, I doc- ument that the number of legislators has a strong positive effect on the extent of government spending, while at the same time it has a signifi- cant negative effect on the relative share of targetable expenditures in the budget. This second main prediction of the theory, namely, that a larger legislative body should be associated with a shift in the composi- tion of the government budget away from pork toward relatively more public goods, ceteris paribus, is unique to the model and, to the best of my knowledge, has never been empirically documented before. Using an instrumental variable strategy, I show that these findings are robust to a possible endogeneity of legislative size. Related literature.—This paper stands at the intersection of several

strands of research in political economy. First, the paper contributes to the literature on legislative behavior and organization by providing a ra- tionale for “universalism” in Congress. This empirically relevant feature of legislative decision making generates what is sometimes referred to as

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

“the law of 1/n” ðWeingast, Shepsle, and Johnson 1981Þ, which postulates that ðthe inefficiency ofÞ government spending is increasing in the size of the legislature.5 Weingast et al. and, more recently, Primo and Snyder

770 journal of political economy

ð2008Þ formally derive this law in models that are very similar to the pres- ent framework but assume universalism from the outset and do not con- sider the possibility of delegates in the federal assembly communicating privately held information.6

Second, the empirical part of the analysis fits well into the burgeoning literature that examines the effects of political institutions on outcomes in politics and economic policy. In particular, Bradbury and Crain ð2001Þ, Gilligan and Matsusaka ð2001Þ, Baqir ð2002Þ, Perotti and Kontopoulos ð2002Þ, and Egger and Koethenbuerger ð2010Þ empirically study the law of 1/n and find support for the predicted positive relationship between government spending and legislature size. My results confirm this find- ing but go beyond these studies in that I am able to relate the size of the legislature to the composition of government spending as well. For this reason, the paper is also related to Persson, Roland, and Tabellini ð2000Þ, Lizzeri and Persico ð2001Þ, Milesi-Ferretti, Perotti, and Rostagno ð2002Þ, and Persson and Tabellini ð2004Þ, which study how electoral rules and the types of democratic political institutions matter for the level and the composition of government spending.7 The empirical work in this liter- ature mostly uses cross-country comparisons, however, which are prone to suffer from an omitted variables problem. Those are less likely to arise if one exploits within-country variation as I do in the present paper. Finally, the paper contributes to the literature on fiscal federalism. In

particular, one key element in the classic theory of fiscal federalism ðOates 1972Þ is that a central government will provide a single, uniform level of public output in all jurisdictions, a presumption that has come under crit- icism ðSeabright 1996; Lockwood 2002; Besley and Coate 2003Þ.8 In con-

5 There have been several earlier papers that explain the norm of universalism. They rely on a simple expected value comparison between the rewards to legislators in the coalition

as the whole against the uncertainties associated with minimal winning coalitions. As noted by Niou and Ordeshook ð1985Þ, however, such a comparison fails to model directly the decision processes within a legislature and, thus, fails to explain why it is in no member’s self-interest to defect from such a norm.

6 Although powerful, the logic in Weingast et al. ð1981Þ and Primo and Snyder ð2008Þ is unsatisfactory from a theoretical point of view because the legislature is assumed to adopt the “norm” of universalism, unanimously passing omnibus bills that ensure every legislator the projects he or she desires, even though all members would collectively be better off agreeing to a lower level of overall spending.

7 See Besley and Case ð2003Þ for an overview of results with an emphasis on US in- stitutions.

8 Empirically, it is not difficult to document cases in which a central government pro- vides public goods in a discriminatory manner across regions, although a tendency toward “equal treatment” arguably remains. On the theoretical side, it is unclear why a central government does not differentiate between districts.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

trast,policyuniformityatthefederallevelendogenouslyarisesinthepresent model precisely in those circumstances in which centralization is Pareto preferred to decentralization. Although previous contributions to the lit-

communication in federal politics 771

erature on fiscal federalism have looked at informational asymmetries be- tween a central government and regional governments ðsee, in particular, Klibanoff and Morduch 1995; Klibanoff and Poitevin 2013Þ, I am not aware of any similar result. To my knowledge, the only other paper that explicitly addresses policy uniformity is by Harstad ð2007Þ and is quite different in focus. The author considers a model in which two politically autonomous regions negotiate an agreement over local public goods with spillovers under asymmetric information. Although policy uniformity is not an equilibrium phenomenon as in the present framework, Harstad shows that a mutual commitment to policy harmonization ðuniform pol- iciesÞ may be advantageous in interregional negotiations because it re- duces delay in bargaining. The paper proceeds as follows. Section II introduces the basic frame-

work, derives the equilibrium conditions, and shows how the information transmitted in equilibrium depends on various parameters of the model. The theoretical model is taken to the data in Section III. Section IV gives a brief discussion of the findings and presents conclusions. All proofs are relegated to the Appendix.

II. The Model

A. The Basic Framework

Consider an economy divided into n constituencies or districts indexed by i ∈ f1, . . . , ng. Each district is composed of a continuum of homo- geneous households with exogenous income y and mass unity. For ex- positional purposes, it is convenient to think of these districts as being geographically distinct, and I will often refer to them as regions. How- ever, it is equally possible to interpret them as broadly defined con- stituencies, which are separated along observable demographic or eco- nomic ðrather than geographicÞ lines and share a common objective with regard to the policy under consideration. There are n 1 1 goods in the economy, one private consumption good x and n public projects gi ∈ f0, 1g, one for each district. If the latter are geographic entities ðregions, states, municipalitiesÞ, the assumption is that policies are targeted to a particular locality and, as such, have the natural interpretation of public projects in infrastructure, recreation, urban renewal, or the environment. Otherwise, the projects can be thought of as entitlement programs tar- geted to the respective constituency ðfamilies with children, sayÞ. Public spending on projects is financed by nondistortionary local income taxes ti ∈ R, which may differ across districts.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

Within each district, a household gains a private benefit vi > 0 from the locally realized public project. In addition, there are spillovers from projects in other districts j ≠i, denoted by g . The cost of realizing proj-

772 journal of political economy

ect gi is ci ∈ ð0, yÞ. All costs and benefits are measured in terms of the private consumption commodity. The utility of a household in re- gion i from the consumption vector ðg1, . . . , gn, xiÞ is thus equal to

ui 5 vi gi 1 o j ≠i

gij gj 1 xi;

and the Pareto-optimal allocation of public projects is characterized by

g *i ðviÞ 5 1 if vi 1 ji ≥ ci 0 otherwise;

�

where ji ; ojgji measure the aggregate spillover of the public project associated with district i. The parameter measuring the private benefit vi from their own project gi is private information of the individuals be- longing to the respective constituency.9 People outside the constituency, including a central authority, know only that the vi’s are independently distributed according to some smooth distribution function FiðviÞ over the full support Vi 5 ½0; �vi�. All other variables and parameters of the economy are common knowledge. To eliminate trivial cases, I will assume in the remainder that ji ∈ ½ci 2 �vi; ci� for all projects gi, i 5 1, . . . , n. Since people within a district are identical, we can represent each

district by a single agent i 5 f1, . . . , ng who acts on behalf of all citizens. This agent’s payoff from a policy vector fðgi; tiÞgi51; : : : ;n is

ui 5 vigi 1 o j ≠i

gji gj 1 y 2 ti;

which also captures the aggregate surplus generated in district i. Under local authority, the decision over project gi lies with the local district, and the agent should simply be thought of as an elected policy maker who directly determines local policy. In this case, each district is financially responsible for its own ðand only its ownÞ project; that is, the ðlocalÞ bud- get constraint reads ti 5 cigi. The agent chooses gi to maximize

uLi 5 vi gi 1 o j ≠i

gji gj 1 y 2 ci gi; ð1Þ

taking the public good supply in all other districts j ≠ i as given. In equi- librium,

9 I will assume that vi is “soft” information; i.e., agents cannot certify or “prove” the value of vi to others, even if they would like to.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

g Li ðviÞ 5 1 if vi ≥ ci 0 otherwise;

� i 5 1; : : :; n;

communication in federal politics 773

where the superscript L stands for local authority. Because each district’s government takes only the benefit received by its own citizens into ac- count, the policy decisions are Pareto suboptimal:10 there is underpro- vision of public projects if the externality is positive, ji > 0, and over- provision if the externality is negative, ji < 0. Moreover, the size of the distortion, measured by the expected difference between optimal and actual expected surplus,

DLi 5

��Eci2ji ci

ðvi 1 ji 2 ciÞFiðviÞ �� > 0;

is increasing in the degree of spillovers jjij. If the decision over project gi lies with a central ðfederalÞ authority, the

district representative may be thought of as a regional delegate to the federal assembly, an appointed local public official, or a lobbyist who advances the constituency’s interests in the central government. Total federal expenditures oicigi are still funded by taxes on local residents ti, i 5 1, . . . , n, but since funds are often shared at the federal level, the link between a district’s tax bill and the implementation of its project will generally not be perfect; that is, local tax revenues could be raised in- dependently from local project realization. To fix ideas, I will assume that there is some arbitrary but exogenous sharing rule ti 5 o

j51sijcjgj, where sij, with o

i51sij 5 1, denotes district i’s share in the funding of project gj. The central authority’s objective is to choose policies ðg1, . . . , gnÞ to max- imize total surplus:

uC 5 o i

ui 5 o i

� vigi 1 jigi 1 y 2 o

j51

sijcjgj

� : ð2Þ

There are two interpretations to this benevolent objective function. The first is that the central government is a policy maker whose constit- uency consists of the entire economy: while local policy makers care only about their own ðregionalÞ district, the center cares about the welfare of

10 The presumption that politically autonomous jurisdictions do not coordinate their policies is standard in the literature and appears natural in many circumstances. If inter-

jurisdictional contracts are enforceable, however, it is also conceivable that regions engage in Coasian bargaining. So far, only a few papers have studied this possibility of Pareto- improving contractual arrangements under decentralization. See Kessler, Lülfesmann, and Myers ð2009Þ for a model with efficient bargaining and Harstad ð2007Þ for an analysis of bargaining under asymmetric information.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

all citizens. There is an alternative and perhaps more compelling inter- pretation, however. Suppose that all districts choose one agent as a del- egate to send to the national legislature. Since the federal legislature is

774 journal of political economy

an enduring institution in which delegates interact repeatedly through communicating, bargaining, and voting over a fairly long period of time, one can expect regional representatives to overcome inefficiencies caused by majoritarian decision rules ðminimum winning coalitionsÞ and nego- tiate their way to the Pareto frontier. This is what Besley and Coate ð2003Þ have called a “cooperative” legislature. In this view, the above objective represents a utilitarian bargaining solution that could be mutually ac- ceptable under the veil of ignorance; that is, legislators—prior to receiv- ing private information—agree to implement ex post a policy vector that maximizes the joint surplus.11 As we will see shortly, however, this ðex anteÞ efficient objective does not translate into ex post efficiency since credible communication is not always feasible once delegates receive private in- formation on their local project. Before I analyze this case, however, it is instructive to study the case without communication as a benchmark. Thus, suppose that the federal government does not know the re-

alization of the vector of district-specific preference parameters v 5 ðv1; : : :; vnÞ. Given prior beliefs FiðviÞ, federal policies will satisfy

ðg1; : : :; gnÞ ∈ arg maxE � o i

� vigi 1 jigi 2 o

sijcjg j

��

or, for all vi ∈ Vi and i 5 1, . . . , n,

g EEi ðviÞ 5 1 ⇔ E½vi� 1 ji ≥ ci; ð3Þ

where E½�� is the unconditional expectation operator and the superscript EE stands for the ex ante efficient decision, which naturally is the optimal policy for an uninformed federal authority that maximizes total surplus. From ð3Þ, the lack of information for the central government yields

uniform policies gi, at least up to observable differences: consider any two districts i and j whose projects are ex ante indistinguishable, that is, ðci; ji; ViÞ 5 ðcj; jj; VjÞ. Even though local benefits will generally differ,

11 Some would argue that creating a forum for communication to debate policies and negotiate mutually beneficial agreements is one of the main purposes of a national as- sembly. Evidence backs this view: in the US House, for instance, minimum winning coa-

litions are the exception rather than the rule. In the European Union, the number of representatives is relatively small, which makes it likely that they will exploit the benefits of cooperation. Also, many decisions require unanimity, which may force legislators to co- operate. Also note that if the districts are sufficiently symmetric, implementing a joint surplus-maximizing objective under the veil of ignorance would not require side “pay- ments” ðlogrollingÞ.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

vi ≠ vj, a federal authority without further information must treat them equally; that is, either g EEi ðviÞ 5 g EEj ðvjÞ ; 1 or g EEi ðviÞ 5 g EEj ðvjÞ ; 0. More- over, the fact that federal policy is insensitive toward local preferences

communication in federal politics 775

under incomplete information implies that decentralized policy making may be preferable. Indeed, because a local authority is “closer to the people,” it is easy to see that it will often be strictly better if spillovers are very small, jjij → 0. For very large spillovers, in contrast, the inefficiency due to the lack of coordination under local authority dominates and cen- tral authority is the desirable mode of governance. Obviously, these con- clusions mirror Oates’s ð1972Þ classic arguments in favor of or against de- centralization within a framework of locally dispersed private information, taking the informational disadvantage of the federal government as given ðsee also Oates 1999Þ. While alluding to the lack of information on the part of the central

government to explain policy uniformity is appealing at first glance, it remains unsatisfactory on second thought. If the federal government does not know regional preferences and if this is what prevents it from adapting policies that are better suited for the local constituencies, why does it not ask local public officials? More generally, what prevents re- gions from communicating their preferences to the federal government? Indeed, is communication not what a federal assembly with regional del- egates is all about? I therefore next turn to the main part of the analysis, which endogenizes the lack of information at the federal level by allow- ing for communication between local representatives and the federal government ðrespectively, a debate among members in the legislatureÞ.

B. Legislative Communication

Now suppose that authority over spending and taxation rests with the

federal government but information can flow between the central gov- ernment and the districts ðlocal officialsÞ in the sense that the latter can communicate their local benefit vi to the former. In doing so, however, they have to take into account the fact that the center—upon having received the information communicated by the districts and possibly updated its prior on v—will implement its most preferred policy. There are two interpretations of this communication consistent with the two views of a central authority laid out above. First, if the central authority is a federal government that is distinct from the local representatives, one could imagine the center consulting regional representatives and offi- cials on the project before making a decision. Alternatively, the central authority may simply be a federal assembly, which itself is composed of regional delegates. In this interpretation, the information transmission stage can be seen as formalizing delegates communicating with each other—a policy debate. The constraint they operate under, however, is

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

that, at the end of the day, the assembly will pass a bill that maximizes joint surplus.12

Formally, communication and policy selection under the central au-

776 journal of political economy

thority are described by the following three-stage game. In the first stage, nature chooses the vector of local benefits v according to FiðviÞ, and each district representative learns the vi ∈ Vi for its local project gi. The sec- ond stage is a communication stage in which the central government consults with local officials or, alternatively, the delegates in the federal assembly engage in a political debate. At a very general level, this com- munication can be formalized by the exchange of messages mi. Upon observing mi, the central authority forms new beliefs miðmi; �Þ over vi ∈ Vi. Given m 5 ðm1; : : :; mnÞ, the central authority then implements a policy vector ðg1, . . . , gnÞ that maximizes expected social surplus. Note that be- cause the federal government always chooses its most preferred policy, conditional on beliefs m, the only thing communication may achieve is to change m. Any communication is therefore “cheap talk” and could in principle be quite complicated ðthe exchange of messages could be con- ditional and repeatedÞ. Under our assumptions that ðaÞ private infor- mation is not correlated across districts and ðbÞ as far as a single district representative is concerned, the decision on her own local project is in- dependent of what happens in other districts, however, it is easy to see that the cheap talk game for each district can be analyzed separately. Moreover, there is no loss of generality restricting attention to a single message mi that is transmitted from the local representative to the central authority, who then decides on gi given its updated beliefs miðmiÞ.

1. Equilibrium Analysis

Cheap talk games with a single sender have been studied extensively in the seminal contribution of Crawford and Sobel ð1982Þ, who consider a generic version of the game in which a better-informed sender can send arbitrary messages to a receiver who eventually makes an irreversible de- cision that affects the well-being of both. The authors show that the Bayesian Nash equilibria of the game will be characterized by a partition

12 One might argue that allowing the federal government to design a mechanism to elicit the private information of the districts—as opposed to the communication outlined above—would be a natural next step. A mechanism ðcompete contractÞ requires full com-

mitment, however, and by definition, budget authority rests with the federal government in a centralized regime. Importantly, this includes the right to “renege” on promises made, especially when the result would be a Pareto improvement. By assumption, the only way the federal government can commit not to implement its most preferred policy is to decen- tralize, i.e., formally place projects under local authority. By the same token, I also do not consider the possibility of regions “cooperating” under decentralization. Consistent with the central government’s inability to commit, the analysis also implicitly assumes that a local policy maker cannot commit to any policy that does not maximize the respective objective function. See also the discussion following theorem 1.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

of the parameter space ðin our case, ViÞ into connected sets ðintervalsÞ. In each of these equilibria, the sender optimally and truthfully announces to which interval the realized state of nature belongs given the receiver’s

communication in federal politics 777

decision rule, and the receiver maximizes her expected utility, given her updated beliefs based on the correct but coarse information about the state of nature she received from the sender. Before I analyze equilibria with meaningful communication, however,

it is important to note that there always exists an equilibrium in which no information is transmitted. Given that the sender’s message is unrelated to his private information, the receiver rationally does not update her be- liefs and picks the optimal action on the basis of her prior.13 Conversely, since the receiver “ignores” the message sent, any message is consistent with an equilibrium. Of course, there may be other equilibria with finer partitions in which

informative communication is feasible. Since there is one decision made for every message sent, the number of actions taken in equilibrium is limited by the number of elements of the partition ðcorresponding to in- tervals of ViÞ. The converse is also true, however: if there are two mes- sages that trigger identical actions, then we can combine them into a single message without changing the equilibrium outcome. Since there are at most two decisions in the present model, we can thus without loss of generality assume that the message space is f0, 1g; a strategy for dis- trict i is then a mapping mi : Vi → f0; 1g while a strategy for the federal government is g Ci : f0; 1g → f0; 1g. Because preferences satisfy single cross- ing, the finest partition of Vi thus has two intervals, with mi 5 0 if vi belongs to one interval and mi 5 1 if vi belongs to the other interval. Moreover, any message sent by agents with private information on the partition containing higher ðrespectively, lowerÞ values of vi must trigger gi 5 1 ðrespectively, gi 5 0Þ; otherwise, it would be optimal for local rep- resentatives with extreme values of vi to change their strategy ðlieÞ. Lemma 1. Communication is limited under centralization. In par-

ticular, for each district i, there are at most two types of Bayesian Nash equilibria ðup to labeling differencesÞ:

a. Communication is completely uninformative. In this equilibrium, the local representative sends a message miðviÞ 5 miðv0iÞ for all vi, v 0 i ∈ Vi, and the centralized policy consequently satisfies

g Ci ðmiðviÞÞ 5 g EEi ðviÞ ∀ vi ∈ Vi:

13 For instance, the sender could truthfully announce that the realized state belongs to

the entire parameter space.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

b. Communication is informative but incomplete. The correspond- ing equilibrium is characterized by a cutoff value ~vi ∈ ð0;�viÞ such that

778 journal of political economy

miðviÞ 5 1 if vi ∈ ½ ~vi; �vi�

0 otherwise

�

and

g Ci ðmiðviÞÞ 5 miðviÞ ∀ vi ∈ Vi:

Proof. The proof follows directly from Crawford and Sobel ð1982Þ and the above discussion. QED

The equilibrium in part b, which I refer to as the informative commu-

nication equilibrium below, is illustrated in figure 1. Since giðmiðviÞÞ 5 miðviÞ, communication in this type of equilibrium

has a simple structure. The local representatives makes a “recommen- dation” as to whether or not their policy project should be realized, and the federal government follows the recommendation. In equilibrium, all proposals are thus rubber-stamped. Corollary. Whenever communication between the federal gov-

ernment and a representative of district i 5 1, . . . , n is informative, the equilibrium policy is characterized by universalism; that is, each represen- tative asking for a project is assured its approval ðregardless of whether the project is socially desirable or notÞ. Lemma 1 shows that whenever the informative communication equi-

librium exists, it is not unique. So why should we focus on this equilib- rium; that is, is it a natural candidate for equilibrium selection? The an- swer is given in lemma 2. Lemma 2. The equilibrium with informative communication ex ante

Pareto dominates the equilibrium in which no information is transmitted. In what follows, I will assume that agents coordinate on the Pareto-

superior equilibrium, provided that it exists. It remains to analyze when this is the case. To this end, consider an informative communication equilibrium and assume that the representative of district i follows his or her equilibrium strategy as prescribed in lemma 1, that is, sends a

FIG. 1.—Informative communication

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

message miðviÞ 5 1 for values vi ≥ ~vi and miðviÞ 5 0 for values vi < ~vi, re- spectively. For the federal authority to follow its own equilibrium strategy of rubber-stamping the district’s proposal given beliefs m, we therefore

communication in federal politics 779

must have

Evi≥~vi fvi 1 ji 2 cig ≥ 0 and Evi<~vi fvi 1 ji 2 cig ≤ 0;

where the first ðsecondÞ inequality ensures that the government opti- mally chooses g Ci ðmiÞ 5 1 ðrespectively, g Ci ðmiÞ 5 0Þ after inferring vi ≥ ~vi ðrespectively, vi < ~viÞ upon hearing the message mi 5 1 ðrespectively, mi 5 0Þ. Next, consider the representative of the district. Since talk is “cheap” ðsending messages is costlessÞ, for the communication strategy to be op- timal given the government’s prescribed equilibrium response of rubber- stamping, it must be the case that the representative prefers gi 5 1 to gi 5 0 whenever vi ≥ ~vi and gi 5 0 to gi 5 1 otherwise. Since vi is distrib- uted with full support over the interval Vi, a representative with prefer- ence parameter ~vi must be indifferent between both outcomes. Denote by si 5 sii a district’s cost share in its own local project. From ð1Þ, when ci is replaced with sici, whatever the ðexpectedÞ values of gj, j ≠ i, the differ- ence between a district’s payoff between gi 5 1 and gi 5 0 is vi 2 sici. Indifference at vi 5 ~vi thus requires

~vi 5 sici: ð4Þ

Hence, we have the following lemma. Lemma 3. In equilibrium, there is informative communication be-

tween the central government and district i if and only if

Efvi 1 ji 2 cijvi ≥ sicig ≥ 0 ð5Þ and

Efvi 1 ji 2 cijvi < sicig ≤ 0: ð6Þ

In summary, we can characterize the information that is transmitted through a political debate and the subsequent course of action under

centralization as follows. First, communication is imperfect in general, which implies that centralized decisions are never efficient, as would be the case under perfect information. Second, whenever informative com- munication is feasible, the central authority in effect rubber-stamps local proposals or, put differently, the central legislature operates with univer- salistic criteria: every district interested in carrying out a project is as- sured approval. However, the federal government still can—and generally will—discriminate among districts: in equilibrium, given the communi- cated information, the decision rule giðmiÞ depends on the realization of

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

the local benefit vi. Third, whether or not meaningful communication is in fact feasible depends on the characteristics of the public project ðji; ciÞ as well as on the fiscal cost sharing rule s .14 The following subsection stud-

780 journal of political economy

ies these relations in greater detail.

2. How Much Information Flows? Local-Bias Effect and Common-Pool Effect

Lemma 3 has shown that the credibility of the shared information and, by extension, the centralized equilibrium policy decision is determined by two exogenous variables: the extent of the spillovers and the rule gov- erning cost sharing. To develop some understanding of the mechanisms underlying ðcredibleÞ information sharing, consider si 5 1; that is, the district “pays the piper” even though the federal government “calls the tune.” Informative communication then requires ~v 5 ci; that is, the dis- trict proposes to implement its local project under federal authority whenever it would have implemented the project itself under local au- thority. Were the federal authority always to follow the local recommen- dation, the outcome under both governance structures would be iden- tical. If spillovers are significant, however, the federal government will sometimes optimally “overrule” the district, effectively undermining any informative communication. To see this, suppose first that spillovers are negative. Then, the federal

government may want to scrap the project even though the local repre- sentative is in favor. Because federal approval is gained less often than is desired by his constituency, the local representative’s incentives to truth- fully communicate vi are diminished. In particular, the desire to coun- teract the federal reluctance to realize the project by overstating local benefits vi may become sufficiently strong as to render any meaningful communication infeasible. Formally, this happens for ðin absolute termsÞ large values of ji that do not satisfy condition ð5Þ. Similarly, if spillovers are positive, the federal government will want to realize the project more often than is desired by the local population. Again, informative com- munication will become infeasible because at some point the incentive of the local representative to counteract federal activism by understat- ing local benefits vi is too pronounced: condition ð6Þ is violated for suf- ficiently large ðpositiveÞ values of ji. The federal government then ratio- nally ignores the local representative’s information and realizes the project against the expressed will of the local population. Because the motive to overstate ðrespectively, understateÞ the local

value of the project primarily depends on the discrepancy between the

14 It is easy to construct examples to assess how pervasive informed communication ðand, as a result, universalismÞ is in equilibrium. If vi is uniformly distributed on ½0; �vi�, e.g., con- ditions ð5Þ and ð6Þ reduce to ji ≤ cið1 2 12 siÞ ≤ ji 1 12 �vi.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

private value and the social value of the policy, we may call this effect the local-bias effect. Also note that because of the simple structure of the model, the local-bias effect operates only through a threshold value of j

communication in federal politics 781

above ðrespectively, belowÞ which the informative equilibrium ceases to exist. It has no influence on the equilibrium policy as long as informa- tion is transmitted. In practice, of course, a project that is decided on at the federal level

is almost certain to be funded at the federal level as well. In what follows, I will therefore focus on the empirically more relevant case in which projects that are under federal jurisdiction are federally funded as well. Assuming a balanced budget, this de facto means that a jurisdiction will bear a share of the project cost equal to its share of the federal tax base. For example, irrespective of any ðobservable or unobservableÞ differ- ences in the tax bases across districts, cost sharing will necessarily take place whenever the federal budget is at least partly financed by a uni- form tax instrument such as a federal income tax or a federal consump- tion tax. In the simplest case of a central budget financed by a uniform tax on identical tax bases, for instance, we would have si 5 1=n. More generally, I will make the following assumption.15

Assumption 1. The nature of federal taxation is such that costs are de facto shared on the federal level. Moreover, the district tax shares decline in the number of districts. For all districts i 5 1, . . . , n, we have

0 < siðnÞ < 1 and siðnÞ > siðn 1 1Þ ∈ ð0; 1Þ:

Public projects thus have “pork barrel” features: they largely benefit a single district at a cost to all others. These types of projects capture the

important elements of many real-life policies and are commonly as- sumed in the literature on legislative politics ðe.g., Weingast et al. 1981; Grossman and Helpman 2005Þ. Recall from the previous sections that how funds are raised and costs shared under federal authority had no impact on federal policies in the benchmark cases of perfect informa- tion and imperfect information without communication, respectively. Once we allow for communication between the federal government and

15 To see how the local tax burden ti varies with the decision of the local project gi, we can differentiate the federal budget to obtain dti=dgi 5 ci=y ∈ ð0; 1Þ, so siðnÞ ∈ ð0; 1Þ follows

directly from previous assumptions. Naturally, one can also think of situations in which siðnÞ is negative or exceeds one. For example, the federal government may promise a greater federal share in future local spending if the district agrees to a project that has high positive spillovers and low local returns. Similarly, there may be ðimplicitÞ penalties in- volved if districts push for projects with large negative spillovers. But this requires project- ðdistrict-Þ specific subsidies or penalties that dominate a district’s share in the federal tax base that naturally results from uniform federal taxes such as income, payroll, or con- sumption taxes. In either case, allowing for si < 0 or si > 1 would not alter the results significantly.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

local officials ðamong local delegates in a federal assembly, respectivelyÞ, this is no longer true: in conjunction with the lack of information on the part of the federal government, cost sharing now creates a common-pool

782 journal of political economy

problem. Project costs are not entirely borne by the local population, which leads local representatives to underestimate the cost of the proj- ect, ceteris paribus. This in turn creates an incentive to overstate the lo- cal benefits vis-à-vis the federal authority, again affecting the transmis- sion of credible information.16

How the common-pool effect and the local-bias effect work together is summarized in theorem 1. Theorem 1. Consider any informative communication equilibrium.

For parameter values ji < ð1 2 siÞci, the federal policy gi is characterized by overprovision, and this tendency will grow with the number of dis- tricts. For values ji > ð1 2 siÞci, the federal policy gi is characterized by underprovision. This tendency is mitigated as the number of districts increases to some optimal n*, which minimizes the loss of information and grows again thereafter. If ji 5 ð1 2 siÞci, communication is perfect and the federal policy decision gi is socially optimal. The theorem states how informative the communication between a

central and a local government on the local policy depends on the ex- tent of the policy’s spillovers and on how costs are shared at the federal level. To understand the result intuitively, suppose first that ji < ð1 2 siÞci. The informative communication equilibrium ðassuming it existsÞ is then characterized by overprovision since the project is realized whenever vi ≥ sici but should be realized only for values vi ≥ ci 2 ji > sici. Adding additional districts implies s 0i < si and clearly makes matters worse: an already overprovided public project is prone to be even more overpro- vided as the cost share of the constituency declines and the incentive to overstate its value increases further. Hence, the combination of imper- fect communication of privately held information and a common-pool problem endogenously generates diseconomies of ðorganizationalÞ scale: the more districts there are, the more difficult it becomes to truthfully communicate a project’s true benefits in the political process, and the more distorted the resulting policy decision will be. The consequences are best seen in the limit case in which si → 0. From ð4Þ, if informative communication is feasible at all, we must have g ICi ðviÞ 5 1 almost always under federal authority. Intuitively, since the local district’s share of the cost is almost nil, it has a strong incentive to overstate local benefits in order to persuade the federal government to realize the project. But we know that the latter always approves whenever it listens to the former, which is necessarily the case if communication is meaningful. For values

16 In a more general formulation with variable project size, this effect holds whenever locally earmarked expenditures grow more rapidly with project scale than local taxes.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

ji > ð1 2 siÞci, on the other hand, the local-bias effect is sufficiently strong that the informative communication equilibrium is characterized by un- derprovision: the project is realized for values v ≥ s c > c 2 j . The local-

communication in federal politics 783

i i i i i

bias effect and the common-pool effect then work in opposite directions, and which effect dominates will depend on the extent of the local bias ðjiÞ and on the degree of cost sharing ðthe number of districts nÞ. In particular, additional districts now have the benefit of counteracting the underprovision problem: more information will flow and the decision will be more efficient as we increase the number of districts from n to n 1 1 ðignoring the integer problemÞ. Eventually, though, the common- pool problem will dominate, assuming si → 0 as n grows sufficiently large. In such a situation, there obviously is an optimal organizational size n* that balances the common-pool effect with the local-bias effect and thus minimizes the loss of information under federal authority. Theorem 1 also points to a case in which federally chosen policies are

efficient, namely, if ji 5 ð1 2 siÞci or si 5 1 2 ji=ci. Note that this value of the cost share parameter corresponds to what is commonly known as the Clarke-Groves mechanism, which induces truth telling in dominant strategies. As explained earlier ðsee n. 12Þ, the problem with setting the cost shares optimally is that this requires full commitment and complete contracting on the part of all parties involved. In contrast, we adopt the incomplete contracting approach by assuming that by its very nature, the central authority cannot commit to any policy that does not ex post maximize its objective ðrecall that the Clarke-Groves mechanism is not balanced and, hence, not renegotiation-proofÞ. It is difficult to envision the federal government writing binding contracts on policies it has budget authority over in the absence of an external enforcement mechanism. Also, a government may not be able to fully bind its successor. An entirely different, but perhaps equally compelling, argument is that optimally leg- islated tax shares will necessarily be both project specific and district spe- cific. As local needs differ and potential projects change frequently ðas reflected in annual budgetsÞ, the tax code must vary across districts and over time, which could give rise to considerable transaction costs: non- uniform taxation will result in inefficient reallocation of the tax base, and frequent changes create uncertainty hindering investment.

C. Assigning Fiscal Authority

I now briefly address how the local-bias effect and the common-pool

effect together translate into the overall efficiency of policies chosen on a federal level and then compare the outcome under federal authority with that under local authority. Note that the comparison is not trivial: as we saw earlier, for instance, communication under central authority im- proves as the local bias shrinks. Thus we would expect centralization to

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

do well for small ðabsoluteÞ values of ji. At the same time, though, the disadvantage of decentralized authority also disappears for small values of j . Which effect dominates is thus not a priori clear. The analysis will

784 journal of political economy

shed light on which governance structure ðlocal vs. federal authorityÞ the population of an economy would decide on. In doing so, I will assume that this choice over fiscal authority is made so as to maximize social surplus and thus coincides with the objective of the federal government. Suppose first that the project under consideration has negative spill-

overs. Projects in this category include investments with the potential for environmental damage such as water dams or power plants or in- frastructure that intensifies intrajurisdictional competition ðe.g., by at- tracting mobile capitalÞ. In this situation, the local-bias effect is positive; that is, the local representative will want to overstate the project’s ben- efits even if the district is financially fully responsible, si 5 1. Irrespective of how much costs are shared on the federal level, the common-pool and the local-bias effects thus work in the same direction, namely, to generate a bias toward overprovision. We find the following theorem. Theorem 2 ðNegative spilloversÞ. Suppose that ji < 0. Then, local

authority over policy gi is socially preferred if jjij ≤ Efvi 2 cijvi ≥ cig and federal authority over gi is socially preferred otherwise. Moreover, when- ever federal authority is socially preferable, the policy decision will be made under ignorance and is characterized by underprovision. Indeed, the project will not be undertaken. Assuming that the surplus-maximizing governance structure is cho-

sen and recalling the discussion at the end of Section II.A, the following corollary immediately follows. Corollary. Policy decisions with negative spillovers over which fed-

eral authority is socially preferred will not distinguish between ex ante indistinguishable districts: all such policies will be uniform. Next, consider policies with positive spillovers, for example, projects

that improve health care, foster education, or protect the environment. In this situation, the local-bias effect is negative; that is, the local rep- resentative will want to understate the project’s benefits if the district is financially fully responsible, si 5 1. The common-pool and local-bias ef- fects now work in opposite directions. Theorem 3 ðPositive spilloversÞ. Suppose that ji > 0. Then local au-

thority over gi is socially preferred if ji ≤ 2Efvi 2 cijsici ≤ vi ≤ cig and fed- eral authority is preferred otherwise. Moreover, the federal policy deci- sion is made under ignorance for all values ji > 2Efvi 2 cij0 ≤ vi ≤ sicig. In this case, the federal policy decisions are characterized by overprovision. Again assuming that the socially preferred governance is chosen, we

can make the following conclusion. Corollary. Policy decisions with positive spillovers over which fed-

eral authority is socially preferred will not distinguish between ex ante

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

indistinguishable districts if either the number of districts is small or spillovers are pronounced. To summarize, the model yields the following conclusions. First, com-

communication in federal politics 785

munication between the federal government and its local counterparts necessarily remains incomplete. The resulting inefficiency of federal spending decisions is determined by the interplay between the local-bias effect and the common-pool effect, which in turn depend on the ex- tent of local spillovers ji and the number of districts n. In particular, centralized political decision making results in overspending in policy domains with negative or small external effects, whereas, too, under- spending persists in policy domains with large external effects. Second, whenever meaningful communication is feasible, federal spending de- cisions will be characterized by a form of universalism: each representa- tive asking for a project gets the project approved. Third, if meaningful communication is not feasible, federal policy decisions are insensitive to local preferences: federal policies are uniform; that is, all districts are treated alike, and either all projects are funded or none, possibly against the expressed will of the local population. In either case, the central gov- ernment’s information about the local consequences of its policies is in- complete. Similarly—but not identically—to Oates’s ð1972Þ original argument, I

thus identify a trade-off between a loss of coordination under local au- thority and a loss of information under federal authority, where the lat- ter becomes more severe the larger the number of districts. Theorems 2 and 3 relate to this trade-off: from society’s point of view, local policy decisions are preferable whenever the associated externalities are small and whenever the number of districts ðor preference heterogeneity, as measured by the variance of viÞ is large. Moreover, optimal federal au- thority necessarily involves uniformity of public good provision for all policies with negative spillovers. Akin to Oates, therefore, federal author- ity is preferred when externalities are large in absolute terms. At the same time, however, the nature of the externality matters: while ðsmallÞ positive externalities can result in “informed” decisions at the federal level, and therefore Pareto-superior outcomes compared to the ex ante efficient ðuninformedÞ federal decision, the same is not true with negative ex- ternalities. Also, if authority were optimally assigned in practice, we should observe “universalism” only if spillovers are positive and the de- gree of spillovers or the cost share parameter is not too large.

D. Empirical Implications

Some of these results are worth investigating empirically. To derive com-

parative statics properties of the model that can be linked to available data on expenditures of US municipalities, I will take central authority

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

for granted and introduce some heterogeneity in the model by sup- posing that the central government has a continuum of potential proj- ects for each district, identified by their project cost c ∈ ½cmin; cmax�, omit-

786 journal of political economy

ting the index i for ease of notation. It is helpful to distinguish between “public good” projects with j > 0 and “pork” projects with j 5 0 ðthe case in which j < 0 is analogous to the pure pork caseÞ. Also, since the data do not allow for an adequate measure of the efficacy of government spending, the focus here will be on dollar expenditures. First consider figure 2A, which illustrates the situation for public good

projects ðj > 0Þ. The figure is based on the case in which an informative equilibrium exists for all values of c and n; in particular, it assumes that public good projects are ex ante efficient ðE½v� 1 j > cmaxÞ, which ensures that for s → 0 ðn → `Þ, meaningful information transmission is still fea- sible even if costs are high. Because credible information always flows, projects are realized whenever v ≥ sðnÞc, which occurs with probability 1 2 F ðsðnÞcÞ. When we aggregate over c, total expenditures on public goods Epg are given by the area underneath 1 2 F ðsðnÞcÞ. An increase in the number of legislators increases ðdistrictsÞ from n to n 1 1, and a corresponding drop in the local jurisdiction’s cost share from sðnÞ to sðn 1 1Þ causes 1 2 F ð�Þ to shift upward by a factor proportional to c: the budget on public goods expands by an amount DEpg , the area shaded in dark gray. Next, consider pork projects ðj 5 0Þ. Again, if the informative equi-

librium exists, spending on pork Epk is equal to the area underneath 1 2 F ðsðnÞcÞ and will increase in n ceteris paribus. Assuming that high- cost pork projects are not ex ante efficient ðE½v� ≤ cmaxÞ, however, cred- ible information will not flow for sufficiently high project costs and large n, which will prompt the federal authority to drop projects for which E½vjv ≥ sðnÞc� < c. In this case, meaningful communication re- quires that c is below some threshold �cðnÞ implicitly defined by E½vjv ≥ sðnÞ�c� 5 �c, with �cðnÞ decreasing in n.17 Figure 2B illustrates this case. Assuming that local benefits for public goods and pork projects are drawn from the same distribution and comparing Epg with Epk, we see that the total budget is strictly larger for public goods than for pork ðthis is also true in the data, as we will see belowÞ. Moreover, an increase in n now has two effects: it increases the likelihood of project realization of all projects that are still “on the agenda,” that is, for which c ≤ �c holds. At the same time, it lowers �c, as the highest-cost projects are no longer realized since meaningful communication became impossible. The over- all effect DEpk is ambiguous; it depends on whether the darker gray area is

17 Depending on F ðvÞ, cðnÞ may not be unique, in which case the argument is more cumbersome but still valid.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

larger or smaller than the lighter gray area. For smaller n, or if E½v� > cmax in contrast, we will have �c > c max, and the second effect ðthe gray areaÞ disappears, yielding an unambiguous increase in the pork budget, sim

FIG. 2.—A, Public good expenditures; B, pork expenditures

monotone, since some public good projects will also be discontinued as n increases. There is some n̂, however, beyond which no further public good projects are discontinued, a which point the above argument applies.

communication in federal politics 787

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

ilar to figure 2A. To summarize, under the above assumptions the comparative statics

of an increase in the number of legislators n is as follows: expenditures on public good projects will increase, while the effect on pork spending is positive for small n and ambiguous for large n since some high-cost pork projects will be dropped from the agenda if E½v� < cmax. In either case, the increase in public good spending always weekly exceeds the increase in pork spending, and strictly so if �cðnÞ < cmax ðn largeÞ.18 Finally, total government expenditures increase if the rise in public good spend- ing dominates the reduction ðif anyÞ in pork spending. This will generally depend on the distribution function and other parameters of the model. For the special case in which F ð�Þ is the uniform distribution and assum- ing cmin 5 0, cmax 5 �v, and s 5 1=n, for example, it is straightforward to show that both pork spending and total spending unambiguously in- crease in n ðthe share of pork spending in the overall budget, however, still declines in nÞ. Of course, how government spending and its com- position vary with legislative size is ultimately an empirical question, which I address in the following section.

18 If some public good projects are ex ante inefficient, E½v� 1 j < cmax, the effect of n on the public good spending and relative size of public good to pork spending is no longer

III. Empirical Analysis

A. Data Description

788 journal of political economy

In this section, I will bring some of the empirical implications of the theory to data from a cross section of city governments in the United States. These municipalities exhibit substantial variation in terms of in- stitutional structures and policy outcomes and have several advantages over cross-country data. First, there are more observations than in typi- cal cross-country studies. Second, to estimate the effect of institutions on policies, one needs to assume that the units of observation exhibit the same functional relationship between the institutional variable in ques- tion ðin our case, the size of their legislatureÞ and policy outcomes, which is more plausible in the case of subnational units that share a common historical and institutional background. Countries are arguably much more heterogeneous. For this reason, omitted variables are also likely to be endemic to cross-country empirical work. Countries may be different for a variety of reasons, and since it is impossible to control for these dif- ferences in practice, the observed correlations may arise because coun- tries with different historical or economic backgrounds systematically dif- fer in their institutions and their spending patterns. Exploiting variation within a country, in contrast, decreases the danger of unobserved het- erogeneity biasing the results. Finally, of course, local jurisdictions are in- teresting in their own right because they affect economic outcomes and policies. The data are drawn from three different sources, combining informa-

tion about political institutions ðform of governmentÞ, population char- acteristics ðmunicipal demographicsÞ, and public finance ðcity spendingÞ. The data on form of government are taken from a survey that is con- ducted every 5 years by the International City Management Association ðICMAÞ; the analysis focuses mainly on 2001, but I will also use infor- mation from previous years to account for possible changes in govern- ment form over time.19 The information on municipal finances comes from the 2002 Census of Governments conducted by the US Bureau of the Census, which also provides the data on city demographics from the 2000 Census of Population. The measure on government spending I use is general municipal expenditures per capita. In the data, this is broken down into a multitude of different categories ranging from expenditures on fire protection, police protection, roads and transit services, parks and recreation, utilities, and the like to budgetary items such as debt service, airport maintenance, liquor stores, insurance, and so forth. A full de- scription of all categories is available from the Census of Governments

19 Surveys are sent to roughly 7,000 municipalities, with a response rate ranging from 50 to 70 percent.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

classification manual provided by the Census Bureau.20 Table 1 provides summary statistics of the data. As I explain below, some of these items can be categorized fairly easily

communication in federal politics 789

into measures of spending on “pork” versus “public goods.” The first type comprises goods and services for which legislators are likely to see the local benefits to their constituents outweigh the local cost ðassuming that they are shared with the population as wholeÞ. At the municipal level, examples are parks and recreational facilities such as swimming pools as well as public libraries. The second type includes those goods and ser- vices for which either there is an aggregate ðcitywideÞ benefit that jus- tifies their provision or costs are locally concentrated, or both. These types of services include police stations, fire houses, as well as expendi- tures on projects that municipal planners classify as “locally unwanted land uses” and that are commonly known as “not in my backyard” de- velopments. Examples of the latter would be landfills and incinerators, power plants, halfway houses and other correctional facilities, hospitals, and public transit stations. The empirical work to follow investigates the relationship between the

number of city councilors ðlegislatorsÞ and the amount as well as com- position of city spending, using a number of other plausible determi- nants of municipal expenditures as control variables.

B. Results

1. Government Spending

The first question that emerges from the theory is whether or not the well-established “law of 1/n” holds, which states that an increase in the number of legislators should be associated with higher per capita spend- ing. Recall from the previous section that this relationship is consistent with the model’s prediction, provided that public goods “dominate” the budget. As one can see from table 1, in the data, both the median and mean ratios of pork to public good spending are below one. Accordingly, I estimate a regression of the form

ln ðyiÞ 5 a 1 blnðCsizeiÞ 1 gXi 1 ei; ð7Þ where yi is per capita government expenditures in municipality i, Csizei is the number of councilors in this municipality’s legislative body, and Xi is a vector of controls that mainly comprises demographic characteristics but also includes variables that are important for the institutional envi- ronment the municipal governments operate in.

20 Available at the US Census Bureau website for federal, state, and local government statistics ðhttp://www2.census.gov/govs/pubs/classification/2006_classification_manual .pdfhttp://www2.census.gov/govs/pubs/classification/Þ.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

T A B L E 1

S u m m a r y S t a t i s t i c s

V a ri a b le

N M in im

u m

M ax im

u m

M e a n

M e d ia n

S ta n d a rd

D e vi a ti o n

U n it

E x p e n d it u re s p e r ca p it a

1 ,4 3 1

.1 4

9 .7 4

1 .3 9

1 .1 2

.9 6

$ 1 ,0 0 0 p e r p e rs o n

R a ti o p o rk

o ve r p u b li c g o o d s e x p e n d it u re s

1 ,3 9 9

0 5 4 .5 7

.4 2 3

.2 8 4

1 .5 1

R at io

P o rk

e x p e n d it u re s p e r ca p it a

1 ,4 0 1

0 .8 8 3

.1 0 3

.0 8 2

.0 8 9

$ 1 ,0 0 0 p e r p e rs o n

P u b li c g o o d e x p e n d it u re s p e r ca p it a

1 ,4 0 1

0 2 .0 5

.3 0 7

.2 8 3

.3 6 8

$ 1 ,0 0 0 p e r p e rs o n

P o p u la ti o n si ze

1 ,4 2 9

1 0 ,0 1 8

3 ,6 9 4 ,8 3 4

5 2 ,9 3 3

2 5 ,2 1 6

1 3 8 ,4 6 6

N u m b e r o f p e o p le

P o p u la ti o n d e n si ty

1 ,4 2 9

7 3 .8 5

5 6 ,2 7 8 .9 5

2 ,9 7 7 .9 4

2 ,9 9 7 .2 3

2 ,8 3 3 .2 4

P e o p le / sq u a re

m il e

In co

m e p e r ca p it a

1 ,4 2 9

7 ,4 0 8

1 0 9 ,2 1 9

2 2 ,5 4 5

1 9 ,9 7 1

9 ,6 6 6 .7 8

$ p e r p e rs o n

P o p u la ti o n w it h h ig h sc h o o l d ip lo m a

1 ,4 2 9

.2 1 5

.9 8 8

.8 0 9

.8 2 4

.1 0 9

P e rc e n ta g e

P o p u la ti o n o ve r 6 5 ye a rs

o f a g e

1 ,4 2 9

.0 2 5

.5 7 4

.1 3 1

.1 2 9

.0 5 3

P e rc e n ta g e

O w n e r- o cc u p ie d h o u se s

1 ,4 2 9

.2 1 5

.9 8 6

.6 4 1

.6 3 3

.1 3 7

P e rc e n ta g e

E th n ic

h o m o g e n e it y

1 ,4 2 9

.2 3 5

.9 7 1

.6 5 4

.6 6 0

.1 8 5

R at io

M e d ia n to

m e an

in co

m e

1 ,4 2 9

.4 5 9

.9 4 9

.7 8 9 4

.7 9 7

.7 3 5

R at io

C o u n ci l si ze

1 ,4 2 6

3 4 0

6 .7 6

7 2 .2 5

N u m b e r o f p e o p le

C o u n ci l si ze

in 1 9 7 6

5 6 8

3 2 5

6 .6 6

6 2 .4 0

N u m b e r o f p e o p le

W a rd

e le ct io n s

1 ,3 9 3

0 1

.4 0 1

0 .4 9 0

In d ic a to r va ri a b le

M a yo r- co

u n ci l fo rm

o f g o ve rn m e n t

1 ,4 2 9

0 1

.3 3 6

0 .4 7 3

In d ic a to r va ri a b le

N o t e . — T h e d a ta

a re

fr o m

th e 2 0 0 2 C e n su s o f G o ve rn m e n ts , th e 2 0 0 1 IC

M A

fo rm

o f g o ve rn m e n t su rv e y, a n d th e 2 0 0 0 C e n su s o f P o p u la ti o n . T o

fa ci li ta te

co m p a ri so n w it h e ar li e r st u d ie s ðB

a q ir 2 0 0 2 Þ, ci ti e s b e lo w 1 0 ,0 0 0 p o p u la ti o n a re

d ro p p e d fr o m

th e sa m p le . T h e in d e x o f e th n ic h o m o g e n e it y is

d e fi n e d a s E th n 5 o

is 2 i ∈ f0

;1 g,

w h e re

s i d e n o te s th e sh a re

o f th e p o p u la ti o n o f ra ce

i in

th e to ta l p o p u la ti o n o f th e ci ty , a n d i ∈ fw

h it e , b la ck , A m e ri ca n

In d ia n , H is p an

ic , A si a n a n d P a ci fi c Is la n d e r, o th e rg . A ci ty is cl a ss ifi e d a s h a vi n g w a rd

e le ct io n s if o ve r h a lf o f th e co

u n ci lo rs

a re

e le ct e d in

d is tr ic ts a n d a s

“a t la rg e ” o th e rw

is e .

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

The baseline specification in table 2 reports the resulting parameter estimates of ordinary least squares ðOLSÞ regressions; heteroskedasticity- robust standard errors are in parentheses. The explanatory variables are

communication in federal politics 791

population size, density, per capita income, educational attainment ðthe percentage of residents with a high school diplomaÞ, population age structure ðthe percentage of the population over age 65Þ, and the frac- tion of houses that are occupied by their owners as opposed to renters. All control variables except the fractions are in log form to facilitate the interpretation as elasticities. Consistent with the theory, an increase in council size is associated

with higher government spending per capita. The estimated effect is not only statistically significant but also qualitatively important. The point estimate of the coefficient is b 5 0:2 percent, which for the median city with seven councilors would roughly translate into 6 percent higher per capita spending if the size of the council increased from seven to nine members. Given average spending of $1,400 in 2000 for the cities in the sample, this amounts to an increase of $84 in annual expendi- tures per inhabitant. The result confirms Baqir’s ð2002Þ earlier finding, indicating that the

relationship between spending and the size of the legislature is stable over time.21 The effects of the other covariates are largely as one would expect and are highly significant. Government size increases with population size, average income, and the percentage of people over age 65 and de- creases with population density and the fraction of owner-occupied hous- ing. Perhaps surprisingly, a better-educated population is associated with lower government spending if one controls for income ðonce the income variable is dropped from the regression, the coefficient becomes positive and significantÞ. To check how sensitive the results are, I ran additional regressions al-

lowing for further controls that are likely to affect government spending and may be correlated with council size: income equality as measured by the ratio of median to mean household income; an index of ethnic ho- mogeneity; the city’s electoral system, namely, whether candidates are elected from the entire city ð“at large”Þ or from districts ð“ward”Þ; and the city form of government, which determines whether the mayor has sub- stantive authority over the budget ðin the “mayor-council” formÞ or not ðin the “council-manager” formÞ. The first two variables capture demo- graphic heterogeneity in the local electorate. This could be important if

21 The point estimate is somewhat smaller than Baqir’s estimate of .302, which can be explained by the presence of two additional controls in the above regression, namely, population density and the fraction of owner-occupied housing. The latter in particular is

highly significant ðit accounts for almost half of the explained variation in spendingÞ and is negatively correlated with council size. Excluding both explanatory variables from the regression results in a coefficient of b 5 :32.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

bigger councils are associated with a more ethnically diverse population structure, which in turn could cause government spending to be elevated The second two variables consider further institutional factors. Since

TABLE 2 OLS Regressions for Government Size

Dependent Variable: Spending per Capita

Independent Variable ð1Þ ð2Þ ð3Þ ð4Þ Council size .223*** .215*** .158** .174**

ð.052Þ ð.061Þ ð.054Þ ð.085Þ Population size .079*** .089*** .079*** .079***

ð.016Þ ð.018Þ ð.011Þ ð.015Þ Density 2.228*** 2.232*** 2.192*** 2.201***

ð.019Þ ð.021Þ ð.043Þ ð.031Þ Income per capita .531*** .588*** .574*** .513***

ð.067Þ ð.073Þ ð.109Þ ð.096Þ % high school diploma 2.568*** 2.783*** 2.839** 2.735***

ð.195Þ ð.222Þ ð.368Þ ð.255Þ % of population over 65 1.58*** 1.51*** 1.32*** 1.43***

ð.239Þ ð.262Þ ð.245Þ ð.267Þ % owner-occupied housing 21.67*** 21.82*** 21.66*** 21.53***

ð.124Þ ð.148Þ ð.208Þ ð.186Þ Median/mean income .163 2.009 2.134

ð.175Þ ð.262Þ ð.279Þ Ethnic homogeneity .199** .069 2.012

ð.099Þ ð.108Þ ð.125Þ Ward elections .006 2.007 2.025

ð.003Þ ð.039Þ ð.033Þ Strong mayor 2.036 2.034 2.057

ð.033Þ ð.030Þ ð.038Þ Regional fixed effects No No Yes No State fixed effects No No No Yes Observations 1,427 1,387 1,387 1,394 R2 .22 .26 .26 .32

Note.—Dependent variable is total municipal spending per capita. The standard error reported in parentheses are heteroskedasticity robust and are clustered at the state ðre spectively, regionalÞ level for the regressions with fixed effects. * Significant at 10 percent. ** Significant at 5 percent. *** Significant at 1 percent.

792 journal of political economy

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

s -

ward cities tend to have more councilors, on average, failing to control for the electoral system may lead to spurious results if the conventional wis- dom that cities with a ward-based electoral system spend more than cit- ies in which legislators are elected at large holds. Similarly, the mayor— who is generally chosen in citywide elections—has a veto over the budget in mayor-council, and one would therefore expect spending to be lower under this form of government ðsee Coate and Knight 2011Þ, which could bias the estimates on the city council effect since there are systematically more council members in cities with strong mayors than in cities with weak mayors.

Column 2 of table 2 reports the corresponding results.22 We see that the estimated coefficient on the council size remains largely unaffected, indicating in particular that the positive council size effect is going nei-

communication in federal politics 793

ther through an effect of local heterogeneity in the composition of the electorate nor through institutional characteristics that shape municipal policy making. Notably, the reported regressions do not include intergovernmental

revenue ðtransfers from higher-level governmentsÞ. Those are positively correlated with council size in the data and thus increase its estimated effect on spending. Whether or not one would want to control for inter- governmental transfers is largely a question of what one seeks to measure. In particular, it is not obvious that this correlation necessarily reflects some connection between the size of the transfer and an unobserved municipal characteristic that itself is correlated with council size, which would bias the coefficient estimate upward. Rather, an equally likely ex- planation is that a large council in line with the model desires large ex- penditures for, say, large public infrastructure projects ðbridges, highways, public transit, etc.Þ and for that very purpose seeks auxiliary resources from higher-level governments.23

The regressions in columns 3 and 4 add either regional or state fixed effects, respectively. The regions considered correspond to the US Cen- sus Bureau’s nine “divisions” within the continental United States.24 Re- gional fixed effects are worth studying if unobserved regional spending preferences are correlated with preferences for smaller governments. In comparison to the state fixed-effects model, this specification has the advantage that the minimum number of observations on the division level is considerably larger than at the state level, which improves the pre- cision of the estimates. Of course, state-specific factors may also be impor-

22 The finding that both income inequality and ethnic homogeneity are positively as- sociated with government spending can largely be attributed to holding the fraction of owner-occupied houses constant: the latter is positively correlated with either measure of

homogeneity in the population, and once it is dropped from the regression, the coeffi- cients become negative ðand significantly so for income equalityÞ, suggesting that more homogeneous population structures translate into fewer government expenses, on aver- age, as one would expect. The effect of the ward variable is not significantly different from zero, and the data thus do not confirm the expectation that city councils whose members are predominantly elected in districts spend significantly more than cities in which council members are predominantly elected at large.

23 As an additional robustness check, I carried out all regressions in table 2 and table 3 accounting for intergovernmental transfers to the municipal government. The results are encouraging in that the main coefficient of interest, namely, the effect of council size, remains significant, although the size of the effect is diminished. The regressions are obtainable from the author on request.

24 The Census Bureau groups states into four broad “regions” and divides each of them into two or more “divisions.” The four census regions are Northeast, Midwest, South, and West. The results reported included fixed effects for the nine divisions, which are New England, Mid-Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain, and Pacific.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

tant: state laws will generally affect both the extent of municipal spend- ing ðe.g., through the assigned number of municipal functionsÞ and the local governance structure. In both regressions, we see that the estimated

794 journal of political economy

coefficient on council size drops in value but remains positive and sig- nificant.

2. Composition of Government Spending

The theory predicts that an increase in the number of legislators ðelec- toral districts, respectivelyÞ should be associated with a relatively higher increase in public good expenditures than in pork expenditures, ceteris paribus. To the best of my knowledge, this prediction is unique to the present model and has never been investigated empirically before. The regression model is

ln ðriÞ 5 a 1 blnðCsizeiÞ 1 gXi 1 ei; ð8Þ where ri is the log ratio of government outlays on pork relative to spend- ing on public goods in municipality i, Csizei is the number of councilors in this municipality’s legislative body, and Xi is the vector of controls. The specific spending categories I used to account for spending on particu- laristic goods ðporkÞ were general expenditures on public libraries, parks, and recreational/cultural-scientific facilities ðthe latter include everything from golf courses, playgrounds, and swimming pools to stadiums, mu- seums, zoos, and celebrations, including public support of cultural activ- itiesÞ. The functional categories that constitute the spending on public goods are general expenditures on fire and police protection, public tran- sit, and direct expenditures on construction in sewage and solid waste management.25

The findings are reported in table 3 with the same set of controls as in the previous section. The dependent variable in the first set of regres- sions is the ðloggedÞ ratio of spending on pork to spending on public goods. The results from the baseline specification in column 1 strongly support the theory: the relationship between council size and the ratio of spending on pork relative to public goods is negative and highly sig- nificant; that is, cities with larger councils spend relatively more on pub- lic goods than on pork, ceteris paribus. We also see that the signs of the coefficients on the controls are largely as expected, which again suggests that the model is not grossly misspecified: larger cities spend more on pork relative to public goods while denser cities spend less ðthough the effect of density is only marginally significantÞ. The fraction of the pop-

25 The categorization was driven primarily by data availability and common sense. I should emphasize, however, that the results that follow are not sensitive to including more spending functions into any of the two broad groups of expenditures or to excluding any

one of the categories mentioned above.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

ulation with a high school diploma is an important predictor for the pat tern of spending, as is homeownership: an electorate that is more edu cated and owns rather than rents demands, on average, more pork ðparks

TABLE 3 OLS Regressions for Composition of Government Expenditures

Dependent Variable

Ratio of Pork over Public Goods Pork ð5Þ

Public Goods ð6ÞIndependent Variable ð1Þ ð2Þ ð3Þ ð4Þ

Council size 2.529*** 2.495*** 2.263** 2.225** 2.103 .369*** ð.909Þ ð.111Þ ð.118Þ ð.112Þ ð.102Þ ð.065Þ

Population size .096*** .111*** .068** .048 .213*** .107*** ð.027Þ ð.029Þ ð.029Þ ð.030Þ ð.030Þ ð.015Þ

Density 2.078** 2.063** 2.029 2.031 2.142*** 2.086*** ð.035Þ ð.036Þ ð.038Þ ð.043Þ ð.035Þ ð.023Þ

Income per capita 2.014 2.001 .204** .125* .572*** .587*** ð.106Þ ð.121Þ ð.137Þ ð.151Þ ð.127Þ ð.772Þ

% high school diploma 2.04*** 1.94*** 1.32*** 1.42*** 1.04*** 2.909*** ð.351Þ ð.351Þ ð.383Þ ð.426Þ ð.398Þ ð.212Þ

% of population over 65 2.377 2.622 2.432 2.632 .832** 1.66 ð.474Þ ð.481Þ ð.501Þ ð.548Þ ð.486Þ ð.262Þ

% owner-occupied housing .533** .532** .296 .423 2.762*** 2.134***

ð.211Þ ð.211Þ ð.237Þ ð.302Þ ð.229Þ ð.158Þ Ethnic homogeneity .374** .409** .364* .192 2.174**

ð.160Þ ð.182Þ ð.197Þ ð.169Þ ð.087Þ Ward electoral system .012 2.034 2.016 .008 2.014

ð.052Þ ð.053Þ ð.055Þ ð.054Þ ð.024Þ Strong mayor 2.123** 2.045 2.023 2.147** 2.013

ð.053Þ ð.057Þ ð.065Þ ð.056Þ ð.029Þ Regional fixed effects No No Yes No No No State fixed effects No No No Yes No No Observations 1,380 1,348 1,340 1,348 1,349 1,393 R2 .11 .13 .17 .23 .14 .25

Note.—Dependent variable in cols. 1–4 is total municipal spending in areas categorized as pork relative to municipal spending in functions categorized as public goods, in col. 5 i is per capita spending on pork ðparks, recreational facilities, and librariesÞ, and in col. 6 it i per capita spending on public goods ðpolice and fire protection, public transit, and ex penditures on the construction of sewage and waste management facilitiesÞ. All variable except percentages are in log form. Robust standard errors, in parentheses, are clustered a the state ðregional, respectivelyÞ level in the regressions with fixed effects. * Significant at 10 percent. ** Significant at 5 percent. *** Significant at 1 percent.

communication in federal politics 795

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

- - ,

t s - s t

libraries, etc.Þ. Average income does not affect the composition of the budget in any significant way if one controls for education and home- ownership; it is easily confirmed, though, that dropping the latter vari- ables from the regression leads to a large and significantly positive esti- mated effect of income. Column 2 reports the coefficients from the regression using the full

set of covariates. We see that cities with ethnically more homogeneous

populations tend to spend more on pork relative to public goods, while the opposite is true for cities with a strong mayor.26 The latter effect could be explained by the fact that strong mayors are usually chosen in a

796 journal of political economy

separate citywide election, which gives them an incentive to seek support from broad coalitions of the population, implementing policies with a public good character rather than policies specifically targeted at certain ðgeographically concentratedÞ groups. By the same token, though, cities with ward elections in which a councilor’s constituents live in smaller districts should spend more on pork relative to public goods. A number of recent papers have investigated the relationship between electoral rules and government spending theoretically ðPersson et al. 2000; Liz- zeri and Persico 2001; Milesi-Ferretti et al. 2002Þ and share a common prediction that at-large systems tilt the composition away from pork spending toward spending on public goods benefiting larger groups of the population. The result that the coefficient on the ward variable was insignificant is inconsistent with this view and is similar to the results of Baqir ð2002Þ, who also finds little support for this hypothesis. Again, the effect of the number of councilors on the composition of

municipal spending continues to hold qualitatively if I control for un- observed taste and institutional parameters by including either regional or state fixed effects, respectively. Results from those regressions are re- ported in columns 3 and 4. As before, the magnitude of the coefficient is smaller with regional fixed effects and drops further to about half of its former value in the regression with state fixed effects. The qualitative direction of the effect remains unchanged, though: the size of the city council affects the ratio of pork relative to public goods in a positive way and—together with the effects of education and ethnic homogeneity—is the only explanatory variable that is significant in determining the ob- served expenditure patterns. To assess the estimate quantitatively, note that a point estimate of b 5 2:5 roughly implies a 15 percent drop in the share of pork relative to public goods in government spending if council size increased from a median of seven members to the next- higher level of nine members, clearly a substantial reduction. Finally, as an additional test, columns 5 and 6 report separate regres-

sions for per capita spending in either of the two categories, respectively. The average slope of the relationship between the number of councilors and the level of pork expenditures is negative but not significantly dif- ferent from zero, which is certainly consistent with the theory ðrecall

26 One should note that the former effect is sensitive to the definition of the public goods variable. As an additional sensitivity analysis, I ran all regressions with police ex- penditures excluded from spending on public goods. The respective estimated coefficient

on ethnic homogeneity became markedly smaller and eventually insignificant in the fixed- effects regressions ðnone of the other coefficients were qualitatively affected in this ex- erciseÞ.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

fig. 2BÞ. In contrast, I find a strong and positive relation between council size and public good expenditures: larger council cities spend more on public goods than smaller council cities, all else equal. Again, this result

communication in federal politics 797

is implied by the theory and is strongly significant.

3. Instrumental Variable Regressions

The assumption underlying the identification strategy in the previous sections is that the number of councilors is exogenous. While changing the number of legislators is not a straightforward matter, one concern with the estimates presented above is that cities whose spending pat- terns are different have adjusted the size of their legislative body in dif- ferent ways.27 One could imagine, for instance, that municipalities were subject to shocks to the preferences of the electorate that drove the de- mand for institutional reform, those with stronger ðweakerÞ preferences for a large government or more universalistic public spending increas- ing ðdecreasingÞ their council size over time. More generally, the con- cern is that both the size of the council and spending patterns are cor- related with omitted right-hand-side variables that are not captured by regional or state fixed effects.28

To deal with this issue, I employ an instrumental variable ðIVÞ strategy, using both size of the municipal council in 1976 and the electoral system ðward or at-largeÞ as instruments for the number of legislators in 2001. Not surprisingly, council sizes in 1976 and 2001 are highly correlated ðthe covariance is .74Þ, and given the length of the elapsed time period, it is not unreasonable to presume that any effect that the number of leg- islators 24 year ago has on today’s spending is going through today’s number of legislators ðsee Baqir ½2002� for a similar argumentÞ. The elec- toral system is another potentially valid instrument: whether councilors are elected at large or in wards is fairly strongly correlated with coun- cil size ðthe covariance is .39Þ and had no effect on the amount or the composition of spending in the first set of IV regressions when instru- mented only for council size by council size in 1976. For brevity of exposition, table 4 shows only the estimated coefficient

on the main explanatory variable of interest, namely, the number of leg- islators in 2001 using council size in 1976 and electoral system ðward or

27 The responses in the ICMA questionnaire from 2001 indicate that only about 2 per- cent of cities in the sample have attempted to change the size of their council in the 5 years

leading up to the survey. Out of those, roughly two-thirds saw the attempted change approved ðchanges to the structure and form of municipal government frequently require voter approval, e.g., through a referendumÞ.

28 Municipalities are regulated by state statutes. While some states impose upper limits on the number of councilors in a municipality ðas a function of its population size, e.g.Þ, cities often enjoy considerable freedom in choosing election procedures, number of coun- cilors, and the type of election ðby district vs. at largeÞ.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

at-largeÞ as instruments. The regressions are estimated two-stage least squares ð2SLSÞ of the baseline model and the extended model, where the latter is estimated both with and without fixed effects. The counci

TABLE 4 Estimated Coefficients on Council Size for Different 2SLS Regressions

Dependent Variable and Model Baseline Extended Regional Fixed

Effects State Fixed Effects

Per capita expenditures: OLS .243*** .227*** .296** .087

ð.072Þ ð.078Þ ð.125Þ ð.103Þ

2SLS .361*** .365*** .381*** .142 ð.085Þ ð.095Þ ð.115Þ ð.148Þ

Pork/public good expenditures: OLS 2.686*** 2.697*** 2.395** 2.392**

ð2.132Þ ð.141Þ ð.124Þ ð.169Þ

2SLS 2.719*** 2.719*** 2.404* 2.393 ð.171Þ ð.173Þ ð.219Þ ð.254Þ

Note.—Instruments are the number of councilors in 1976 and the electoral system ðward or at-largeÞ. Number of observations is 546 for the first set and 538 for the second set. Robust standard errors, in parentheses, are clustered at the regional and state levels for the fixed-effects regressions. * Significant at 10 percent. ** Significant at 5 percent. *** Significant at 1 percent.

susceptible to outliers. For instance, the value of the coefficient on council size increase considerably from .87 to .138 if a single observation is deleted. The observation in question is the city of Winfield, Kansas, which has a very small council of only three members in both 1976 and 2001 and spends per capita more than twice as much as the median city in Kansas In the IV regression, the respective change is from .142 to .201.

798 journal of political economy

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

size in 1976 is available only for a smaller number of cities, resulting in 545 observations. To facilitate comparison, the first row therefore rep- licates the original OLS regression for the various specifications of the model with the smaller sample size. With the exception of the last spec- ification, which includes state fixed effects, all estimated coefficients using the smaller sample are very similar to the estimates from the full sample.29

Both instruments have considerable bite: the F -statistic for the first stage is 124, and t-values of 17 for the council size in 1976 and 5 for the ward indicator variable make them the two primary predictors of council size in 2001. Since the model is overidentified, I can also test whether the overidentification restrictions are satisfied. With a Hansen-Sargan test, the null hypothesis that the instruments are jointly valid cannot be re- jected for any of the specifications below.

29 Including fixed effects for 48 states in this smaller sample soaks up a lot of variation in the data and leaves some states with fewer than a handful of observations, which likely contributes to the larger standard errors. The estimated b in these regressions is also highly

The endogenous variable of interest in the first set of regressions is government spending per capita. When I instrument for the 2001 coun- cil size in those regressions, the estimated coefficients increase in mag-

communication in federal politics 799

nitude for all specifications, which is an indication that the original es- timates were too low because of omitted variables. Indeed, a variant of the Durban-Wu-Hausman test ðthat is robust to violations of conditional homoskedasticityÞ shows that exogeneity of council size in 2001 is re- jected for the baseline model at the 10 percent level but cannot be re- jected for any other specification.30

The second set of regressions uses an IV strategy to study the effect of council size on the composition of spending. Here, the magnitude and pattern of the 2SLS point estimates are very similar to those of the OLS estimates, suggesting that endogeneity is not a problem for the compo- sition of the budget. A Durban-Wu-Hausman test confirms this: the hy- pothesis that the OLS estimates are unbiased cannot be rejected. As be- fore, the coefficient drops in magnitude once fixed effects are introduced but remains negative ðthough insignificant for the state fixed effectsÞ. To summarize, the results from the empirical section show that the

size of the legislative body has a significant effect on the size and the composition of government spending. Cities with a larger council spend more per capita on average. At the same time, they tend to spend less on pork barrel projects and more on public goods than cities with a smaller council that are otherwise comparable, leading to a strong and negative relationship between council size and the ratio of pork versus public good expenditures.

IV. Discussion and Concluding Remarks

Federal authority results in political failure in the present model despite

the fact that the legislature’s decision making is efficient; that is, the legislature acts “cooperatively” as in, for example, Weingast et al. ð1981Þ. An alternative and noncooperative approach to the behavior of legis- latures is to assume that central decisions are made by majority rule. This view is adopted by Lockwood ð2002Þ, who studies a model of legislative bargaining in the spirit of Baron and Ferejohn ð1989Þ in which decisions require a minimum winning coalition to form. Besley and Coate ð2003Þ consider both cooperative and noncooperative ðmajority ruleÞ legisla-

30 Baqir ð2002Þ finds qualitatively similar results using 1960 council size to instrument for 1990 council size. One ðplausibleÞ explanation he proposes for the observed increase in the

coefficient is that, over time, cities with growing populations could have taken advantage of economies of scale in government spending ðthus reducing their per capita spendingÞ while at the same time having to increase representation of the city council. This would lead to a downward bias in the estimated cross-sectional relationship between spending and council size in 2000, which could partly be corrected for by the IV approach.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

tures.31 If legislators act noncooperatively and decisions are made by majority rule, centralized policies will be tilted toward those regions that are part of the ruling majority and consistently disadvantage the minor-

800 journal of political economy

ity regions. In the present framework, this effect would add another di- mension of conflict, which likely renders impossible a meaningful com- munication with jurisdictions outside the minimum winning coalition. Otherwise, however, the flavor of the result would be preserved. In par- ticular, as long as the costs of local public goods are still financed through a central budget, information transmission for those jurisdictions that belong to the minimum winning coalition should still be governed by the interplay of the common-pool effect and the local-bias effect.32

The results in the present paper also shed light on the question whether centralization increases the size of government. In Besley and Coate ð2003Þ, regions seek to attract a larger share of central spending by delegating bargaining to representatives with high values for the local public good. In Persson and Tabellini ð1994Þ, local policy makers use contributions to persuade a central legislator to allocate public spend- ing toward their regions; in the noncooperative equilibrium, the central legislator supplies too many public goods. In all these studies, centrali- zation for political decision making results in an overprovision of public goods. In the present paper, this is not true, but rather depends on the local-bias effect. In particular, while centralization increases spending in policy domains in which projects have negative or positive but small spillovers, centralization may actually decrease public spending in pol- icy domains with significant positive spillovers. This finding is consis- tent with casual evidence in the European Union. The European Union’s common agricultural policy, in which spillovers are absent or even neg- ative, is arguably characterized by overspending. In other areas with sig- nificant spillovers such as the environment, in contrast, spending is re- markably low.

31 By incorporating political economy considerations and focusing on the ðdistribu- tionalÞ conflict among regions, these papers thus look at federalism from an angle that is quite different from Oates’s original insight. In contrast, the model in the present paper

pays tribute to the need to explicitly account for the incentive of political actors ðalthough it admittedly does so in a fairly rudimentary wayÞ but is much closer in the spirit of Oates.

32 Legislators will continue to be subject to the common-pool problem since more dis- tricts still imply a lower cost share of one’s own jurisdiction on any given project. Similarly, the motive to manipulate the perceived value of one’s own project will still depend on the discrepancy between that value and whatever value the decisive legislator ðagenda setterÞ places on the project, giving rise to a local-bias effect. While these two main effects are thus likely to survive in a framework with legislative bargaining, it is difficult to assess the full characteristics of any communication equilibrium without an explicit model in mind. Moreover, any information provided may shift the composition of the minimum winning coalition, which will further complicate matters. Also note that direct majority vote is not a suitable model for legislative decision making in this framework of distributive politics. Because of the multidimensionality of the policy vector, a majority vote equilibrium ðCon- dorcet winnerÞ does not exist.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

Finally, in order to keep the analysis concise and tractable, I have made specific assumptions concerning the nature of the public projects and the privately held information. In particular, the informational asym-

communication in federal politics 801

metry related only to a project’s idiosyncratic benefits, not to its cost or its spillovers. Moreover, project costs and benefits were additive separable across districts, implying that the efficient local project choice could be made independently of what happened in other districts. Some of these assumptions could easily be altered. For instance, models in which dis- tricts are better informed about their cost of raising public funds or about the spillover effect of other projects ðrather than about their own project benefitÞ could be analyzed in an analogous fashion, with similar conclusions. Similarly, one could easily allow the federal government to collect additional information at the local level, provided that some re- sidual uncertainty remains.33 It would also be fairly straightforward to study a situation in which the realizations of local benefits are correlated, say, because of some common shock to the economy. How much infor- mation a local representative can credibly transmit in equilibrium would then hinge on what others have already said. My conjecture is that this effect could make sequential communication desirable; depending on the nature of the correlation, the order of speeches would matter and there may well be an optimal sequence of speakers in the assembly that maximizes informational efficiency. Going one step further, one could consider a scenario in which proj-

ect decisions interrelated, for example, because the federal budget is fixed. The communicated information of others then directly affects the social desirability of each project. Again, the incentives of local delegates to truthfully convey their information will depend on the communica- tion strategies of other delegates in equilibrium. Because of the interde- pendency in communication strategies, this case is much more difficult to analyze.34 A formal analysis of these cases is beyond the scope of the pres- ent paper but constitutes a fruitful avenue of future research.

33 Technically, what is most important is that the variable capturing the privately held information continues to be one-dimensional. Obviously, this would still be the case if

districts knew more than the federal authority about the project cost or about the potential spillovers ðas opposed to project benefitsÞ. It is easy to see, then, that communication would still be influenced by the common-pool and the local-bias effects in ways analogous to the current model, yielding qualitatively comparable results. Also, it is straightforward to allow a federal government to gather information on its own, provided that local representatives know which information the government possesses ði.e., there is no two-sided informa- tional asymmetryÞ. In this case, one can just think of the uncertainty regarding v as a residual uncertainty that is present once the federal government has updated its beliefs from its own information ðfor instance, the support of V could have narrowed as a result of federal informationÞ. As long as some residual uncertainty is left, the results will carry through.

34 In a series of examples with three delegates who have received private signals about the desirability of a common project, Austen-Smith ð1990Þ shows that there are two op-

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

Appendix

Proof of Lemma 2

802 journal of political economy

Note that in the equilibrium without information transmission, either gi ; 1 or gi ; 0, irrespective of vi. Now consider an informative equilibrium and a local delegate with information vi ∈ ½0;~viÞ who sends a message mi 5 0 that triggers gi 5 0 in equilibrium. Since this agent can alternatively send a message mi 5 1 that would prompt the central authority to set gi 5 1, he or she must weakly prefer the former, and strictly so whenever vi < ~vi. The converse is true for del- egates with information vi ∈ ½~vi; �vi�. The local districts are therefore better off. By a similar revealed-preference argument, the central authority is also strictly better off. Since information is being transmitted, it changes the decision giðviÞ for some states of the world, which must be strictly better than sticking with the ðuninformedÞ decision that prevails without the information. QED

Proof of Theorem 2

For ji < 0, the project is overprovided under local authority. Theorem 1 tells us

that it is also overprovided under centralization if the informative communica- tion equilibrium exists. But because the project is undertaken even more often in the latter, local authority is strictly preferred. Formally,

DCi 5 Eci2ji si ci

ðvi 1 ji 2 ciÞdFiðviÞ > Eci2ji ci

ðvi 1 ji 2 ciÞdFiðviÞ 5 DLi :

Next, note that local authority is also socially preferred whenever jjj is suffi- ciently small that the federal authority would always realize the project without additional information, that is, if the informative communication equilibrium does not exist because ð6Þ is violated. Hence, if centralization is ever optimal, it must be the case that ðaÞ the federal authority acts under ignorance ðno infor- mation flowsÞ and ðbÞ ð5Þ is violated, which implies gi ; 0. The expected surplus lost under centralization is then

DCi 5 E �vi

ci 2ji

ðvi 1 ji 2 ciÞdFiðviÞ > 0: ðA1Þ

The expected loss in surplus under local authority is

DLi 5 2Eci2ji ci

ðvi 1 ji 2 ciÞdFiðviÞ > 0: ðA2Þ

Comparing ðA1Þ with ðA2Þ, we see that local authority is preferred if

posing effects: on the one hand, delegates have an incentive to share information insofar as their preferences are correlated. On the other hand, information may be withheld because

of diversity in ideal points.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

E �vi

ci2ji

ðvi 1 ji 2 ciÞdFiðviÞ ≥ 2Eci 2ji ci

ðvi 1 ji 2 ciÞdFiðviÞ

communication in federal politics 803

⇔ E �vi

ðvi 1 ji 2 ciÞdFiðviÞ ≥ 0

⇔ E �vi

ðvi 2 ciÞdFiðviÞ ≥ 2ji½1 2 F ðciÞ�

⇔ 2Efvi 2 cijvi ≥ cig ≤ ji:

Finally, note that for ji < 2Efvi 2 cijvi ≥ cig, condition ð5Þ is indeed violated; that is, the informative communication equilibrium under centralization does not exist. For those values of ji we must have g Ci ðviÞ ; 0 for all vi ∈ Vi, which proves the second part of the theorem. QED

Proof of Theorem 3

Consider first parameter values 0 < ji ≤ ð1 2 siÞci. For those values, ð6Þ holds triv-

ially. Hence, if the informative communication equilibrium does not exist in this range, ð5Þ must be violated and we must have giðviÞ ; 0; that is, the project would never be undertaken under federal authority. But then local authority is always preferred because it at least ensures that gi 51 in some states of the world. If the informative communication equilibrium exists in this range, we know from the- orem 1 that the project is undertaken too often. The expected surplus lost is

DCi 5 2Eci 2ji si ci

ðvi 1 ji 2 ciÞdFiðviÞ > 0: ðA3Þ

Conversely, the project is undertaken not often enough under local authority. The corresponding expected loss in surplus is

DLi 5 Eci ci 2ji

ðvi 1 ji 2 ciÞdFiðviÞ > 0: ðA4Þ

When we compare ðA3Þ with ðA4Þ, local authority is preferred if

2Eci 2ji si ci

ðvi 1 ji 2 ciÞdFiðviÞ ≥ Eci ci 2ji

ðvi 1 ji 2 ciÞdFiðviÞ

⇔ 2Eci si ci

ðvi 1 ji 2 ciÞdFiðviÞ ≤ 0

⇔ 2Eci sici

ðvi 2 ciÞdFiðviÞ ≥ ji½F ðciÞ 2 F ðsiciÞ�

⇔ 2Efvi 2 cijsici ≤ vi ≤ cig ≥ ji:

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

Finally, note that for ji > 2Efvi 2 cijsici ≤ vi ≤ cig, both conditions ð5Þ and ð6Þ hold; that is, the informative communication equilibrium under centralization indeed exists ðand the outcome is preferred to that under local authorityÞ.

804 journal of political economy

Next, consider parameter values ð1 2 siÞci < ji. If the informative communi- cation equilibrium exists in this range, theorem 1 tells us that it will be charac- terized by underprovision. But because the project is undertaken even less often under local authority, federal authority is strictly optimal. Formally,

DCi 5 Esi ci ci2ji

ðvi 1 ji 2 ciÞdFiðviÞ < Eci ci2ji

ðvi 1 ji 2 ciÞdFiðviÞ 5 DLi :

If the informative communication equilibrium does not exist for ji > ð1 2 siÞci, we must have ð6Þ violated,

Esici 0

ðvi 1 ji 2 ciÞdFiðviÞ > 0 ⇔ jiF ðsiciÞ > 2 Esi ci 0

ðvi 2 ciÞdFiðviÞ;

ji > 2Efvi 2 cij0 ≤ vi ≤ sicig: Federal authority is still better since

DCi 5 2Eci 2ji 0

ðvi 1 ji 2 ciÞdFiðviÞ < Eci ci 2ji

ðvi 1 ji 2 ciÞdFiðviÞ 5 DLi

for those values of ji. Federal policy decisions will then be made under igno- rance and will be characterized by excessive provision. Finally, note that ji >

2Efvi 2 cij0 ≤ vi ≤ sicig will hold if either j is large or the cost share parameter si is large ðthe number of districts n is smallÞ. QED

References

Austen-Smith, David. 1990. “Information Transmission in Debate.” American J. Polit. Sci. 34:124–52.

Baqir, Reza. 2002. “Districting and Government Overspending.” J.P.E. 110:1318– 54.

Baron, David P., and John Ferejohn. 1989. “Bargaining in Legislatures.” American Polit. Sci. Rev. 83:1181–1206.

Besley, Timothy, and Anne Case. 2003. “Political Institutions and Policy Choices: Evidence from the United States.” J. Econ. Literature 41:7–73.

Besley, Timothy, and Stephen Coate. 2003. “Centralized versus Decentralized Provision of Local Public Goods: A Political Economy Approach.” J. Public Econ. 87:2611–37.

Bradbury, John Charles, and W. Mark Crain. 2001. “Legislative Organization and Government Spending: Cross-Country Evidence.” J. Public Econ. 82:309–25.

Coate, Stephen, and Brian Knight. 2011. “Government Form and Public Spend- ing: Theory and Evidence from U.S. Municipalities.” American Econ. J.: Econ. Policy 3:82–112.

Crawford, Vincent P., and Joel Sobel. 1982. “Strategic Information Transmis- sion.” Econometrica 50:1431–51.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

Dessein, Wouter. 2002. “Authority and Communication in Organizations.” Rev. Econ. Studies 69:811–38.

Egger, Peter, and Marko Koethenbuerger. 2010. “Government Spending and

communication in federal politics 805

Legislative Organization: Quasi-Experimental Evidence from Germany.” Amer- ican Econ. J.: Appl. Econ. 2:200–212.

Epstein, David. 1998. “Partisan and Bipartisan Signaling in Congress.” J. Law, Econ., and Org. 14:183–204.

Faguet, Jean-Paul. 2004. “Does Decentralization Increase Government Respon- siveness to Local Needs? Evidence from Bolivia.” J. Public Econ. 88:867–93.

Gilligan, Thomas W., and Keith Krehbiel. 1989. “Collective Choice without Pro- cedural Commitment.” In Models of Strategic Choice in Politics, edited by Peter C. Ordeshook, 295–314. Ann Arbor: Univ. Michigan Press.

Gilligan, Thomas W., and John G. Matsusaka. 2001. “Fiscal Policy, Legislature Size, and Political Parties: Evidence from State and Local Governments in the First Half of the 20th Century.” Nat. Tax J. 54:57–82.

Grossman, Gene M., and Elhanan Helpman. 2005. “Party Discipline and Pork Barrel Politics.” Working Paper no. 11396, NBER, Cambridge, MA.

Harstad, Bård. 2007. “Harmonization and Side Payments in Political Coopera- tion.” A.E.R. 97:871–89.

Kessler, Anke S., Christoph Lülfesmann, and Gordon Myers. 2009. “Federations, Constitutions, and Political Bargaining.” Discussion Paper no. 4244, Centre Econ. Policy Res., London.

Klibanoff, Peter, and Jonathan Morduch. 1995. “Decentralization, Externalities, and Efficiency.” Rev. Econ. Studies 62:223–47.

Klibanoff, Peter, and Michel Poitevin. 2013. “A Theory of ðDe-ÞCentralization.” Manuscript, Kellogg School Management, Northwestern Univ.

Krishna, Vijay, and John Morgan. 2001. “Asymmetric Information and Legislative Rules: Some Amendments.” American Polit. Sci. Rev. 95:435–52.

Lizzeri, Alessandro, and Nicola Persico. 2001. “The Provision of Public Goods under Alternative Electoral Incentives.” A.E.R. 91:225–39.

Lockwood, Ben. 2002. “Distributive Politics and the Costs of Centralization.” Rev. Econ. Studies 69:313–37.

Milesi-Ferretti, Gian Maria, Roberto Perotti, and Massimo Rostagno. 2002. “Electoral Systems and Public Spending.” Q.J.E. 117:609–57.

Niou, Emerson M. S., and Peter C. Ordeshook. 1985. “Universalism in Con- gress.” American J. Polit. Sci. 29:246–58.

Oates, Wallace E. 1972. Fiscal Federalism. New York: Harcourt. ———. 1999. “An Essay on Fiscal Federalism.” J. Econ. Literature 37:1120–49. Perotti, Robert, and Yianos Kontopoulos. 2002. “Fragmented Fiscal Policy.” J. Public Econ. 86:191–222.

Persson, Torsten, Gerard Roland, and Guido Tabellini. 2000. “Comparative Politics and Public Finance.” J.P.E. 108:1121–61.

Persson, Torsten, and Guido Tabellini. 1994. “Does Centralization Increase the Size of Government?” European Econ. Rev. 38:765–73.

———. 2004. “Constitutional Rules and Fiscal Policy Outcomes.” A.E.R. 94:25– 45.

Primo, David M., and James M. Snyder Jr. 2008. “Distributive Politics and the Law of 1/n.” J. Politics 70:477–86.

Seabright, Paul. 1996. “Accountability and Decentralization in Government: An Incomplete Contracts Model.” European Econ. Rev. 40:61–89.

Weingast, Barry R., Kenneth A. Shepsle, and Christopher Johnson. 1981. “The Political Economy of Benefits and Costs: A Neoclassical Approach to Distrib- utive Politics.” J.P.E. 89:642–64.

This content downloaded from ��140.211.65.60 on Sun, 18 Nov 2018 17:38:07 UTC��

All use subject to https://about.jstor.org/terms

__MACOSX/Economics Resources/._Federalism - Kessler.pdf

Economics Resources/Taxation in the United States - GRUBER File.pdf

__MACOSX/Economics Resources/._Taxation in the United States - GRUBER File.pdf

Economics Resources/Political Economy - GRUBER File.pdf

__MACOSX/Economics Resources/._Political Economy - GRUBER File.pdf

Economics Resources/Cost-Benefit Analysis - GRUBER File.pdf

__MACOSX/Economics Resources/._Cost-Benefit Analysis - GRUBER File.pdf

Economics Resources/STATE RANKINGS 2017 .pdf

__MACOSX/Economics Resources/._STATE RANKINGS 2017 .pdf

Economics Resources/Theory of Collective Action - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Theory of Collective Action - HOLCOMBE File.pdf

Economics Resources/State Fiscal Rankings in 2018 - NORCROSS File.pdf

Ranking the States by Fiscal Condition 2018 Edition

Eileen Norcross and Olivia Gonzalez

MERCATUS RESEARCH

This paper can be accessed at www.mercatus.org/statefiscalrankings.

The views expressed in Mercatus Research are the authors’ and do not represent official positions of the Mercatus Center or George Mason University.

ABSTRACT

For the fifth and final year, we rank states according to their financial condition. On the basis of FY 2016 financial reports of the 50 states, this study ranks the states’ fiscal sol- vency using 13 indicators that assess the extent to which the states can meet their obliga- tions. State finances are analyzed according to five dimensions of solvency: cash, budget, long-run, service-level, and trust fund solvency. These five dimensions are combined to produce an overall ranking of state fiscal solvency. Nebraska, South Dakota, Tennessee, Florida, and Oklahoma rank as the top five most fiscally solvent states. Kentucky, Mas- sachusetts, New Jersey, Connecticut, and Illinois rank as the bottom five states. This ranking highlights the relative performance of the states in one year, but understand- ing financial health requires looking at the underlying objective performance of each state over time. We complement this year’s ranking with a 10-year trend analysis of the states’ financial performance. We find that although, on average, state budgets have not fallen to the lows they reached during the recession, they also have not quite improved to prerecession levels. There has been a slight decline in average state operating ratios since FY 2014, but most states are still able to match revenues with expenses. Long-term liabilities have, on average, increased over time. Long-term liabilities increased the most significantly in FY 2015, largely as a result of new Government Accounting Standards Board rules that require states to report unfunded pension obligations on their balance sheets. Unfunded pension liabilities remain an ongoing problem for the states, and their magnitude is only more transparently revealed by these reporting changes. Pairing these findings with what we have learned from the past four editions of this study, we con- clude with recommendations for future research that emphasize pairing quantitative and qualitative data in context to analyze state financial condition.

JEL codes: H2, H3, H7, M410, M420

Keywords: state fiscal health, financial ratios, state budget, state finance, state debt, public pensions, OPEB, state borrowing, municipal debt, fiscal solvency

Eileen Norcross and Olivia Gonzalez. “Ranking the States by Fiscal Condition, 2018 Edition.” Mercatus Research, Mercatus Center at George Mason University, Arlington, VA, October 2018.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

CONTENTS

1. RANKING THE STATES 8 Cash Solvency Rankings 10 Budget Solvency Rankings 14 Long-Run Solvency Rankings 17 Service-Level Solvency Rankings 17 Trust Fund Solvency Rankings 19 Overall Ranking of the States 23

2. FISCAL CONDITION TRENDS 26 National Trends 26 Fiscal Implications of Heav y Reliance on Oil Tax Revenues 34 Fiscal Implications of Major Tax Reforms 36 States with Pension Problems 41 States with Consistently Strong Fiscal Performance 42 States with Consistently Weak Fiscal Performance 43

3. CONCLUSION 43

APPENDICES Appendix A. Ranking Methodolog y 45 Appendix B. Data Tables 60 Appendix C. State Profiles 77

ACKNOWLEDGMENTS 179

ABOUT THE AUTHORS 179

ABOUT THE MERCATUS CENTER AT GEORGE MASON UNIVERSITY 180

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLES 1. Descriptive Statistics for Fiscal Year 2016 State Government

Financial Indicators 11 2. Ranking of States by Cash Solvency (Fiscal Year 2016) 13 3. Ranking of States by Budget Solvency (Fiscal Year 2016) 16 4. Ranking of States by Long-Run Solvency (Fiscal Year 2016) 18 5. Ranking of States by Service-Level Solvency (Fiscal Year 2016) 20 6. Ranking of States by Trust Fund Solvency (Fiscal Year 2016) 24 7. Ranking of States by Fiscal Condition

(Fiscal Year 2016, Unweighted) 25 8. Ranking of States by Fiscal Condition

(Fiscal Year 2016, Weighted) 27 9. 15-Year Treasury Bond Interest Rates 34 A1. Financial Statement Data Used to Construct Indicators 45 A2. Financial Indicators Used to Measure Fiscal Condition 48 A3. Ranking the States by Fiscal Condition Using

New Methodolog y (Fiscal Years 2006–2015) 50 B1. Components of Cash Solvency: Cash, Quick, and Current Ratios

for the States (Fiscal Year 2016) 60 B2. Components of Budget Solvency: Operating Ratio and Surplus or

Deficit Per Capita (Fiscal Year 2016) 61 B3. Components of Long-Run Solvency: Net Asset Ratio, Long-Term

Liability Ratio, and Long-Term Liabilities Per Capita (Fiscal Year 2016) 62

B4. Components of Service-Level Solvency: Taxes, Revenues, and Expenses to Total State Personal Income (Fiscal Year 2016) 63

B5. Components of Trust Fund Solvency: Unfunded Pensions and Other Postemployment Benefits as a Percentage of Personal Income (Fiscal Year 2016) 64

B6. State Debt (Fiscal Year 2016) 65 B7. Pension Liabilities under State Discount Rate Assumptions

(Fiscal Year 2016) 67 B8. Pension Liabilities Discounted under Risk-Free Discount Rate

(Fiscal Year 2016) 69 B9. Other Postemployment Benefits: Retiree Health Benefits

(Fiscal Year 2016) 71 B10. Pension Plans (Fiscal Year 2016) 72

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

FIGURES 1. Average Cash to Short-Term Liabilities (Quick Ratio) Trends 28 2. Average Surplus (or Deficit) Per Capita Trends 29 3. Average Revenue-to-Expenses (Operating Ratio) Trends 29 4. Average Long-Term Liability to Total Asset Trends 30 5. Average Net Asset Ratio Trends 31 6. Average Taxes, Revenues, and Expenses Relative to State

Personal Income 33 7. Average Trust Fund Solvency Trends 33 8. States Reliant on Oil Taxes Experience the Most Volatile

Budgets (Operating Ratio) 35 9. Budget Trends (Operating Ratio) for States with Significant Tax

Reforms 37 10. Service-Level Trends (Tax-to-Income Ratio) for States with

Significant Tax Reforms 39 11. Service-Level Trends (Expenses-to-Income Ratio) for States

with Significant Tax Reforms 39

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

For the fifth year in a row, we assess the fiscal health of the states. Each edition of these rankings has provided a snapshot of each state’s fis-cal health by presenting information from states’ audited financial reports in an easily accessible format. The goal for our research has been to establish a consistent set of financial data and basic indicators with which to evaluate individual state performance, better understand the factors that drive changes in performance, and identify areas where financial report- ing may improve. States face many fiscal problems, but these problems are not insurmountable. Studying how each state is performing with regard to a variety of fiscal indicators can help state policymakers address persistent issues and anticipate potential problems.

As with any set of measures, financial indicators and trend lines should be interpreted with caution and in the context of a deeper analysis of each state’s financials, pension systems, rainy day funds, budget and policy reforms, eco- nomic conditions, and fiscal institutions.

For our analysis, we draw primarily from each state’s comprehensive annual financial reports (CAFR) as well as from state actuarial reports.1 The goal of this study, as well as that of previous editions, has been to operationalize the CAFR by applying 13 basic indicators to measure state fiscal health. We calculate this year’s rankings from the states’ fiscal year (FY) 2016 reports and then apply our trend analysis to reports from 2006 through 2016. We hope that by applying our indicators to more years of data, we can reach a better understanding of what constitutes fiscal health.

1. All data except for personal income, population, and information on each state’s pension system and other postemployment benefits (OPEB) are drawn from each state’s CAFR. A CAFR is a full accounting of a state government’s finances, and it includes information on assets, long-term liabili- ties, debt, and cash flow. CAFRs provide the most comprehensive public accounting of state finances that allows for cross-state comparisons and the analysis of state performance over time. Appendix A table A1 lists where data were found for each variable.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

This paper contains three sections. Section 1 presents this year’s ranking. Sec- tion 2 presents an analysis of how states have changed over time, with an emphasis on what can be learned from consistently strong and consistently weak states. We also highlight lessons from oil-producing states, tax-reforming states, and states with the biggest changes in the health of their pension system. Section 3 concludes with key lessons from the rankings and trends and implications for states moving forward. More detailed information regarding the methodology and indicators by which we measure financial condition can be found in the appendices.

Several themes persist from the previous editions of “Ranking the States by Fiscal Condition.” States with long-running structural deficits and large unfunded pension obligations tend to be states that either skipped or reduced their contributions to employee pension and health benefit plans and then issued debt to cover budget shortfalls or pension contributions, effectively adding to future obligations. States that are reliant on natural resource revenues experi- enced dramatic swings in cash, budgetary, and service-level solvency indicators. FY 2015 Government Accounting Standards Board (GASB) reforms that required states to report unfunded pension obligations on their books generally resulted in larger long-term liabilities for states with weaker pension funding levels. With several years of data, we can also see the effect of tax reform in Indiana, Kansas, Michigan, North Carolina, Rhode Island, and Utah.

1. RANKING THE STATES Building on the previous editions of this study, we rank the states according to their fiscal solvency on the basis of their audited financial reports.2 Fiscal solvency cap- tures whether a state is able to meet its short-term and long-term obligations without incurring excessive debt, engaging in budget gimmicks, or using other evasive tactics.3

Each edition of this study has applied a method for assessing financial condition developed by public administration researchers XiaoHu Wang, Lynda Dennis, and Yuan Sen (Jeff ) Tu.4 Their study defined four types of solvency,

2. The most recent CAFRs available for all states at the time of writing were from FY 2016. 3. Eileen Norcross and Olivia Gonzalez, “Ranking the States by Fiscal Condition, 2017 Edition” (Mercatus Research, Mercatus Center at George Mason University, Arlington, VA, June 2017); Eileen Norcross and Olivia Gonzalez, “Ranking the States by Fiscal Condition, 2016 Edition” (Mercatus Research, Mercatus Center at George Mason University, Arlington, VA, June 2016); Eileen Norcross, “Ranking the States by Fiscal Condition” (Mercatus Research, Mercatus Center at George Mason University, Arlington, VA, July 2015). 4. XiaoHu Wang, Lynda Dennis, and Yuan Sen (Jeff) Tu, “Measuring Financial Condition: A Study of U.S. States,” Public Budgeting & Finance 27, no. 2 (2007): 1–21.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

including (1) cash solvency, (2) budget solvency, (3) long-run solvency, and (4) service-level solvency. In 2014, Sarah Arnett applied this method of measuring fiscal condition to produce a ranking of the states on the basis of their relative performance.5 The next edition of “Ranking the States by Fiscal Condition” updated Arnett’s study by changing how service-level solvency is calculated and by adding another dimension of solvency, (5) trust fund solvency, which included total unfunded pension obligations, other postemployment benefits (OPEB), and total state debt.6 Following the first edition, each new report adopted the same methodology, with minor improvements each year.7 The five solvency areas each attempt to measure different aspects of fiscal condition:

1. Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt.

2. Budget solvency measures whether a state can cover its fiscal-year spend- ing using current revenues. It can help address the question of whether the state ran a shortfall during the year.

3. Long-run solvency measures whether a state has a hedge against large long-term liabilities. Are enough assets available to cushion the state from potential shocks or long-term fiscal risks?

4. Service-level solvency captures whether states have enough “fiscal slack” by measuring taxes, revenues, and expenses relative to state personal income. If spending commitments demand more revenues, are states in a good position to increase taxes without harming their economy? Are expenses high relative to the income of state residents, pointing to unsustainable levels of spending?

5. Trust fund solvency measures how much retirement-related debt a state has. How large are unfunded pension liabilities and OPEB liabilities com- pared with the state personal income?8

5. Sarah Arnett, “State Fiscal Condition: Ranking the 50 States” (Mercatus Working Paper, Mercatus Center at George Mason University, Arlington, VA, January 2014). 6. Norcross, “Ranking the States by Fiscal Condition.” 7. In addition to adding a new solvency area, the 2015 edition of “Ranking the States by Fiscal Condition” changed the way service-level solvency is calculated by measuring taxes, revenues, and expenditures as a proportion of personal income. The 2017 edition of the study dropped total state debt from the trust fund solvency area and capped outlier cash values for the cash solvency area. This year’s edition equally weights each solvency area. For a better understanding of how these methodological changes affect the rankings, see appendix A for a backtracked ranking for each year this study has been released. 8. For a more in-depth explanation of each solvency area, see Norcross and Gonzalez, “Ranking the States by Fiscal Condition, 2017 Edition.”

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

The first four dimensions—cash, budget, long-run, and service-level sol- vency—are constructed on the basis of data from the state’s CAFR, particularly its statement of net assets, statement of activities, and change in net position. The fifth dimension of solvency is trust fund solvency, which consists of unfunded pension obligations and OPEB liabilities. Data measuring each state’s unfunded pension obligations come from individual actuarial reports for the state govern- ments’ state-administered pension plans. OPEB data come from CAFR state- ments and the actuarial statements of OPEB plans, where available. Population figures are drawn from the US Census, and personal income data are drawn from the Bureau of Economic Analysis’ regional economic accounts.9

Using the state’s financial statements, we construct 13 fiscal indicators to measure the different dimensions of fiscal health.10 Table 1 provides basic statis- tics, including the mean, median, standard deviation, and maximum and mini- mum values for each ratio in FY 2016. These statistics provide an overview of the average performance of the 50 states for each indicator. The biggest changes from the past year’s fiscal rankings report, which used FY 2015 data, are in three indicators: the change in net position or surplus (deficit) per capita, the long- term liability ratio, and the unfunded-pension-to-state-income ratio.

To rank the states by their short-term and long-term fiscal health pros- pects, the 13 indicators listed in table 1 are bundled according to the dimensions of solvency they measure. Appendix A explains how the individual indicators are standardized and summed to create an index of fiscal solvency. The state profiles in appendix C summarize key information for each state, providing a closer look at the underlying data that make up the final ranking.

Cash Solvency Rankings The first dimension of the ranking, cash solvency, is composed of three indica- tors, or ratios: the cash ratio, the quick ratio, and the current ratio, as displayed by equation 1. These three different ratios measure varying degrees of liquidity of state assets, with the cash ratio being the most liquid and the current ratio being the least liquid. These ratios capture a government’s cash position relative to its

9. United States Census Bureau, “State Population Totals, 2006–2016,” https://www.census. gov/data/datasets/2017/demo/popest/state-total.html; Bureau of Economic Analysis, “Regional Economic Accounts, 2006–2016,” https://www.bea.gov/data/economic-accounts/regional. 10. Appendix A table A1 describes where the line items for each fiscal indicator can be found in each state’s financial statement, and appendix A table A2 provides definitions of each indicator.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

short-term liabilities. They indicate whether a government can meet bills that are due over a 30- to 60-day horizon.

Cash solvency = cash ratio + quick ratio + current ratio (1)

As table 1 shows, the states’ mean cash ratio in FY 2016 is 2.22, meaning states have 2.22 times more cash than short-term liabilities, on average. The aver- age quick and current ratios for FY 2016 are 2.99 and 3.22, respectively. As a rough guideline, healthy current ratios should exceed two, and cash and quick ratios should be greater than one.11 Most states have enough cash to cover short- term liabilities, on the basis of these minimum benchmarks.

11. Steven Finkler, Financial Management for Public, Health, and Not-for-Profit Organizations (Upper Saddle River, NJ: Prentice Hall, 2012).

TABLE 1. DESCRIPTIVE STATISTICS FOR FISCAL YEAR 2016 STATE GOVERNMENT FINANCIAL INDICATORS

Financial indicators n Mean Median Standard deviation Maximum Minimum

Cash ratioa 50 2.22 1.50 2.53 17.07 0.42

Quick ratioa 50 2.99 2.45 2.53 17.38 0.92

Current ratioa 50 3.22 2.63 2.56 17.92 1.05

Operating ratio 50 1.01 1.03 0.09 1.16 0.52

Surplus (deficit) per capita 50 −$72.45 $135.94 $1,038.13 $529.95 −$6,945.82

Net asset ratio 50 −0.17 0.00 0.74 0.77 −2.98

Long-term liability ratio 50 0.63 0.39 0.79 3.88 0.04

Long-term liability per capita 50 $4,386.94 $3,010.80 $4,137.26 $18,928.22 $282.34

Tax income ratio 50 0.06 0.06 0.02 0.10 0.00

Revenue income ratio 50 0.13 0.13 0.03 0.23 0.09

Expenses income ratio 50 0.13 0.12 0.04 0.26 0.08

Pension income ratio 50 0.43 0.40 0.16 0.91 0.17

OPEB income ratiob 48 0.04 0.03 0.05 0.21 0.00

Source: Authors’ analysis of the FY 2016 CAFRs for all states.

Notes: CAFR = comprehensive annual financial report; FY = fiscal year; OPEB = other postemployment benefits. a. These are the descriptive statistics for the cash, quick, and current ratios before the outliers have been capped. The maximum values change to 7.72, 9.81, and 9.00 for the cash, quick, and current ratios, respectively, after capping Alaska as an outlier. b. OPEB-to-income ratios are reported for only 48 states because two states, Nebraska and South Dakota, do not report unfunded OPEB liabilities.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Table 2 ranks the states according to cash solvency. The rank is derived from a z-score, or a standardized value of the summed cash solvency indicators, which measures by how many standard deviations an individual state’s score is above or below the mean for all 50 states. For example, Ohio’s cash index is 2.26 standard deviations above the mean, giving the state a rank of ninth place for cash solvency. Ohio’s cash metrics, or indicators, show that it has a strong cash position, with between three to four times the cash needed to cover its short- term bills. By contrast, Washington has a cash index of −1.23, or about one stan- dard deviation below the mean. Washington’s cash, quick, and current ratios are 1.33, 2.05, and 2.48, respectively. These metrics indicate that although Washing- ton has sufficient cash relative to minimum benchmarks, it still performs below the mean performance of the states.

The top five states that performed well in this area, relative to other states, in FY 2016 are Alaska, Wyoming, South Dakota, Florida, and Montana. The bottom five states are Arizona, Pennsylvania, Massachusetts, Illinois, and Connecticut.

Alaska, Wyoming, and South Dakota’s high level of cash solvency is due to these states’ restricted permanent funds. Although each of their permanent funds is structured differently, they all restrict cash in some way. Alaska has cash, quick, and current ratios of 17.07, 17.38, and 17.93, respectively, in FY 2016 primar- ily because of $69.15 billion in cash, cash equivalents, investments, and receiv- ables recorded on its statement of net position. However, $44.79 billion of this is restricted for the state’s permanent funds, meaning it cannot readily be accessed for meeting short-term bills.12 Similarly, Wyoming’s high cash solvency indicators reflect the state’s reported $23.09 billion in cash on hand, $11.46 billion of which is restricted as nonspendable within the Permanent Mineral Trust Fund or the Common School Land Fund.13 Of South Dakota’s $3.14 billion in cash on hand, $663.56 million is restricted in funds held as permanent investments.14 However, as will be seen in the next few sections, Alaska does poorly in budget solvency and trust fund solvency (ranking 50th in both areas), and Wyoming does poorly in budget (47th), service-level (37th), and trust fund (37th) solvency. South Dakota performs relatively better in these other areas despite its restricted funds. The main takeaway here is that large cash ratios do not necessarily imply robust fiscal health. Although saving money for specific purposes may be fiscally responsible

12. Alaska FY 2016 CAFR, p. 19. 13. “Cash on hand” is used here to refer to cash, cash equivalents, investments, and receivables on a statement of net position. Wyoming FY 2016 CAFR, p. 36. 14. South Dakota FY 2016 CAFR, p. 34.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Rank State Cash index Rank State Cash index

1 Alaska 11.12 26 Iowa −1.02

2 Wyoming 8.82 27 Indiana −1.08

3 South Dakota 6.11 28 Virginia −1.10

4 Florida 5.08 29 Washington −1.23

5 Montana 3.68 30 New Jersey −1.29

6 Alabama 2.92 31 Louisianaa −1.29

7 Idaho 2.80 32 Colorado −1.53

8 North Dakota 2.70 33 Delaware −1.58

9 Ohio 2.26 34 Texas −1.66

10 Tennessee 2.03 35 Michigan −1.72

11 Arkansas 1.95 36 New Hampshire −1.73

12 Nebraska 1.70 37 Rhode Island −1.75

13 Oregon 0.88 38 West Virginia −1.98

14 Missouri 0.84 39 Wisconsin −2.12

15 Utah 0.61 40 Kentucky −2.25

16 Georgia 0.35 41 Maryland −2.29

17 Minnesota 0.24 42 Kansas −2.34

18 Hawaii −0.01 43 Maine −2.36

19 Mississippi −0.25 44 New York −2.51

20 Oklahoma −0.37 45 California −2.55

21 New Mexico −0.46 46 Arizona −2.68

22 South Carolina −0.49 47 Pennsylvania −2.84

23 North Carolina −0.59 48 Massachusetts −3.09

24 Nevada −0.68 49 Illinois −3.17

25 Vermont −0.80 50 Connecticut −3.26

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The cash solvency index is the sum of the standardized values of the cash, quick, and current ratios. a. New Jersey’s cash solvency score is –1.2853, and Louisiana’s is −1.2946. New Jersey is ranked 30th, and Louisiana is ranked 31st, although the rounded scores are the same.

TABLE 2. RANKING OF STATES BY CASH SOLVENCY (FISCAL YEAR 2016)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

to some extent, there exists a point at which this behavior exhibits diminishing marginal returns, especially if other financial needs are not being met.

Florida’s strong cash position is due to high levels of cash that are not restricted by permanent funds but are in some way set aside within the state’s rainy day fund. The state reports $20.35 billion of pooled investments with the State Treasury, $1.4 billion of which is part of the state’s Budget Stabilization Fund to be accessed in the case of a fiscal emergency.15 These are indicators of a strong short-term position.

Montana’s cash solvency rank is due to cash and equity in pooled invest- ments of $1.68 billion and $2.29 billion, respectively, giving it a strong short-term position. According to the FY 2016 CAFR, Montana has a total general fund bal- ance of $271.3 million.16 An analysis of states’ recession readiness indicates that to weather an average recession, Montana would need $465 million in budget reserves.17

It should be noted that one state, Alabama, has consistently performed well in the cash solvency area but has also consistently been late in filing its CAFR. Alabama released its FY 2016 CAFR on February 28, 2018. Public finance research suggests that late financial report filing can be associated with poor financial management or can act as an early sign of fiscal distress.18 Although Alabama has between 3.66 and 4.89 times the cash needed to cover short-term obligations in FY 2016, it should not be overlooked that the state’s reporting prac- tices could be greatly improved.

Budget Solvency Rankings Equation 2 displays the indicators that make up the second dimension of this ranking, budget solvency, which measures whether a state’s revenues match its expenses. The first indicator is the operating ratio, the proportion of total

15. Florida FY 2016 CAFR, p. 67. 16. In 2017, Montana instituted a budget stabilization fund; the state had previously relied on any remaining balances in its general fund to meet budget shortfalls. See Title 17, Chapter 7, Part 1 17-7-30 Budget Stabilization Reserve Fund—Rules for Deposits and Transfers (https://leg.mt.gov/bills/mca/ title_0170/chapter_0070/part_0010/section_0300/0170-0070-0010-0300.html). 17. Erick Elder, “Weathering the Next Recession: Is Montana Prepared?” (https://www.mercatus .org/publication/weathering-next-recession-how-prepared-montana), in “Weathering the Next Recession: How Prepared Are the 50 States?” (Mercatus Research, Mercatus Center at George Mason University, January 2016), p. 24. 18. Kloha, Philip, Carol S. Weissert, and Robert Kleine, “Someone to Watch over Me: State Monitoring of Local Fiscal Conditions,” American Review of Public Administration 35, no. 3 (2005): 236–55.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

revenues available to cover total expenses. A ratio greater than one indicates that revenues exceed expenses and thus that the state can pay for budgeted spending in the fiscal year. An operating ratio of less than one is a red flag indicating that the state is vulnerable to cash flow problems in the event of a fiscal setback. In FY 2016, the average operating ratio was 1.01, meaning that most states’ revenues were able to cover expenses.

Budget solvency = operating ratio + surplus (or deficit) per capita (2)

The second indicator is the surplus (or deficit) per capita, which is mea- sured as the state’s change in net assets divided by the state’s population. The change in net assets, also known as the change in position, captures the change in direction of the state’s overall financial position between the previous and cur- rent years. An increase in net assets is considered a surplus, whereas a decrease is considered a deficit. Most states reported a decline in position, or deficit, in FY 2016 of $72.45 per capita, on average. As section 2 will later explain, this is the first year since FY 2009 that most states have experienced a deficit. Note also that states with weak operating ratios tend to record a deficit.

Average surpluses per capita decreased by $222.43 in FY 2016 from the previous year. This means that, on average, states’ net position declined. As table 1 displays, the maximum surplus per capita also dropped significantly from the previous year, falling from $2,810.21 to $529.95. This steep drop in average net position is largely due to the effect of declining oil prices in North Dakota. North Dakota’s surplus per capita in FY 2015 was $2,810.21, but this figure fell to a defi- cit of $137.47 in FY 2016. This change leaves North Carolina’s surplus per capita of $529.95, a slight increase from $492.64 in FY 2015, as the new maximum sur- plus for FY 2016. For both years, Alaska remains the state with the largest deficit per capita.

Together, the operating ratio and the surplus or deficit per capita form the budget solvency index, which allows us to rank the states according to bud- get solvency, as seen in table 3. The top five states in this index in FY 2016 are Nevada, North Carolina, Georgia, Utah, and South Carolina. The bottom five states are Illinois, Wyoming,19 Connecticut, New Jersey, and Alaska.

19. Wyoming stands out as a state that dropped significantly in budget solvency from FY 2015 to FY 2016. This decline was primarily driven by a drop in revenues (−12%) that exceeded the state’s rise in expenses (5%). The largest areas that contributed to the drop in revenues included capital grants and contributions (−80%) and taxes (−20%). The largest tax decreases came from federal mineral royal- ties (−38%), mineral severance (−33%), miscellaneous (31%), and sales and use taxes (−19%).

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Rank State Budget index Rank State Budget index

1 Nevada 2.30 26 Missouri 0.38

2 North Carolina 1.94 27 Marylandd 0.38

3 Georgia 1.15 28 South Dakota 0.36

4 Utah 1.14 29 Virginia 0.30

5 South Carolinaa 1.14 30 West Virginia 0.22

6 Florida 1.08 31 Pennsylvania 0.16

7 Tennessee 1.04 32 Colorado 0.10

8 Mississippi 0.96 33 Ohio 0.09

9 Vermont 0.91 34 Oregon 0.04

10 New Hampshire 0.86 35 New Yorke 0.04

11 Hawaii 0.83 36 Indiana −0.06

12 Minnesota 0.81 37 Nebraska −0.09

13 Idaho 0.78 38 North Dakota −0.33

14 Montanab 0.78 39 Kentucky −0.37

15 Arizona 0.74 40 Louisiana −0.43

16 Maine 0.72 41 Oklahoma −0.63

17 California 0.69 42 Delaware −0.85

18 Wisconsin 0.68 43 New Mexico −0.94

19 Washington 0.66 44 Kansas −0.95

20 Arkansas 0.64 45 Massachusetts −1.12

21 Rhode Island 0.55 46 Illinois −1.36

22 Alabama 0.49 47 Wyoming −1.40

23 Iowa 0.47 48 Connecticut −1.56

24 Texasc 0.47 49 New Jersey −2.04

25 Michigan 0.46 50 Alaska −12.25

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The budget solvency index is the sum of the standardized values of a state’s operating ratio and its surplus (deficit) per capita ratio. a. Utah’s budget solvency score is 1.1374, and South Carolina’s is 1.1353. Utah is ranked fourth, and South Carolina is ranked fifth, although the rounded scores are the same. b. Idaho’s budget solvency score is 0.7839, and Montana’s is 0.7809. Idaho is ranked 13th, and Montana is ranked 14th, although the rounded scores are the same. c. Iowa’s budget solvency score is 0.4730, and Texas’s is 0.4683. Iowa is ranked 23rd, and Texas is ranked 24th, although the rounded scores are the same. d. Missouri’s budget solvency score is 0.3832, and Maryland’s is 0.3787. Missouri is ranked 26th, and Maryland is ranked 27th, although the rounded scores are the same. e. Oregon’s budget solvency score is 0.0377, and New York’s is 0.0355. Oregon is ranked 34th, and New York is ranked 35th, although the rounded scores are the same.

TABLE 3. RANKING OF STATES BY BUDGET SOLVENCY (FISCAL YEAR 2016)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Long-Run Solvency Rankings Long-run solvency is measured by three indicators: the net asset ratio, the long- term liability ratio, and the long-term liability per capita ratio, as displayed by equation 3. The first of these indicators, the net asset ratio, is the proportion of net assets, or assets that are left over after a state government has paid its debts, relative to the government’s total assets. The greater the amount of net assets relative to total assets, the more the government has on hand to cover long-term liabilities. The average net asset ratio in FY 2016 is −0.17. The second indicator, the long-term liability ratio, represents the proportion of long-term liabilities relative to total assets. It includes liabilities like outstanding bonds, loans, claims and judgments, pensions, OPEB, and compensated employee absences. On aver- age, states held long-term liabilities representing 63 percent of their total assets in FY 2016. The third long-run solvency indicator is long-term liabilities per capita. In FY 2016, states held an average of $4,386.94 per person in long-term liabilities.

Long-run solvency = net asset ratio + long-term liability ratio + long-term liability per capita ratio (3)

Average long-term liabilities relative to assets have worsened slightly since FY 2015. The maximum long-term liabilities relative to assets increased from 3.60 to 3.88 (New Jersey), whereas the minimum improved slightly from 0.05 to 0.04 (Nebraska). Long-term liabilities per capita increased by $115.04, on aver- age. The maximum long-term liability per capita increased from $16,820.87 to $18,928.22 (New Jersey), and the minimum decreased from $378.61 to $282.34 (Nebraska).

Table 4 ranks the states according to long-run solvency. States that per- formed well in this ranking by holding relatively lower levels of long-term liabilities include Nebraska, South Dakota, Oklahoma, Tennessee, and Idaho. States that performed poorly because of their higher levels of long-term lia- bilities included Kentucky, Connecticut, Massachusetts, Illinois, and New Jersey.

Service-Level Solvency Rankings Service-level solvency attempts to measure how much “fiscal slack” states have (to raise taxes or increase spending) through the calculation of three

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE 4. RANKING OF STATES BY LONG-RUN SOLVENCY (FISCAL YEAR 2016)

Rank State Long-run index Rank State Long-run index

1 Nebraska 10.98 26 Michigan −0.23

2 South Dakota 4.39 27 Georgia −0.30

3 Oklahoma 3.82 28 Kansas −0.35

4 Tennessee 3.45 29 Mississippi −0.40

5 Idaho 2.80 30 New Hampshire −0.43

6 Alaska 2.34 31 Indiana −0.44

7 Wyoming 2.13 32 Ohio −0.45

8 North Carolina 1.96 33 Colorado −0.54

9 North Dakota 1.84 34 West Virginia −0.76

10 Utah 1.37 35 Maine −0.78

11 South Carolina 0.99 36 Washington −0.93

12 Iowa 0.74 37 Pennsylvania −0.95

13 Montana 0.61 38 Louisiana −1.02

14 New Mexico 0.56 39 New York −1.08

15 Missouri 0.26 40 Delaware −1.12

16 Nevada 0.14 41 Vermont −1.19

17 Floridaa 0.14 42 Hawaii −1.32

18 Virginia 0.08 43 Rhode Island −1.62

19 Alabama 0.03 44 Maryland −1.69

20 Arizona 0.02 45 California −1.73

21 Texas −0.01 46 Kentucky −2.71

22 Minnesota −0.10 47 Connecticut −3.60

23 Arkansas −0.15 48 Massachusetts −3.86

24 Wisconsin −0.18 49 Illinois −5.14

25 Oregon −0.22 50 New Jersey −5.35

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The long-run solvency index is the sum of the standardized values of the net asset, long-term liability, and long-term liability per capita ratios. a. Nevada’s long-run solvency score is 0.1384, and Florida’s is 0.1379. Nevada is ranked 16th, and Florida is ranked 17th, although the rounded scores are the same.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

ratios: those of total taxes, revenues, and expenses relative to state personal income, as displayed by equation 4. All else equal, a lower value for each of these ratios is better for a state’s fiscal health. States with especially high lev- els of taxes, revenues, and expenses relative to state personal income are at greater financial risk should they experience a sudden downturn. They are more likely to have difficulty responding to increased demands on their bud- gets or increasing costs associated with pensions and OPEB obligations when hard times hit.

Service-level solvency = tax-to-income ratio + revenue-to-income ratio + expenses-to-income ratio (4)

Table 5 ranks the states according to service-level solvency. States with low levels of taxes, revenues, and expenses as a percentage of personal income are ranked at the top. The best-performing states when it comes to service-level sol- vency include Nevada, Alaska, New Hampshire, Virginia, and Florida, whereas the weakest-performing states are West Virginia, Vermont, Delaware, North Dakota, and New Mexico.

Although a lower value of each of the indicators that compose service-level solvency is generally better for a state’s fiscal health, greater context is required to interpret these metrics. These ratios provide a starting point to understand how each state compares when it comes to the financial burden it places upon its citizens. However, other factors, such as the structure of a state’s tax system and the nature of its spending, need to be considered when evaluating the fiscal effect of states’ tax and budget decisions.

Trust Fund Solvency Rankings Trust fund solvency captures the portion of a state’s long-term liabilities that includes risk-adjusted pension obligations and OPEB relative to state per- sonal income, as displayed by equation 5. OPEB are benefits other than pen- sions that are paid to former employees; they largely consist of retiree medical insurance, but they may also include ancillary benefits such as life insurance. The long-term liability solvency area captures some portion of these liabili- ties, but not the entirety of them. As described in appendix A table A1, the liability numbers used for long-run solvency are taken from the states’ state- ments of net assets and of activities. Until FY 2015, states only reported their deficit in annual contributions to the pension system as part of their long-term

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Rank State Service-level

index Rank State Service-level

index

1 Nevada 4.36 26 Arizona −0.26

2 Alaska 4.19 27 Connecticut −0.32

3 New Hampshire 3.77 28 California −0.49

4 Virginia 3.39 29 Idaho −0.56

5 Florida 3.33 30 Washington −0.60

6 South Dakota 2.80 31 Michigan −0.68

7 Nebraska 2.69 32 Wisconsin −0.69

8 Missouri 2.21 33 Maine −0.87

9 Utah 1.97 34 Montana −0.92

10 Kansas 1.89 35 Massachusetts −0.95

11 Oklahoma 1.53 36 Minnesota −1.14

12 Tennessee 1.50 37 Wyoming −1.38

13 Texas 1.32 38 New Yorkb −1.38

14 Illinois 1.31 39 Rhode Island −1.56

15 Colorado 1.08 40 Oregon −1.58

16 North Carolina 1.04 41 Iowac −1.58

17 Maryland 0.77 42 Hawaii −2.05

18 Indiana 0.71 43 Kentucky −2.10

19 Georgiaa 0.71 44 Mississippi −2.33

20 New Jersey 0.69 45 Arkansas −3.00

21 Alabama 0.49 46 West Virginia −3.08

22 South Carolina 0.31 47 Vermont −3.28

23 Pennsylvania 0.25 48 Delaware −3.36

24 Louisiana −0.13 49 North Dakota −3.37

25 Ohio −0.21 50 New Mexico −4.43

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The service-level solvency index is the sum of the standardized values of the tax-, revenue-, and expenses-to-income ratios. a. Indiana’s service-level solvency score is 0.7081, and Georgia’s is 0.7068. Indiana is ranked 18th, and Georgia is ranked 19th, although the rounded scores are the same. b. Wyoming’s service-level solvency score is −1.3815, and New York’s is −1.3829. Wyoming is ranked 37th, and New York is ranked 38th, although the rounded scores are the same. c. Oregon’s service-level solvency score is −1.5781, and Iowa’s is −1.5807. Oregon is ranked 40th, and Iowa is ranked 41st, although the rounded scores are the same.

TABLE 5. RANKING OF STATES BY SERVICE-LEVEL SOLVENCY (FISCAL YEAR 2016)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

liabilities.20 That number—the deficiency in pension funding since 1997—did not provide an accurate picture of the governments’ true financial positions.21 However, as of FY 2015, accounting guidance GASB 68 requires states to report their net pension obligation as part of their long-run liabilities.

Trust fund solvency = pension-to-income ratio + OPEB-to-income ratio (5)

An additional standard, GASB 67, provides new guidance for state gov- ernments in selecting the discount rate used to measure the present value of their unfunded liabilities. Previously, under GASB 25, governments selected the expected rate of return on plan assets with which to calculate the present value of their unfunded liabilities. This expected rate of return is based on pen- sion plan asset portfolios, which are typically invested in a mix of equities, fixed income, and alternatives. However, this aspect of GASB 25 was criticized in that it effectively measured a government-guaranteed, and therefore riskless, obliga- tion with reference to risky assets. On average, state plans used a discount rate of 7.52 percent to calculate the present value of plan liabilities in FY 2015, which is much higher than the return on bonds. The difference between these two dis- count rates is the risk premium that plans are assuming.

GASB 67 attempts to correct GASB 25 on this matter by suggesting a “blended approach” in applying a discount rate to value pension liabilities. For the portion of the liability that is backed by assets, the expected rate of return on pension assets (the higher discount rate) may be used to calculate the present value of the liability. For the portion of the liability that is not backed by assets, a low-risk return on tax-exempt municipal bonds (a lower discount rate) is to be used. The effect of this “blended rate” approach depends on when the plan is estimated to run out of assets. A plan projected to run out of assets sooner would apply the lower discount rate to a greater portion of its liability, resulting in a much larger present value for the unfunded liability. This effect points to an incentive for plans to project pension asset run-out dates far into the future, which will enable them to apply the higher discount rate, resulting in smaller present values for unfunded liabilities and lower annual contributions.

The effect of GASB 68 and GASB 67 is mixed. GASB 68’s inclusion of unfunded pension liabilities on the balance sheet is an improvement in

20. Norcross and Gonzalez, “Ranking the States by Fiscal Condition, 2017 Edition.” 21. Sheila Weinberg and Eileen Norcross, “GASB 67 and GASB 68: What the New Accounting Standards Mean for Public Pension Reporting” (Mercatus on Policy, Mercatus Center at George Mason University, Arlington, VA, June 2017).

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

transparency and accounting. But measurement problems still remain in the reporting of pension liabilities.22 The discretion states have in determining when they estimate a plan will run out of assets results in subjective and inconsistent application of discount rates. States vary in how stringently they apply the dis- count rate to measure underfunding. For example, New Jersey applied the lower return on municipal bonds to more of its pension plans’ liabilities in FY 2016, leading to the reporting of much higher liabilities than similarly situated states such as Illinois, which in FY 2015 projected that its major plans will not run out of assets until 2065.23 Because of the inconsistent application of GASB 67, many states still continue to understate the full value of their pension liabilities.24

In addition to ongoing variation in the measurement of plan underfunding, states only recognize the portion of their pension or OPEB liability for which the state government is responsible. The net pension liability that they report does not measure the entire unfunded liability for plans in which both local and state governments participate and contribute. Our interest is in determining the fiscal health of all pension plans that are administered by the state government, regard- less of the degree to which the state government is a participating or contributing employer to that plan.

For these reasons, we continue to include the trust fund solvency area to help account for persisting gaps in financial reporting of pension and OPEB lia- bilities.25 We use the most recent actuarial reports of pension and OPEB plans that states offer to their employees to complement information found on state CAFRs. We also include the full unfunded liability of state-administered plans to alert state governments to the fiscal condition of pension systems for which they have administrative responsibility. These shortfalls present a possible contingent liability to the state should the participating local government experience fis- cal stress. As with previous editions of the fiscal rankings, this survey does not include plans that are locally administered and locally funded.26

22. Sheila Weinberg and Eileen Norcross, “A Judge in Their Own Cause: GASB 67/68 and the Continued Mismeasurement of Public Pension Liabilities,” Journal of Law, Economics & Policy 14, no. 1 (2017): 61–90. 23. Weinberg and Norcross, “Judge in Their Own Cause,” 72. 24. Weinberg and Norcross, “Judge in Their Own Cause,” 72. 25. The trust fund solvency indicator was first introduced in Norcross, “Ranking the States by Fiscal Condition,” and has been applied in each edition since. In previous editions, trust fund solvency also included total debt outstanding. However, this component was eventually dropped because most debt is captured in long-run solvency, making the trust fund solvency indicator redundant. 26. A more in-depth description of the pension and OPEB data that were collected can be found in Norcross and Gonzalez, “Ranking the States by Fiscal Condition, 2017 Edition,” 8.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Average pension-to-income ratios increased by 8 percent since FY 2015, from 0.35 to 0.43. This means that in FY 2016, unfunded pension liabilities account for 43 percent of state personal income, on average. The increase from FY 2015 is largely due to the drop in the discount rate used to value pension liabil- ities on a fair-market basis.27 The maximum pension-to-income ratio increased from 0.73 to 0.91 (Alaska), and the minimum pension-to-income ratio increased from 0.16 to 0.17 (Tennessee).

Table 6 ranks the states according to trust fund solvency. The strongest states in this area include Oklahoma, Kansas, Tennessee, Nebraska, and Indiana. The five states with the weakest trust fund solvency are Illinois, Mississippi, Ohio, New Mexico, and Alaska. It should be stressed that pension and OPEB underfunding are measured relative to state personal income. How a state performs in the ranking is due to the size of its unfunded pension liability as compared with the relative wealth of its residents. For this reason, New Jersey—a high-income state—per- forms better than Alaska, although New Jersey’s pension underfunding is a press- ing budgetary problem, whereas Alaska’s defined-benefit pension plans are closed to new entrants and not increasing in size. This indicates that it is important not to rely on the pension-to-income metric alone but to pair it with a qualitative and complete assessment of the individual pension plans in question. The metric is useful, however, in that it highlights the magnitude of pension underfunding and the risk underfunding poses to state finances, which has not been consistently recognized or measured by states and local governments.

Overall Ranking of the States To construct an overall fiscal ranking of the states, the scores from the five dimen- sions of solvency are equally weighted and added together, as displayed by equa- tion 6.28 Table 7 ranks the states according to fiscal condition. The top five states are Nebraska, South Dakota, Tennessee, Florida, and Oklahoma. The bottom five states are Kentucky, Massachusetts, New Jersey, Connecticut, and Illinois.

Fiscal Condition Index = (cash solvency × 0.20) + (budget solvency × 0.20) + (long-run solvency × 0.20) + (service-level solvency × 0.20) + (trust fund solvency × 0.20) (6)

27. Section 2 expands on how interest rates have changed between 2006 and 2016. 28. This contrasts with the approach to the FY 2013, FY 2014, and FY 2015 rankings, which applied stronger weights to the short-term solvency areas. See appendix A for an explanation of this change.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Rank State Trust fund index Rank State Trust fund index

1 Oklahoma 5.62 26 Arkansas −0.29

2 Kansas 4.26 27 Rhode Island −0.36

3 Tennessee 3.25 28 Georgia −0.38

4 Nebraska 2.12 29 West Virginia −0.42

5 Indiana 1.75 30 Michigan −0.50

6 Wisconsin 1.18 31 Colorado −0.53

7 Florida 0.99 32 Minnesota −0.55

8 New Hampshirea 0.99 33 Missouri −0.56

9 Arizona 0.91 34 Alabama −0.72

10 Virginia 0.78 35 South Carolina −0.74

11 Delaware 0.58 36 Connecticut −0.84

12 North Dakota 0.57 37 Wyoming −0.87

13 South Dakota 0.56 38 New Jersey −0.90

14 North Carolina 0.44 39 Louisianab −0.90

15 Texas 0.26 40 Montana −1.03

16 Massachusetts 0.25 41 California −1.10

17 Maryland 0.15 42 Oregon −1.21

18 Vermont 0.11 43 Kentucky −1.33

19 Washington 0.07 44 Hawaiic −1.33

20 Utah 0.02 45 Nevada −1.44

21 Idaho 0.01 46 Illinois −1.49

22 Maine 0.00 47 Mississippi −1.57

23 New York −0.02 48 Ohio −1.66

24 Pennsylvania −0.19 49 New Mexico −1.76

25 Iowa −0.23 50 Alaska −1.93

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The trust fund solvency index is the sum of the standardized values of the pension- and OPEB-to-income ratios. a. Florida’s trust fund solvency score is 0.9887, and New Hampshire’s is 0.9873. Florida is ranked seventh, and New Hampshire is ranked eighth, although the rounded scores are the same. b. New Jersey’s trust fund solvency score is −0.8980, and Louisiana’s is −0.9016. New Jersey is ranked 38th, and Loui- siana is ranked 39th, although the rounded scores are the same. c. Kentucky’s trust fund solvency score is −1.3253, and Hawaii’s is −1.3325. Kentucky is ranked 43rd, and Hawaii is ranked 44th, although the rounded scores are the same.

TABLE 6. RANKING OF STATES BY TRUST FUND SOLVENCY (FISCAL YEAR 2016)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE 7. RANKING OF STATES BY FISCAL CONDITION (FISCAL YEAR 2016, UNWEIGHTED)

Rank State Fiscal

condition index Rank State Fiscal

condition index

1 Nebraska 3.48 26 Wisconsin −0.23

2 South Dakota 2.84 27 Arizona −0.25

3 Tennessee 2.25 28 Colorado −0.28

4 Florida 2.12 29 Iowa −0.32

5 Oklahoma 2.00 30 Washington −0.41

6 Wyoming 1.46 31 Oregon −0.42

7 Idaho 1.17 32 Michigan −0.53

8 Utah 1.02 33 Maryland −0.54

9 North Carolina 0.96 34 Maine −0.66

10 Nevada 0.93 35 Pennsylvania −0.71

11 Alaska 0.69 36 Mississippi −0.72

12 New Hampshire 0.69 37 Louisiana −0.76

13 Virginiaa 0.69 38 Hawaii −0.77

14 Alabama 0.64 39 Vermont −0.85

15 Missouri 0.63 40 Rhode Island −0.95

16 Montana 0.62 41 New York −0.99

17 Kansas 0.50 42 California −1.04

18 Georgia 0.31 43 West Virginia −1.20

19 North Dakota 0.28 44 Delaware −1.26

20 South Carolina 0.24 45 New Mexico −1.41

21 Indiana 0.17 46 Kentucky −1.75

22 Texas 0.07 47 Massachusetts −1.76

23 Ohio 0.01 48 New Jersey −1.78

24 Minnesota −0.15 49 Connecticut −1.91

25 Arkansas −0.17 50 Illinois −1.97

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The fiscal condition index is the sum of the cash, budget, long- run, service-level, and trust fund solvency indices equally weighted as follows: (0.20 × cash solvency score) + (0.20 × budget solvency score) + (0.20 × long-run solvency score) + (0.20 × service-level solvency score) + (0.20 × trust fund solvency score). a. Alaska’s fiscal condition solvency score is 0.6946, New Hampshire’s is 0.6911, and Virginia’s is 0.6905. Alaska is ranked 11th, New Hampshire is ranked 12th, and Virginia is ranked 13th, although the rounded scores are the same.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

In this year’s edition of the fiscal rankings, we present the ranking of the states on an unweighted basis. We also provide a ranking calculated with the same weights used in the most recent four editions of this study for comparison in table 8. The weighted ranking places more emphasis on the short term and therefore rewards states with permanent trusts and a high level of cash. Whereas it is important for states to maintain a reserve fund to cover periods of volatility, holding an excess of cash is not necessarily indicative of strong fiscal health, as Alaska and North Dakota demonstrate. Additionally, service-level solvency was assigned a weight of 10 percent in previous years because of the subjective nature of that metric. In the current study, service-level solvency is weighted equally (20 percent) with the other dimensions, arguably giving it more prominence than it previously had. There are good reasons to give more weight to certain dimensions over others, but we stress that ultimately, the relative ranking of a state does not mean as much as the metrics underlying that ranking.

In comparing the two methods, we can see that the bottom five states remain largely unchanged and that the rest of the rankings incur only minor changes. In the weighted ranking, Wyoming makes the top five because of its large levels of cash, but it gets pushed down in the unweighted ranking by Nebraska, Tennessee, and Oklahoma.

2. FISCAL CONDITION TRENDS For this section, we collected the same data used for section 1, including histori- cal data, to look at how states performed between fiscal years 2006 and 2016 as a whole. After looking at the 13 financial indicators over this time period, we discovered national trends within each area of solvency. Within these trends, national themes emerged. In particular, states that rely heavily on oil tax rev- enues experience more fiscal stress than other states and show more volatile operating ratios that reflect their difficulty in matching revenues with expenses when oil prices decline. Additionally, we compare significant tax reforms made by several states and their effects on the fiscal health of those states.

National Trends

Cash solvency. Cash trends across the states have, on average, improved over time. Each indicator of cash solvency—the cash, quick, and current ratios—shows an upward trend since FY 2006, with the biggest dip in cash available relative to

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE 8. RANKING OF STATES BY FISCAL CONDITION (FISCAL YEAR 2016, WEIGHTED)

Rank State Fiscal

condition index Rank State Fiscal

condition index

1 South Dakota 3.04 26 Indiana −0.20

2 Florida 2.60 27 Texas −0.26

3 Wyoming 2.59 28 Iowa −0.30

4 Nebraska 2.14 29 Washington −0.35

5 Tennessee 1.89 30 Vermont −0.40

6 Idaho 1.48 31 Wisconsin −0.47

7 Montana 1.43 32 Colorado −0.50

8 Alabama 1.17 33 Kansas −0.57

9 Utah 0.95 34 Michigan −0.58

10 Nevada 0.87 35 Arizona −0.61

11 North Carolina 0.82 36 Maine −0.74

12 Oklahoma 0.75 37 Maryland −0.75

13 North Dakota 0.73 38 Rhode Island −0.77

14 Missouri 0.62 39 Louisiana −0.81

15 Ohio 0.59 40 California −0.98

16 Arkansas 0.56 41 Pennsylvania −1.03

17 Georgia 0.53 42 West Virginia −1.04

18 South Carolina 0.28 43 New Mexico −1.05

19 Minnesota 0.19 44 New York −1.11

20 Virginia 0.15 45 Delaware −1.24

21 New Hampshire 0.13 46 Kentucky −1.53

22 Alaska 0.06 47 New Jersey −1.72

23 Oregon 0.02 48 Massachusetts −1.93

24 Hawaii −0.18 49 Illinois −2.12

25 Mississippia −0.18 50 Connecticut −2.16

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

Note: CAFR = comprehensive annual financial report. The weighted fiscal condition index is the sum of the cash, bud- get, long-run, service-level, and trust-fund solvency indices weighted as follows: (0.35 × cash solvency score) + (0.35 × budget solvency score) + (0.10 × long-run solvency score) + (0.10 × service-level solvency score) + (0.10 × trust fund solvency score). a. Hawaii’s fiscal condition solvency score is −0.1815, and Mississippi’s is −0.1820. Hawaii is ranked 24th, and Mississippi is ranked 25th, although the rounded scores are the same.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

short-term debt occurring shortly after the recession in FY 2009. We plot the quick ratio in figure 1 to illustrate this, but all three ratios follow this general trend.

Budget solvency. Budget solvency shows more variation across the states and over time. Overall, most states have experienced surpluses—that is, their net position has moved in a positive direction—until recently, as shown in figure 2. In FY 2016, most states experienced deficits, or declines in net position, for the first time since FY 2009, with an average deficit of −$72 per capita. The largest aver- age deficit experienced before that, −$556 per capita, occurred after the reces- sion. However, there has been quite a bit of variation in surplus (or deficit) per capita trends since FY 2006, so these averages should be interpreted cautiously and paired with a closer look at the underlying trends for each state.

Figure 3 displays the trends for the average level of revenues relative to expenses, or the operating ratio. State performance for this indicator has experi- enced less variation across states than the surplus per capita indicator, and it has been relatively robust since FY 2006, with FY 2009 as the weakest year for budget solvency. Following the recession in 2009, states had an operating ratio of 0.93 on average, meaning that most states’ revenues only covered around 93 percent of their expenses. There has been a slight decline in average state operating ratios since FY

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

20162015201420132012201120102009200820072006

FIGURE 1. AVERAGE CASH TO SHORT-TERM LIABILITIES (QUICK RATIO) TRENDS

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

–$600

–$400

–$200

$200

$400

$600

$800

20162015201420132012201120102009200820072006

0.85

0.90

0.95

1.00

1.05

1.10

20162015201420132012201120102009200820072006

FIGURE 2. AVERAGE SURPLUS (OR DEFICIT) PER CAPITA TRENDS

FIGURE 3. AVERAGE REVENUE-TO-EXPENSES (OPERATING RATIO) TRENDS

Source: Authors’ analysis of the FY 2006–2016 CAFRs and US Census for all 50 states.

Source: Authors’ analysis of the FY 2016 CAFRs for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

2014, but operating ratios have not fallen below the recommended level of 1.00; in other words, most states have managed to at least match revenues with expenses. Although state budgets have not fallen to the lows they hit during the recession, they also have not quite improved to prerecession levels. The largest average operating ratio in our sample, 1.09, occurred before the recession, in fiscal years 2006 and 2007.

Long-run solvency. All three indicators of long-run solvency—the long-term lia- bility-to-asset ratio, liabilities per capita, and the net asset ratio—have worsened over time. The most drastic decline took place in FY 2015, largely as a result of GASB 68, which required states to report their unfunded pension liabilities on their balance sheets.

Figure 4 displays how GASB 68 influenced the size of long-term liabilities relative to total assets. Between fiscal years 2014 and 2015, the average long-term liability-to-asset ratio increased by 53 percent, and the liabilities per capita ratio increased by 54 percent. Long-term liabilities, on average, have grown from 29 percent of assets held by states in FY 2006 to 62 percent in FY 2016. Average long-term liabilities per capita have also followed a similar trend, with a low of $2,122 per capita in FY 2006 and a high of $4,387 per capita in FY 2016.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

20162015201420132012201120102009200820072006

GASB 68

ra ti o

FIGURE 4. AVERAGE LONG-TERM LIABILITY TO TOTAL ASSET TRENDS

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

Note: GASB = Government Accounting Standards Board.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

–0.20

–0.15

–0.10

–0.05

–0.00

0.05

0.10

0.15

0.20

20162015201420132012201120102009200820072006

FIGURE 5. AVERAGE NET ASSET RATIO TRENDS

The worsening of states’ long-term financial positions is also reflected in average net asset ratios, which started at 0.13 in FY 2006 and dropped to −0.17 in fiscal years 2015 and 2016, as figure 5 shows. Figure 5 also demonstrates GASB 68’s influence on net asset ratios in 2015. The states’ average net asset ratio declined sixfold, falling from 0.03 in FY 2014 to −0.17 in FY 2015. Reporting larger unfunded pension liabilities increases the size of total long-term liabilities. Any available assets must now be stretched further than before to cover these liabilities.

There are a variety of reasons why a state may show a negative net asset ratio. All the metric indicates is that the state’s liabilities exceed its assets. A negative net position requires a deeper look at the individual state’s finances and its reasons for issuing debt. When states issue debt for ordinary purposes, such as capital construction (e.g., school or road construction), but do not record the underlying asset, this may lead the state to show a negative net asset ratio, although this does not mean the state is in fiscal stress. However, in some cases, states issue debt for extraordinary reasons—to cover budget gaps or to make contributions to the pension system. Issuing debt for spending that should be covered through annual appropriations is a red flag for fiscal imprudence and distress. This demonstrates that there is a spectrum of possible reasons for

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

issuing debt and that different reasons have varying implications for state fiscal health. As with all metrics, it is important to look deeper into the net asset ratio to determine the purpose and overall frequency of debt issuances.

For example, in fiscal years 2015 and 2016, New York reported a deficit in net position because of the issuance of debt for purposes related to tobacco settlements, local governments, infrastructure and transit projects, and obligations related to postemployment benefits for public workers.29 Kentucky’s FY 2016 negative net asset ratio stems from the adoption of GASB 68 and the reporting of pension liabili- ties on its balance sheet.30 The decline in net assets in Massachusetts is attributed to five primary reasons: the implementation of GASB 68, school construction costs, infrastructure and highway projects, debt issued to pay for capital assets held by quasi-public entities, payments to local governments and housing authorities, and OPEB payments.31 In FY 2009, Ohio reported a $6.11 billion deficit for unrestricted government assets because of the issuance of debt to public colleges and universities and to local governments and component units to build schools. Its unfunded lia- bilities of $341.5 million were due to compensated absences for public employees.32

Service-level solvency. Overall, the service-level solvency area provides only a very broad picture of the tax and spending burden placed on residents of the states. The trend lines, when paired with additional analysis, are complemented by budgetary solvency metrics, as detailed earlier in section 2. For tax revenues relative to state personal income, FY 2008 is the year that stands out, as shown in figure 6; the ratio increased to a high of 0.07 and marginally decreased the states’ average fiscal slack. Although tax revenues relative to income have remained somewhat constant, total taxes have experienced a steady increase, dropping only after the 2008 recession. Total revenues relative to income show a trend similar to that of taxes, without much variation over time, but with slightly more variation than taxes. Revenues similarly experienced a drop, but in FY 2009, the year after the recession, rather than in FY 2008. Average expenses relative to income increased between fiscal years 2009 and 2010, reflecting a reduction in fiscal slack for most states. This was likely a result of the recession.

Trust fund solvency. Figure 7 displays the trends in the trust fund solvency indi- cators. Overall, there has been an increase in the average pension-to-income

29. New York 2015 CAFR, p. 23; New York 2016 CAFR, p. 23. 30. Kentucky 2016 CAFR, p. 17. 31. Massachusetts 2016 CAFR, p. 19. 32. Ohio 2009 CAFR, pp. 7–8.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

expenses to income

total revenues to income

tax revenues to income

20162015201420132012201120102009200820072006

ra ti o

0.0

0.1

0.2

0.3

0.4

0.5

OPEB-to-income ratio

pension-to-income ratio

20162015201420132012201120102009200820072006

FIGURE 6. AVERAGE TAXES, REVENUES, AND EXPENSES RELATIVE TO STATE PERSONAL INCOME

FIGURE 7. AVERAGE TRUST FUND SOLVENCY TRENDS

Source: Authors’ analysis of the FY 2006–2016 CAFRs and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

Source: Authors’ analysis of the FY 2006–2016 CAFRs, pension and OPEB actuarial reports, and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

ratio over time. There was a significant uptick in FY 2012 and a drop back down in FY 2013, but there has been a gradual increase since.

The present value of unfunded pension liabilities is measured on a risk- adjusted basis determined by the prevailing rate of return on notional 15-year Treasury bonds on June 30 of each fiscal year.33 As displayed in table 9, interest rates hit a low mark in 2012 and another in 2016, resulting in an increase in the present value of the states’ unfunded pension liabilities in those years. There was not much variation in the average OPEB-to-income ratio, leaving it less informa- tive than the pension-to-income indicator.

For the period during which we applied these indicators to state finan- cial reports,34 several patterns emerged on the basis of economic and fiscal fac- tors that prevailed in particular states. We identified similar patterns of short- term volatility in the fiscal performance of states that rely on natural resources for their primary source of revenue. States with large and growing unfunded pension liabilities and persistent structural deficits showed worsening trends in long-term performance. And states that undertook significant tax reforms showed changes in their fiscal performance.

Fiscal Implications of Heavy Reliance on Oil Tax Revenues States that rely primarily on oil tax revenues to finance government spending are marked by big swings in short-term solvency.35 For example, Alaska, North Dakota, and Wyoming each rely more heavily for their revenue on severance taxes on oil production than do other states. Revenue from severance taxes made

33. This is the case for most years, except for 2007, in which the last available rates in the fiscal year were reported on June 29 by the US Treasury, and 2012, in which they were reported on June 28. Daily records of Treasury yield curve rates are available at https://www.treasury.gov/resource -center/data-chart-center/interest-rates/Pages/TextView.aspx?data=yield. 34. We relied primarily on state CAFRs supplemented by information from state actuarial pension and OPEB reports. 35. Norcross and Gonzalez, “Ranking the States by Fiscal Condition, 2017 Edition.”

TABLE 9. 15-YEAR TREASURY BOND INTEREST RATES

Year 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Interest rate

5.23 5.12 4.29 3.92 3.36 3.64 2.03 2.87 2.81 2.59 1.68

Source: US Department of the Treasury, “Daily Treasury Yield Curve Rates,” accessed October 1, 2018, https://www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=yield.

Note: The 15-year rate is the average between the 10-year and 20-year daily yield curve rates.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

up 41 percent and 19 percent of general revenues for North Dakota and Wyoming, respectively, in FY 2016.36 Alaska typically relies on severance taxes more than any other state. In FY 2016, however, it was unable to generate enough severance tax revenue to pay for the tax credits it had given to oil producers, with the result that it reported on its CAFR a negative amount of severance taxes brought in.37

All three states depend on severance taxes more than other states, and they experience more revenue volatility as a result. Figure 8 displays how Alaska, North Dakota, and Wyoming compare with the national average in their abil- ity to match revenues to expenses. These states have extremely volatile operat- ing ratios relative to the rest of the country. In most years, Alaska’s revenues exceeded its expenses by 50 percent or more. A decline in oil prices, however, resulted in revenues falling short of expenses by 50 percent in FY 2016. North Dakota and Wyoming follow similar paths. In FY 2015, North Dakota’s revenues exceeded expenses by 27 percent, but in FY 2016, they fell short and covered only 98 percent of spending. Wyoming’s FY 2014 revenues exceeded expenses by 48 percent, but then in FY 2016 revenues fell to cover only 93 percent of expenses.

36. North Dakota FY 2016 CAFR, p. 37; Wyoming FY 2016 CAFR, p. 39. 37. Alaska FY 2016 CAFR, p. 38, reports that −$318,546,000 was brought in via severance taxes that year.

0.0

0.5

1.0

1.5

2.0

2.5

benchmark

Wyoming North Dakota

national average

Alaska

20162015201420132012201120102009200820072006

FIGURE 8. STATES RELIANT ON OIL TAXES EXPERIENCE THE MOST VOLATILE BUDGETS (OPERATING RATIO)

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

As figure 8 displays, all states experienced weakened operating ratios as a result of the recession in 2008 and 2009, but the key difference between most states and Alaska, North Dakota, and Wyoming is that most states have recovered and stabilized. In contrast, Alaska, North Dakota, and Wyoming’s budgetary sol- vency indicators are still marked by volatility caused by swings in the price of natu- ral resources. A study by the Pew Research Center notes a similar trend and warns against the highly volatile nature of severance taxes. Their study ranks Alaska, North Dakota, and Wyoming as having the first-, second-, and third-most-volatile revenue streams of the 50 states, respectively, between 1997 and 2016.38

Many other states, including New Mexico, Oklahoma, Montana, Texas, and West Virginia, implement similar severance taxes on oil production, but they amount to 10 percent or less of general revenues. Thus, these states do not rely as heavily on these taxes as do Alaska, North Dakota, and Wyoming. As a result, they do not exhibit as much volatility in their finances, as shown by the operating ratio indicator. The Pew Research Center’s study ranks New Mexico, Oklahoma, Montana, Texas, and West Virginia as having the 8th, 14th, 25th, 26th, and 47th

most volatile revenue streams, respectively, of the 50 states.39

Fiscal Implications of Major Tax Reforms During the period in which we applied these indicators to state finances, between 2006 and 2016, several states undertook major tax reform. Two recent studies examine the effect of these reforms in several states, including Utah, Indiana, North Carolina, Kansas, Rhode Island, and Michigan.

In this section, we review these studies and their findings in light of the fis- cal ranking metrics for states. Figure 9 portrays the operating ratio budget trends for the highlighted tax reform states. We will discuss how their tax reforms affected their ability to match revenues with expenses over time as well as their tax revenue and expense trends relative to total state personal income.

The effects of a state’s tax reform on revenues depend on the design of the reform and on its implementation.40 Utah, Indiana, and North Carolina reduced

38. Pew Research Center, “Revenue Volatility Varies Widely by State and Tax Type,” January 29, 2015, http://www.pewtrusts.org/en/research-and-analysis/articles/2015/01/revenue-volatility -varies-widely-by-state-and-tax-type. 39. Pew Research Center, 2015. 40. According to one analysis, the effect of North Carolina’s reforms resulted in an overall improvement in the state’s business climate from 44th place in 2013 to 11th place in 2018. See Nicole Kaeding and Jeremy Horpedahl, “Help from Our Friends: What States Can Learn from Tax Reform Experiences across the Country,” Tax Foundation, May 15, 2018, https://taxfoundation.org/state-tax-reform-lessons-2018/.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

tax rates while broadening their bases. Base-broadening increases the overall efficiency of a tax system by reducing the distortionary effect of taxes on deci- sion-making.41 Utah moved from a progressive income tax to a flat income tax of 5 percent and replaced income deductions with a credit system, both of which also improved efficiency. Sales taxes were reduced, but certain exemptions for food purchases made using government assistance remained in place, a reform intended to achieve greater equity for low-income residents.42 The effect of these reforms on Utah’s tax revenues from 2006 to 2016 is mixed.43 Taxes as a percent- age of state income declined from 7 percent to 5 percent from 2006 to 2012 and then increased to 6 percent from 2013 to 2016, as figure 10 shows. Expenses as a percentage of income have declined from a high of 12 percent in 2010 to 9 percent in 2016, as shown in figure 11.

41. George R. Crowley, “Case Studies in the Political Economy of Tax Reform” (Mercatus Research, Mercatus Center at George Mason University, 2016). 42. Crowley notes that sales tax for food purchased with government assistance programs was already exempted before the reforms. This feature of sales tax reform is unlikely to achieve the goal of greater equity. See Crowley, “Case Studies in the Political Economy of Tax Reform,” 16. 43. Crowley, “Case Studies in the Political Economy of Tax Reform,” 17–18.

0.8

0.9

1.0

1.1

1.2

1.3

benchmark

Rhode Island Michigan

Utah

North Carolina Kansas Indiana

National average

20162015201420132012201120102009200820072006

FIGURE 9. BUDGET TRENDS (OPERATING RATIO) FOR STATES WITH SIGNIFICANT TAX REFORMS

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

As figure 9 displays, all states experienced dips in revenues as a result of the recession in 2008. Utah’s revenues, however, dropped more significantly than most because of a decline in income and sales taxes that resulted from the state’s 2006 and 2007 tax reforms.44 Utah’s budget solvency indicators show that over the 2006–2016 period, the state generally has operated with revenues exceed- ing expenses by a comfortable margin, except in the years following the state’s reforms and the recession. A decline in expenses helps to account for Utah’s positive operating ratios.

Indiana gradually reduced both its corporate and its income tax rates between 2012 and 2017. Nicole Kaeding and Jeremy Horpedahl of the Tax Foun- dation note that in addition to reducing rates, Indiana simplified its tax code; eliminated some tax incentives, thereby broadening its base; and increased its gas tax. Indiana’s operating ratio over the post-reform period indicates that the state’s revenues either matched its expenses or exceeded them by between 2 and 6 percent. In 2016, Indiana’s operating ratio hit a low point of one, with revenues exactly matching expenses. The state’s ratio of tax to personal income has alter- nated between 6 percent (in 2012) and 5 percent (in 2016) of state income (figure 10), while its expenses have remained steady at 11 percent of state income (figure 11), which explains Indiana’s lower operating ratio in 2016.

Between 2013 and 2017, North Carolina undertook a series of tax reform measures. In 2013, the state changed its personal income tax from a three-bracket structure to a flat rate of 5.75 percent. It also cut its flat corporate tax rate from 6.9 percent to 5 percent, limited or eliminated dozens of tax exemptions, expanded its sales tax base, and repealed its estate tax.45 As George Crowley explains, base broadening and lowering the tax rate meet the criteria of increasing efficiency and convenience, while at the same time, removing tax brackets may decrease equity for lower-income residents.46 The net effect of these changes has been positive, as the state’s financial indicators show. In FY 2010, North Carolina’s operating ratio reached its lowest mark, with the state’s revenues exceeding its expenses by 1 percent. Since the implementation of tax reform, North Carolina’s expenses have decreased, and it stands out in figure 9 because of this. Its rev- enues exceeded its expenses by 12 percent in 2016. North Carolina’s net position

44. Crowley, “Case Studies in the Political Economy of Tax Reform,” 17–18; Kaeding and Horpedahl, “Help from Our Friends,” 3. 45. Tax Foundation and North Carolina Chamber Foundation, North Carolina Illustrated: A Visual Guide to Tax Reform, accessed August 23, 2018, https://interactive.taxfoundation.org/nc-illustrated/ - north-carolina-illustrated. 46. Crowley, “Case Studies in the Political Economy of Tax Reform,” 31.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

0.04

0.05

0.06

0.07

0.08 Utah Rhode Island

North Carolina

Michigan

Kansas

Indiana national average

20162015201420132012201120102009200820072006

0.00

0.05

0.10

0.15

0.20

0.25

Utah

Rhode Island

North Carolina

Michigan

Kansas

Indiana national average

20162015201420132012201120102009200820072006

FIGURE 10. SERVICE-LEVEL TRENDS (TAX-TO-INCOME RATIO) FOR STATES WITH SIGNIFICANT TAX REFORMS

FIGURE 11. SERVICE-LEVEL TRENDS (EXPENSES-TO-INCOME RATIO) FOR STATES WITH SIGNIFICANT TAX REFORMS

Source: Authors’ analysis of the FY 2006–2016 CAFRs and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

Source: Authors’ analysis of the FY 2006-2016 CAFRs and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

has also improved, with its surpluses increasing from $296 per capita in FY 2013 to $530 per capita in FY 2016. The state’s taxes as a percentage of income have remained relatively steady at less than 6 percent annually, as shown in figure 10. Lastly, as shown in figure 11, North Carolina’s expenses have declined from 13 percent of state income in 2010 to 10 percent in 2016, which has contributed to the state’s strong operating ratios.

In contrast to Utah, Indiana, and North Carolina, Kansas lowered its tax rates and narrowed its tax base in 2012, running against the criteria of greater tax-system efficiency. This reform was not coupled with reductions to spend- ing to offset lost income, and it also included a pass-through exemption for the income of sole proprietorships that narrowed the state’s tax base while encour- aging tax avoidance. The effect of the overall reform was revenue negative. Taxes as a percentage of state income fell from 6 percent in FY 2012 to 5 percent in FY 2014, as shown in figure 10, while spending remained at 10 percent of state income, as shown in figure 11. The overall results of the reform are evident in the state’s budget indicators, as displayed in figure 9. In FY 2012, Kansas’s revenue exceeded the state’s expenses by 4 percent, but from then it declined. In fiscal years 2015 and 2016, the state’s revenues covered 98 percent and 94 percent of spending, respectively. Kansas’s net position moved in a negative direction, with deficits of $77 and $283 per capita in fiscal years 2015 and 2016, respectively.

In FY 2010, Rhode Island implemented a revenue-neutral reform of its per- sonal income tax. The state reduced its number of income tax brackets from five to three and lowered its top marginal tax rate from 9.9 percent to 5.9 percent while also eliminating its alternative 6 percent flat tax on personal income. The goal of these reforms was to achieve greater efficiency by streamlining rates and greater equity by eliminating itemized deductions and reducing the number of tax cred- its.47 The reform was projected to shift more of the tax burden to the top 5 percent of income earners. The state’s revenues increased slightly after the reform, remain- ing at roughly 6 percent of state income in 2010, as figure 10 shows, while the state’s expenses dropped sharply from 20 percent of state income to 14 percent of state income after the reform, as shown in figure 11. The result, borne out in figure 9, was that the state’s operating ratio increased steadily from 1.00 in FY 2010 to 1.06 in FY 2015. As Crowley notes, Rhode Island’s modest increase in tax revenue may also be due to overall increases in its GDP and per capita income.48 Rhode Island’s operating ratio has since fallen slightly to 1.03 in FY 2016.

47. Crowley, “Case Studies in the Political Economy of Tax Reform,” 20. 48. Crowley, “Case Studies in the Political Economy of Tax Reform,” 21.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Michigan undertook tax reform twice between 2006 and 2016. In FY 2007, the state replaced its Single Business Tax with the simpler-to-administer Michigan Business Tax, which consisted of a 4.95 percent tax on business income, a 0.8 per- cent tax on gross receipts, and an additional 22 percent surcharge on these activi- ties. The surcharge proved very unpopular, leading to a second set of reforms in 2011 that repealed the Michigan Business Tax and replaced it with a 6 percent flat tax on corporate income. The tax exempted all noncorporate businesses.49 These reforms followed some of the principles of optimal taxation, including increasing efficiency, transparency, and convenience. But, as Crowley finds, they may be criti- cized on equity grounds because they primarily benefit business owners.50 Evalu- ating the effect of these reforms on revenues is complicated by the fact that the Michigan economy experienced a decline in GDP and employment that began in 2000 and continued until 2009.51 Increases in GDP and per capita income preceded the 2011 tax reforms, and job growth began to improve.

Michigan’s operating ratio rose steadily from a low of 0.93 in FY 2009 to a high of 1.08 in FY 2012. The state’s operating ratio has since fallen and stabilized from FY 2014 to FY 2016 in the range between 1.01 and 1.03, as figure 9 displays. Improvement in the state’s operating ratio is likely due to expenses being cut. Michigan’s tax revenues declined from 7 percent to 6 percent between 2006 and 2016, as shown by figure 10. The state’s expenses increased from 13 percent to 16 percent of state income between 2006 and 2009 and then declined to 13 percent in 2016, as figure 11 shows.

States with Pension Problems States with large and growing pension liabilities include Connecticut, Illinois, Kentucky, and New Jersey. The trend lines for these states show increases in both long-run liabilities and unfunded pension liabilities. Between fiscal years 2006 and 2016, their long-term liabilities relative to assets tripled in size. Long- term liabilities were roughly equal to assets in both Illinois and New Jersey in FY 2006, but by FY 2016, they were three times as large as assets. Unfunded pension liabilities have more than doubled in New Jersey over this period, from 20 per- cent to 49 percent of state personal income. In Illinois and Kentucky, unfunded pension liabilities relative to state income tripled over this period from 22 per- cent and 21 percent to 67 percent and 61 percent, respectively. Connecticut’s

49. Crowley, “Case Studies in the Political Economy of Tax Reform,” 23. 50. Crowley, “Case Studies in the Political Economy of Tax Reform,” 24. 51. Crowley, “Case Studies in the Political Economy of Tax Reform,” 25–26.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

long-term liabilities grew over this period from 87 percent of assets to more than 200 percent of assets, and the state’s unfunded pensions relative to state income also more than doubled from 19 percent to 48 percent.

States with Consistently Strong Fiscal Performance Most states have healthy cash ratios of between one to three times the cash needed to cover their short-term bills. To determine which states are “consis- tently strong” in fiscal performance, we examine the solvency indicators on which they show greater variation, such as budget, long-run, and trust fund.

Montana exhibits strong short-term performance in budget solvency, with an operating ratio of at least 1.00 and surpluses, or increases in net position, over the entire period from FY 2006 to FY 2016. Idaho, Iowa, Mississippi, Missouri, North Carolina, North Dakota, Ohio, South Dakota, Tennessee, Texas, Utah, Vir- ginia, West Virginia, and Wisconsin have maintained operating ratios of at least one in all but one year of the period studied. Alabama, Colorado, Florida, Min- nesota, Nebraska, South Carolina, Vermont, and Wyoming each had operating ratios of at least one in all but two years studied.

Most states have kept their long-term liabilities below 50 percent of their assets. States that have surpassed this threshold and consistently kept their long- term liabilities at or below 20 percent of assets include Alaska, Idaho, Montana, Nebraska, North Carolina, North Dakota, Oklahoma, South Dakota, Tennes- see, and Wyoming. Alabama, Indiana, Iowa, Maine, Missouri, and New Mexico kept liabilities at 20 percent of assets during most of the period studied, with increases in recent years. These increases could be due to the implementation of GASB 68, which requires states to record unfunded pension liabilities on their balance sheets.

Pensions and OPEB are two areas in which most states show a consistent decline in solvency, due to growing unfunded liabilities. No state has consistently kept pension underfunding to 20 percent or less of state income across the whole sample. Indiana and Nebraska have come closest, with pension-to-income ratios falling below 0.20 over the past several years. However, their ratios have since risen to 0.23 and 0.22, respectively. Arizona, Arkansas, Colorado, Florida, Idaho, Indiana, Iowa, Kansas, Maine, Massachusetts, Minnesota, Mississippi, Missouri, Montana, Nevada, New Hampshire, North Dakota, Oklahoma, Oregon, Penn- sylvania, Rhode Island, Tennessee, Utah, Virginia, Washington, Wisconsin, and Wyoming all had OPEB-to-income ratios below 0.05 from FY 2006 to FY 2016. Nebraska and South Dakota report no unfunded OPEB liabilities.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

States with Consistently Weak Fiscal Performance A state exhibiting particularly weak short-run performance over the period from FY 2006 to FY 2016 is New Jersey, which recorded expenses exceeding revenues for all years in the sample, pointing to an ongoing structural deficit. Similarly, Illinois’s and Kentucky’s revenues fell short of those states’ expenses in all but one or two years sampled. Louisiana’s operating ratios began a steady decline in FY 2010, with revenues failing to match expenses from that year through FY 2016. Other states with weak operating ratios include California, Connecticut, Hawaii, Maryland, and Massachusetts.

In the long run, several states’ net positions moved in a negative direction in most years, including those of Hawaii, Illinois, Kentucky, Louisiana, Maryland, Massachusetts, New Jersey, and New York. Several states had long-term liabili- ties that exceeded their assets by 50 percent or more for most years, including Connecticut, Illinois, Massachusetts, and New Jersey. California, Rhode Island, and Washington all had liabilities that made up at least 50 percent of their assets for all years in the sample, and New York joined and has remained in this group of states since FY 2009. Furthermore, all of these states except for Washington had negative net asset ratios for each year in the sample. Maryland and Vermont had negative net asset ratios beginning in FY 2009, and Pennsylvania had a negative net asset ratio beginning in FY 2010.

Trust fund solvency has grown weaker for nearly all states over the period sampled. It should be stressed, however, that performance in trust fund solvency is relative to the total income of a state’s residents. For example, in FY 2016 New Jersey had unfunded pension liabilities of 49 percent of state income, and Alaska had unfunded pension liabilities of 91 percent of state income. However, New Jersey is a densely populated state with a high level of personal income, and its growing unfunded pension and OPEB liabilities continue to put it under immense fiscal pressure, as can be seen in the state’s low levels of cash, its struc- tural deficit, and its high ratio of long-term liabilities to assets. Alaska, in contrast, has a small population and low levels of personal income. Because it closed its defined-benefit pension plans to new hires in 2005, its unfunded pension liability should decrease over the coming years, presenting less of a risk to its finances.

3. CONCLUSION In addition to updating the US state fiscal rankings with FY 2016 data, we have applied our 13 basic indicators of fiscal solvency to state fiscal data for the pre- ceding 10-year period. This allowed us to examine trends of fiscal performance

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

across the states. In this data, several patterns are clear. The effects of tax reforms, oil price shocks, pension underfunding, the recession, and accounting reforms are all evident in the 10-year trend lines. States that are the most reliant on oil revenues show volatility in the short run and the most dramatic swings in levels of cash and revenue. In some cases, this leads those states to increase spending beyond a level that could be supported if the revenues were raised in proportion to the incomes of residents. Several states undertook tax reforms in the years we analyzed. Indiana, Kansas, Michigan, North Carolina, Rhode Island, and Utah each passed a variety of measures, including reductions to income, corporate, and sales taxes. In the cases of Indiana, North Carolina, and Utah, these reforms were accompanied by base-broadening or spending reforms that resulted in an overall neutral or positive effect on the states’ financial positions, as shown by improvement in their budgetary solvency indicators. Kansas’s tax reform of 2012, in contrast, cut tax rates but did not undertake spending reforms and narrowed the state’s tax base by including an exemption for sole proprietor- ships. As a result, Kansas’s budgetary solvency has steadily declined over the post-reform years. Over the period from FY 2006 to FY 2016, some states have consistently performed poorly, including Connecticut, Illinois, and New Jersey, all of which have experienced either ongoing structural deficits, a growing reli- ance on debt to fund spending, or underfunded pensions and OPEB liabilities. States with low levels of debt and unfunded pension liabilities and with strong short-term indicators include Nebraska, Tennessee, and Utah.

Financial indicators provide a snapshot of state fiscal performance and make the audited financial reports of state governments more accessible to the public, but they cannot provide a complete picture of fiscal performance. How- ever, the indicators can serve as warning flags and reveal important patterns, such as whether a state is running structural deficits or accumulating excessive levels of debt, or if its finances are at particular risk to economic shocks. The goal of this research is to shed light on how to best assess the short- and long-run fiscal risks states face and then to put this assessment in the context of states’ economic and fiscal institutions. With 10 years of data at our disposal, future research will assess the statistical reliability and validity of the metrics used in this study—that is, how well do these particular metrics assess fiscal condition in the states? The findings of that research may allow us to propose a core set of indicators that state governments might use to monitor signs of fiscal stress and that can be used in conjunction with qualitative analysis and case studies.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

APPENDIX A. RANKING METHODOLOGY This study calculates 13 financial indicators, as described in table A2, to cre- ate five dimensions of fiscal solvency: cash, budget, long-run, service-level, and trust fund. Table A1 describes the financial statements in which each indica- tor is found. Individual indicators are grouped and summed according to the dimension of solvency to which they contribute. For some indicators, a higher value indicates a higher degree of solvency. These include the cash ratio, quick ratio, current ratio, operating ratio, surplus (or deficit) per capita, and net asset ratio. For several other indicators, a lower value indicates higher solvency. To construct a ranking that is intuitive to interpret, the following indicators are transformed by taking their inverse: long-term liability ratio, long-term liability per capita, taxes-to-income ratio, revenue-to-income ratio, expenses-to-income ratio, pension affordability ratio, and OPEB affordability ratio.

Financial statement Line item Definition Notes

Statement of net assets (net position)

Cash Cash balances at the end of the fiscal year

Statement of net assets (net position)

Cash equivalents Short-term, highly liquid investments convertible to cash either readily or within three months of maturity

Statement of net assets (net position)

Investments Liquid resources that are invested to earn a return higher than a bank deposit

Most investments are reported at fair value.

Statement of net assets (net position)

Receivables Funds due from transac- tions with government (timing of these collections may vary, depending on their type)a

There are three types of transac- tions: (a) exchange transactions (e.g., individuals paying the state for college tuition or health services); (b) exchange-like transactions between the state and another party in which the value of the exchange is not equal to the benefits (e.g., the purchase of licenses or permits or regulatory fees); (c) nonexchange transactions, in which the government gives value to another party without receiving equal value in exchange.b

Statement of net assets (net position)

Current assets Assets that are converted into cash or consumed within the year

Statement of net assets (net position)

Current liabilities Obligations due within the year

Obligations include accounts pay- able, short-term debt, and voucher warrants.

TABLE A1. FINANCIAL STATEMENT DATA USED TO CONSTRUCT INDICATORS

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Financial statement Line item Definition Notes

Statement of net assets (net position)

Noncurrent liabilities

Long-term liabilities due over a few years or decades, often with interestc (listed in order of maturity)

Liabilities include outstanding bonds, net pension obligations,d com- pensated absences, and pollution remediation obligations.

Statement of net assets (net position)

Unrestricted net assets

Assets that may be used for any purpose

“Used for any purpose” does not imply the resource is liquid. A deficit in unrestricted net assets may signal the issuance of new debt and does not indicate fiscal trouble.

Statement of net assets (net position)

Restricted net assets (net position)

Assets that are restricted for a particular purpose (e.g., capital projects and debt service)

Assets are restricted by enabling legislation. They may be expend- able, or they may be nonexpendable, such as the principal used to fund an endowment.

Statement of net assets (net position)

Total net assets (total net position)

Combined net assets, including capital assets such as land, buildings, equipment, and infrastruc- ture (e.g., roads, bridges, and tunnels), less any outstanding debt used to acquire those assets

Capital assets are reported net of related debt. The resources needed to repay capital debt must be provided by other sources, since the capital assets themselves cannot be liqui- dated to fund these liabilities.

Statement of net assets (net position)

Total assets Sum of current, noncurrent, and capital assets

Statement of net assets (net position)

Total liabilities Sum of short- and long- term liabilities

Category includes general obligation and revenue bond debts, payments toward OPEB,e and the state’s portion of any unfunded pension.

Statement of activities Total taxes All revenues due from taxes levied

Category excludes grants, charges for services, contributions, transfers, and investment earnings.

Statement of activities Total revenue Total taxes plus program revenue

Category includes unrestricted grants, charges for services, contributions, transfers, and investment earnings.

Statement of activities Total expenses Total spent on govern- mental programs, debt service, unemployment compensation, loans, intergovernmental revenue sharing, lotteries, and the operation of government and commissions

On an accrual basis, expenses include costs that were incurred that year (such as earned pension benefits that will not be paid until a future date).

Statement of activities Changes in net assets

General revenues and changes in net assets totaled and added to net (expense) revenue totals to produce the change in net assets over the reporting period

Governments report the amount of net assets at the beginning of the year and add or subtract changes in net assets for the year to present ending net assets.f

TABLE A1. FINANCIAL STATEMENT DATA USED TO CONSTRUCT INDICATORS (CONTINUED)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Financial statement Line item Definition Notes

Annual reports for state pension plans

Unfunded pension liability

Pension plan assets subtracted from pension plan liabilities to calculate the size of the pension plan’s unfunded liability (or liability without any assets backing it)

These figures are reported in the annual reports of pension plans; in the fiscal rankings, the liability is recomputed on the basis of a low-risk or guaranteed-discount rate.

Notes to the basic financial statement

OPEB liability The OPEB obligation stated in the notes to the basic financial statement

These data were cross-checked with Standard & Poor’s OPEB data.

Source: Dean Michael Mead, An Analyst’s Guide to Government Financial Statements (Norwalk, CT: Governmental Accounting Standards Board, 2012).

a. Dean Michael Mead, An Analyst’s Guide to Government Financial Statements (Norwalk, CT: Governmental Account- ing Standards Board, 2012), 66. Examining receivables balances over time may help to show if the government’s ability to collect monies is increasing or decreasing. b. “Minnesota Management & Budget Statewide Operating Policy,” No. 0104-03, July 12, 2012, revised August 2, 2013. The GASB classifies nonexchange transactions into four types: (a) derived tax revenues, or the payment of income or sales taxes to the state; (b) nonexchange revenues, such as property taxes; (c) government-mandated nonexchange revenues, or federal grants to be used to carry out a mandate; and (d) voluntary nonexchange transactions, such as donations. c. States vary in reporting what is included in noncurrent liabilities. The notes to their financial statements provide more detail. See GASB, “Touring the Financial Statements, Part IV: Note Disclosures,” GASB website, December 2009, http://gasb.org/cs/ContentServer?c=GASBContent_C&pagename=GASB%2FGASBContent_C%2FUsersArticlePage& cid=1176156722430. d. GASB, “GASB Improves Pension Accounting and Financial Reporting Standards,” GASB website, news release, June 25, 2012, http://www.gasb.org/cs/ContentServer?pagename=GASB/GASBContent_C/ GASBNewsPage&cid=1176160126951. According to GASB, net pension obligation (NPO) is the difference between the annual required contribution (ARC) to fund the benefits earned in that year plus the cost of past earned benefits and the employer’s actual fiscal year contribution. See GASB, “Statement No. 27 of the Governmental Accounting Standards Board: Accounting for Pensions by State and Local Governmental Employers” (No. 116-C, Governmen- tal Accounting Standard Series, November 1994). The NPO only recognizes a portion of the annual expense of the pension plan, and it is not a measure of the outstanding pension liability. If the state has historically made the full ARC, its NPO is zero. This standard for recording the expense of the pension plan was replaced in FY 2014 with new guidance, GASB Statement No. 68. See GASB, “Summary of Statement 68 Accounting and Financial Reporting for Pensions—An Amendment of GASB Statement No. 27,” GASB website, June 2012, http://www.gasb.org/jsp/GASB/ Pronouncement_C/GASBSummaryPage&cid=1176160219492. e. OPEB = other postemployment benefits. f. GASB, “Touring the Financial Report, Part II: The Statement of Activities,” GASB website, May 2007, http:// gasb.org/cs/ContentServer?c=GASBContent_C&pagename=GASB%2FGASBContent_C%2FUsersArticlePage& cid=1176156736216.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

Financial Indicator Definition Interpretation Solvency dimension

1 Cash ratio (Cash + cash equivalents + invest- ments)/current liabilities

Higher ratio indicates greater cash solvency

Cash

2 Quick ratio (Cash + cash equivalents + invest- ments + receivables)/current

liabilities

Higher ratio indicates greater cash solvency

Cash

3 Current ratio Current assets/current liabilities Higher ratio indicates greater cash solvency

Cash

4 Operating ratio Total revenues/total expenses 1 or greater indicates budget solvency

Cash

5 Surplus (or deficit) per capita

Change in net assets/population Positive ratio indicates budget solvency

Budget

6 Net asset ratio Restricted and unrestricted net assets/total assets

Higher ratio indicates greater long-run solvency

Long-run

7 Long-term liability ratio

Long-term (noncurrent) liabili- ties/total assets

Lower value indicates greater long-run solvency

Long-run

8 Long-term liability per capita

Long-term (noncurrent) liabilities/population

Lower value indicates greater long-run solvency

Long-run

9 Tax-to-income ratio

Total taxes/state personal income Lower value indicates greater service-level solvency

Service-level

10 Revenue-to- income ratio

Total revenues/state personal income

Lower value indicates greater service-level solvency

Service-level

11 Expenses-to- income ratio

Total expenses/state personal income

Lower value indicates greater service-level solvency

Service-level

12 Pension-to- income ratio

Unfunded pension liability/state personal income

Lower value indicates greater trust fund solvency

Trust fund

13 OPEB-to-income ratio

OPEB/state personal income Lower value indicates greater trust fund solvency

Trust fund

Note: OPEB = other postemployment benefits.

TABLE A2. FINANCIAL INDICATORS USED TO MEASURE FISCAL CONDITION

Following the methodology of last year’s edition, for each indicator within the cash solvency dimension, we use an inner quartile method to establish an upper boundary at which to cap outliers. We do so by separating the data into quartiles and then setting the outer boundary at three times the inner quartile range beyond quartile three. Doing so sets a clear boundary with which we can identify any major outliers to cap.

To arrive at an overall ranking that aggregates each dimension of solvency, the ranking for each dimension is assigned a weight. Unlike that of previous years, this year’s edition gives each area of solvency an equal weight of 20 per- cent. This change was made to provide a more objective index of fiscal condi- tion. Although an argument can be made to weight the short-term more heav- ily, a counterargument could be made for doing the opposite. Cash and budget

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

solvency issues might be a more immediate concern for policymakers on any given day, but long-run and trust fund solvency are just as important and can eas- ily become more relevant when an economic downturn hits. Yet whichever area of solvency is prioritized, choosing to weight some areas more heavily than oth- ers is a subjective judgment. For this reason, we weight each dimension equally and provide our data online so that anyone can easily apply different weights and see how ranks change accordingly.

After applying the weights, the overall solvency score is calculated by sum- ming each of the dimensions of solvency into one final score that the states are then ranked by. The primary methodological change between this year and all previous editions of this study has been the change in weights.52

Although only minor changes have been made each year to improve the fis- cal condition index, these changes add up to a substantial overhaul that can affect the interpretation of changes in rank from year to year. Major methodological changes that have taken place include changes to the way service-level solvency was calculated in the 2015 edition (FY 2013), the capping of outliers in the 2017 edition (FY 2015), and the equal weighting of each subindex area in this edition (FY 2016). Each methodological change was carried into subsequent editions and has now been backtracked to reproduce the rankings with a standardized methodology across all years in which this study has been produced. Table A3 provides the backtracked rankings for fiscal years 2006–2016, for ease of com- parison. The data were adjusted for inflation using 2016 conversion factors, but this only affected the surplus per capita and long-term liability per capita figures. All other indicators in the rankings are relative measures of financial line items, and therefore their interpretation does not change upon adjusting for inflation.

As a result of standardizing the methodology, the replicated rankings from fiscal years 2013–2015 may not be perfectly comparable with the datasets from previous editions of this study. This is because of changes to the way that the index was calculated as well as to the way in which the data were collected and categorized. For example, in the 2017 edition, we classified both bills that are due within one year and those that are due in more than one year among North Dakota’s long-term liabilities; in this edition, we reclassified bills due within one year as current liabilities. These changes helped standardize the rankings over the selected sample but did not drastically change the rankings themselves.

52. For a more detailed description of how the dimensions of solvency are calculated, see Norcross and Gonzalez, “Ranking the States by Fiscal Condition, 2017 Edition.”

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 06

F is

ca l Y

ea r

20 07

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

A la

b am

a 2

5 15

15 37

3 5

8 32

15 35

A la

sk a

4 2

1 50

50 1

1 2

1 50

50 1

A ri

zo na

20 19

7 24

19 22

24 18

12 25

24 24

A rk

an sa

s 5

23 34

42 28

24 7

22 35

42 1

C al

if o

rn ia

43 46

42 25

15 41

41 46

46 29

14 43

C o

lo ra

d o

32 17

16 4

43 16

27 17

15 5

44 15

C o

nn ec

ti cu

t 38

48 39

18 42

43 37

48 34

18 37

D el

aw ar

e 37

32 25

45 9

39 35

34 25

45 7

F lo

ri d

a 19

38 41

6 4

12 13

38 18

6 2

G eo

rg ia

26 20

37 12

10 21

21 23

27 14

11 21

H aw

ai i

7 33

46 44

46 30

15 31

44 43

46 36

Id ah

o 17

8 8

30 18

14 25

7 10

34 8

Ill in

o is

40 50

48 8

40 46

42 50

49 11

45 48

In d

ia na

42 3

26 14

12 13

20 3

38 13

16 11

Io w

a 13

13 32

28 16

20 12

13 28

26 17

K an

sa s

34 25

33 19

21 29

36 29

26 17

23 28

K en

tu ck

y 28

39 36

38 45

38 30

39 50

38 43

Lo ui

si an

a 6

28 24

37 39

23 4

27 8

40 39

M ai

ne 50

15 30

39 32

47 50

15 24

37 31

M ar

yl an

d 36

37 23

9 17

25 47

37 43

9 19

M as

sa ch

us et

ts 46

47 29

26 25

45 44

47 31

28 22

M ic

hi g

an 45

36 49

33 27

44 46

36 48

32 26

M in

ne so

ta 31

24 12

35 30

33 28

21 23

35 32

M is

si ss

ip p

i 27

31 21

43 41

40 18

33 7

44 42

M is

so ur

i 15

16 40

10 31

17 10

20 37

7 29

TA B

LE A

3 . R

A N

K IN

G T

H E

S TA

T E

S B

Y F

IS C

A L

C O

N D

IT IO

N U

S IN

G N

E W

M E

T H

O D

O LO

G Y

( F

IS C

A L

Y E

A R

S 2

0 0

6 –2

0 15

)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 06

F is

ca l Y

ea r

20 07

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

M o

nt an

a 3

6 4

29 33

7 3

5 5

30 36

N eb

ra sk

a 18

1 17

13 6

2 14

1 11

10 3

N ev

ad a

23 29

11 1

34 8

29 28

33 1

38 12

N ew

H am

p sh

ir e

25 14

35 2

11 10

23 14

29 2

20 10

N ew

J er

se y

12 49

50 22

44 34

19 49

47 21

41 38

N ew

M ex

ic o

39 11

43 49

48 49

40 12

20 49

47 50

N ew

Y o

rk 41

45 27

34 20

42 39

45 42

33 15

N o

rt h

C ar

o lin

a 49

27 18

20 5

27 49

24 17

23 4

N o

rt h

D ak

o ta

9 21

5 40

14 19

11 19

4 39

12 20

O hi

o 16

41 38

23 49

32 16

42 21

24 49

O kl

ah o

m a

21 9

10 16

36 18

38 9

14 20

34 25

O re

g o

n 14

35 22

32 23

26 34

35 41

22 27

P en

ns yl

va ni

a 33

26 31

17 26

28 32

25 30

16 25

R ho

d e

Is la

nd 30

43 45

41 47

50 31

44 45

41 48

S o

ut h

C ar

o lin

a 35

34 13

36 38

36 33

32 16

36 40

S o

ut h

D ak

o ta

10 10

28 5

2 4

6 10

22 3

6 3

Te nn

es se

e 8

4 14

11 3

6 8

4 13

12 5

Te xa

s 48

12 6

7 1

9 43

11 9

8 21

U ta

h 1

18 3

21 22

5 2

16 3

19 18

V er

m o

nt 24

30 47

47 7

37 22

30 36

47 9

V ir

g in

ia 22

22 19

3 13

11 26

26 40

4 13

W as

hi ng

to n

29 40

20 27

8 31

17 41

19 27

10 23

W es

t V

ir g

in ia

11 42

9 48

35 35

9 40

6 46

30 29

W is

co ns

in 44

44 44

31 29

48 45

43 39

31 33

W yo

m in

g 47

7 2

46 24

15 48

6 2

48 28

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 08

F is

ca l Y

ea r

20 09

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

A la

b am

a 6

8 44

17 40

15 5

6 8

16 34

A la

sk a

1 2

1 50

50 1

1 2

50 45

50 27

A ri

zo na

36 21

34 24

3 21

48 26

37 24

19 31

A rk

an sa

s 5

20 28

42 29

22 8

19 17

43 28

C al

if o

rn ia

48 47

45 27

17 44

49 47

43 25

26 46

C o

lo ra

d o

27 18

12 6

33 18

32 17

16 6

37 16

C o

nn ec

ti cu

t 37

48 49

19 39

45 42

48 45

15 27

D el

aw ar

e 35

34 38

46 15

41 36

35 34

46 6

F lo

ri d

a 17

38 36

5 5

10 12

37 29

5 10

G eo

rg ia

23 29

29 20

21 24

24 32

24 13

15 19

H aw

ai i

18 35

46 41

47 40

25 36

49 40

42 49

Id ah

o 12

7 10

31 7

9 19

8 32

29 24

Ill in

o is

46 50

47 10

46 48

50 50

44 12

46 50

In d

ia na

13 3

13 18

1 4

13 4

21 20

7 8

Io w

a 15

13 14

26 8

17 15

11 9

31 22

K an

sa s

38 26

26 14

6 20

43 20

19 14

21 24

K en

tu ck

y 34

42 48

38 45

47 34

43 36

39 40

Lo ui

si an

a 7

24 7

40 42

25 7

24 4

38 39

M ai

ne 49

15 25

37 36

42 47

16 5

35 35

M ar

yl an

d 32

39 43

11 23

30 33

38 38

11 17

M as

sa ch

us et

ts 43

46 20

28 30

37 40

46 33

26 18

M ic

hi g

an 45

37 32

34 34

39 41

40 31

37 23

M in

ne so

ta 22

22 23

33 12

27 26

22 23

34 30

M is

si ss

ip p

i 20

31 11

43 32

35 20

28 10

42 47

M is

so ur

i 8

19 15

8 31

13 18

18 20

8 32

TA B

LE A

3 . R

A N

K IN

G T

H E

S TA

T E

S B

Y F

IS C

A L

C O

N D

IT IO

N U

S IN

G N

E W

M E

T H

O D

O LO

G Y

( F

IS C

A L

Y E

A R

S 2

0 0

6 –2

0 15

) (C O N TI N U E D

)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 08

F is

ca l Y

ea r

20 09

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

M o

nt an

a 4

6 5

32 35

7 4

10 2

33 36

N eb

ra sk

a 14

1 9

9 4

2 17

1 12

9 2

N ev

ad a

19 28

39 1

38 8

16 27

41 1

41 9

N ew

H am

p sh

ir e

31 16

37 2

19 11

35 21

26 2

13 10

N ew

J er

se y

28 49

50 23

44 46

22 49

46 22

43 45

N ew

M ex

ic o

44 11

42 49

48 49

21 12

48 50

48 48

N ew

Y o

rk 40

44 31

35 22

38 39

44 40

32 16

N o

rt h

C ar

o lin

a 47

25 24

21 11

29 44

23 25

18 4

N o

rt h

D ak

o ta

16 17

3 39

2 16

9 15

1 41

8 14

O hi

o 11

40 16

22 49

28 11

39 27

23 49

O kl

ah o

m a

25 9

8 15

9 12

14 7

7 21

1 3

O re

g o

n 29

32 33

30 28

33 28

33 42

28 45

P en

ns yl

va ni

a 33

23 22

13 26

23 38

29 30

19 20

R ho

d e

Is la

nd 39

45 41

44 41

50 29

45 15

44 44

S o

ut h

C ar

o lin

a 42

30 30

36 43

43 37

25 14

36 38

S o

ut h

D ak

o ta

3 10

19 3

10 3

2 9

13 4

5 2

Te nn

es se

e 10

4 18

12 24

6 10

3 11

10 3

Te xa

s 41

12 17

7 18

19 31

13 39

7 14

U ta

h 2

14 6

16 20

5 3

14 6

17 25

V er

m o

nt 21

33 35

45 16

36 30

34 22

48 12

V ir

g in

ia 24

27 21

4 13

14 27

30 28

3 11

W as

hi ng

to n

30 41

40 25

14 32

23 41

35 27

9 29

W es

t V

ir g

in ia

9 36

4 47

37 31

6 31

3 49

33 25

W is

co ns

in 26

43 27

29 25

34 45

42 18

30 29

W yo

m in

g 50

5 2

48 27

26 46

5 47

47 31

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 10

F is

ca l Y

ea r

20 11

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

A la

b am

a 7

7 13

19 34

9 9

7 35

17 37

A la

sk a

1 2

1 50

50 2

1 2

1 50

50 2

A ri

zo na

49 29

29 21

23 31

49 26

28 23

24 33

A rk

an sa

s 14

18 30

44 27

27 14

19 40

43 31

C al

if o

rn ia

46 46

45 25

24 41

44 46

43 26

26 46

C o

lo ra

d o

36 14

39 7

37 22

33 15

34 7

39 21

C o

nn ec

ti cu

t 48

47 49

16 31

46 48

47 44

19 32

D el

aw ar

e 34

35 23

45 10

37 41

36 32

47 7

F lo

ri d

a 6

36 6

5 7

5 10

35 26

5 10

G eo

rg ia

27 30

28 17

16 21

27 31

38 11

16 22

H aw

ai i

26 37

46 40

43 49

32 39

47 42

42 49

Id ah

o 19

9 10

28 20

17 21

8 5

31 17

Ill in

o is

50 50

50 11

45 50

50 50

49 18

46 50

In d

ia na

20 3

37 14

3 11

22 3

25 16

6 9

Io w

a 15

15 9

31 18

18 12

11 22

30 23

K an

sa s

43 20

21 13

19 24

39 21

23 14

21 23

K en

tu ck

y 28

43 44

39 41

45 17

44 45

40 44

Lo ui

si an

a 8

25 41

26 39

23 18

30 42

25 41

M ai

ne 44

16 11

34 35

34 47

16 21

37 22

M ar

yl an

d 38

40 43

10 15

26 40

42 48

10 15

M as

sa ch

us et

ts 42

48 47

27 14

48 42

48 39

29 19

M ic

hi g

an 41

39 32

36 29

40 36

38 30

34 35

M in

ne so

ta 32

28 35

35 25

33 31

24 29

36 28

M is

si ss

ip p

i 18

24 12

41 46

32 16

23 14

41 47

M is

so ur

i 12

19 26

9 32

15 13

18 27

8 29

TA B

LE A

3 . R

A N

K IN

G T

H E

S TA

T E

S B

Y F

IS C

A L

C O

N D

IT IO

N U

S IN

G N

E W

M E

T H

O D

O LO

G Y

( F

IS C

A L

Y E

A R

S 2

0 0

6 –2

0 15

) (C O N TI N U E D

)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 10

F is

ca l Y

ea r

20 11

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

M o

nt an

a 3

10 4

29 38

7 3

10 4

27 40

N eb

ra sk

a 10

1 20

6 2

1 11

1 17

6 3

N ev

ad a

16 31

40 1

44 13

20 33

37 2

43 16

N ew

H am

p sh

ir e

23 21

25 2

12 10

37 29

46 1

13 12

N ew

J er

se y

22 49

48 20

36 44

25 49

50 20

38 44

N ew

M ex

ic o

21 11

34 48

48 43

24 13

8 49

48 41

N ew

Y o

rk 45

45 38

37 17

42 45

45 41

38 18

N o

rt h

C ar

o lin

a 40

26 17

24 5

25 38

25 19

24 9

N o

rt h

D ak

o ta

9 12

3 42

11 14

7 12

3 45

11 13

O hi

o 5

38 27

22 49

20 8

37 20

22 49

O kl

ah o

m a

17 8

31 15

1 4

19 9

15 15

1 4

O re

g o

n 33

33 36

32 47

38 29

28 12

35 45

P en

ns yl

va ni

a 39

32 33

23 21

29 43

32 36

21 25

R ho

d e

Is la

nd 29

44 22

43 42

47 34

43 33

32 33

S o

ut h

C ar

o lin

a 31

22 16

38 40

35 28

20 11

39 36

S o

ut h

D ak

o ta

2 6

5 4

4 3

2 6

7 3

5 3

Te nn

es se

e 13

5 14

12 26

8 6

4 18

12 4

Te xa

s 35

13 18

8 13

16 35

14 16

9 14

U ta

h 4

17 8

18 28

6 4

17 6

13 27

V er

m o

nt 25

34 15

47 8

36 23

34 13

48 2

V ir

g in

ia 24

27 19

3 9

12 26

27 24

4 12

W as

hi ng

to n

30 41

42 30

6 30

30 40

9 28

8 25

W es

t V

ir g

in ia

11 23

7 46

30 28

15 22

10 44

30 31

W is

co ns

in 47

42 24

33 22

39 46

41 31

33 20

W yo

m in

g 37

4 2

49 33

19 5

5 2

46 34

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 12

F is

ca l Y

ea r

20 13

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

A la

b am

a 9

7 31

16 36

13 7

20 37

16 33

A la

sk a

1 2

1 50

50 2

1 2

1 50

50 2

A ri

zo na

42 22

12 24

23 28

43 22

10 21

22 28

A rk

an sa

s 15

21 39

44 26

31 14

21 39

42 28

C al

if o

rn ia

48 46

44 21

39 43

47 46

23 27

39 43

C o

lo ra

d o

29 15

19 8

37 20

26 16

20 7

38 19

C o

nn ec

ti cu

t 49

48 41

27 34

48 48

47 43

28 37

D el

aw ar

e 22

37 48

47 14

41 21

39 42

46 12

F lo

ri d

a 13

36 11

5 11

10 5

32 5

4 10

G eo

rg ia

33 32

35 15

21 24

34 31

32 17

19 24

H aw

ai i

32 39

47 42

42 47

32 40

29 44

43 45

Id ah

o 18

10 14

30 17

21 15

8 7

31 20

Ill in

o is

50 49

45 23

46 50

50 49

45 23

44 50

In d

ia na

20 4

32 19

7 16

23 3

30 13

6 12

Io w

a 14

14 15

32 2

9 13

15 17

34 23

K an

sa s

38 19

22 10

19 22

35 18

27 10

16 22

K en

tu ck

y 36

44 46

39 44

46 39

45 47

40 45

Lo ui

si an

a 19

31 42

25 41

27 22

34 49

22 42

M ai

ne 46

17 33

40 24

40 49

17 24

39 30

M ar

yl an

d 44

42 43

14 20

36 42

43 46

14 17

M as

sa ch

us et

ts 40

47 40

29 18

45 46

48 48

30 18

M ic

hi g

an 34

30 6

35 33

33 30

30 21

32 31

M in

ne so

ta 31

24 29

31 30

34 24

23 12

38 34

M is

si ss

ip p

i 16

27 21

41 47

30 17

27 33

41 47

M is

so ur

i 17

20 34

7 28

17 19

19 26

8 29

TA B

LE A

3 . R

A N

K IN

G T

H E

S TA

T E

S B

Y F

IS C

A L

C O

N D

IT IO

N U

S IN

G N

E W

M E

T H

O D

O LO

G Y

( F

IS C

A L

Y E

A R

S 2

0 0

6 –2

0 15

) (C O N TI N U E D

)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 12

F is

ca l Y

ea r

20 13

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

M o

nt an

a 4

9 8

28 40

11 6

9 6

29 35

N eb

ra sk

a 10

1 25

6 4

1 11

1 16

6 5

N ev

ad a

21 28

26 2

45 14

20 25

15 2

46 11

N ew

H am

p sh

ir e

25 33

23 1

10 8

36 33

31 1

13 9

N ew

J er

se y

35 50

50 17

38 49

37 50

50 19

40 49

N ew

M ex

ic o

12 12

37 49

48 37

16 12

35 48

49 39

N ew

Y o

rk 45

45 38

36 22

42 45

44 44

37 21

N o

rt h

C ar

o lin

a 41

23 27

22 12

26 44

24 18

25 8

N o

rt h

D ak

o ta

5 11

2 46

8 5

4 10

2 49

9 6

O hi

o 6

35 20

20 49

18 8

35 34

20 48

O kl

ah o

m a

11 8

7 13

1 4

12 7

22 12

1 4

O re

g o

n 23

26 28

37 43

32 18

29 9

36 41

P en

ns yl

va ni

a 47

34 36

18 25

35 41

37 40

18 24

R ho

d e

Is la

nd 37

43 18

34 32

38 38

42 19

35 27

S o

ut h

C ar

o lin

a 27

18 13

38 35

29 25

11 14

26 32

S o

ut h

D ak

o ta

2 6

5 3

9 3

2 5

13 3

7 3

Te nn

es se

e 7

5 16

11 5

6 9

6 25

11 4

Te xa

s 39

13 10

9 16

19 29

14 8

9 15

U ta

h 8

16 9

12 27

12 10

13 4

15 25

V er

m o

nt 24

40 30

48 3

39 28

36 41

47 2

V ir

g in

ia 30

29 17

4 15

15 27

28 28

5 14

W as

hi ng

to n

28 41

24 26

13 25

33 41

36 24

11 27

W es

t V

ir g

in ia

26 25

49 45

31 44

31 26

38 45

26 42

W is

co ns

in 43

38 4

33 6

23 40

38 11

33 3

W yo

m in

g 3

3 3

43 29

7 3

4 3

43 36

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 14

F is

ca l Y

ea r

20 15

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

A la

b am

a 10

20 33

15 36

18 9

19 40

19 36

A la

sk a

1 2

1 50

50 2

1 7

50 8

50 11

A ri

zo na

42 21

21 20

15 26

44 21

17 26

20 30

A rk

an sa

s 13

25 35

43 27

29 12

24 23

45 30

C al

if o

rn ia

47 46

23 28

41 43

46 45

11 29

41 42

C o

lo ra

d o

34 15

19 8

30 19

34 32

33 12

29 23

C o

nn ec

ti cu

t 48

47 50

27 37

48 49

47 4

22 35

D el

aw ar

e 28

41 46

46 12

44 25

39 43

47 8

F lo

ri d

a 7

28 7

3 10

7 5

16 10

4 6

G eo

rg ia

29 30

30 14

22 23

23 28

24 15

22 20

H aw

ai i

27 40

43 42

43 45

20 42

34 44

42 43

Id ah

o 14

9 6

33 16

15 10

5 6

30 12

Ill in

o is

49 49

41 24

44 49

50 49

46 20

44 50

In d

ia na

25 3

26 19

6 13

19 29

13 16

5 17

Io w

a 22

16 24

30 23

25 24

13 32

38 25

K an

sa s

35 19

38 10

20 21

37 20

45 11

18 22

K en

tu ck

y 43

45 44

40 45

50 45

46 37

41 45

Lo ui

si an

a 19

35 47

23 42

30 27

38 47

24 38

M ai

ne 44

23 40

35 25

38 41

35 12

35 24

M ar

yl an

d 41

43 45

17 17

34 43

44 39

18 14

M as

sa ch

us et

ts 50

48 48

31 19

47 48

48 48

32 19

M ic

hi g

an 33

31 37

29 31

33 36

25 35

34 31

M in

ne so

ta 17

22 15

38 33

28 18

23 15

39 32

M is

si ss

ip p

i 15

29 29

41 47

36 17

30 36

43 47

M is

so ur

i 16

17 31

7 32

17 16

15 18

7 33

TA B

LE A

3 . R

A N

K IN

G T

H E

S TA

T E

S B

Y F

IS C

A L

C O

N D

IT IO

N U

S IN

G N

E W

M E

T H

O D

O LO

G Y

( F

IS C

A L

Y E

A R

S 2

0 0

6 –2

0 15

) (C O N TI N U E D

)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

S ta

F is

ca l Y

ea r

20 14

F is

ca l Y

ea r

20 15

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

C as

h Lo

ng -r

un B

ud g

et S

er vi

ce -l

ev el

Tr us

t F

un d

FC I

M o

nt an

a 6

6 10

32 40

12 6

14 7

31 40

N eb

ra sk

a 9

1 14

6 3

1 11

1 29

6 3

N ev

ad a

20 24

20 2

46 11

22 22

16 2

46 16

N ew

H am

p sh

ir e

37 33

39 1

13 10

39 33

42 1

9 12

N ew

J er

se y

36 50

49 21

39 46

35 50

49 23

39 49

N ew

M ex

ic o

21 12

11 48

49 40

26 11

21 50

49 46

N ew

Y o

rk 46

44 34

34 21

41 42

40 25

36 16

N o

rt h

C ar

o lin

a 40

18 8

18 9

22 31

9 5

17 13

N o

rt h

D ak

o ta

3 10

3 49

7 6

4 8

1 49

7 6

O hi

o 12

34 18

22 48

20 14

31 41

25 48

O kl

ah o

m a

11 7

16 12

1 4

13 4

8 10

1 2

O re

g o

n 18

27 17

37 38

32 15

27 14

37 43

P en

ns yl

va ni

a 45

38 42

16 26

35 47

36 38

21 26

R ho

d e

Is la

nd 38

42 22

36 28

39 38

43 9

40 28

S o

ut h

C ar

o lin

a 23

11 13

25 35

24 21

10 19

27 34

S o

ut h

D ak

o ta

2 5

9 4

11 3

2 3

26 3

10 3

Te nn

es se

e 8

8 36

11 2

8 7

2 20

13 2

Te xa

s 32

14 4

9 8

14 29

18 30

9 15

U ta

h 5

13 5

13 24

9 8

12 3

14 23

V er

m o

nt 26

36 27

47 5

37 30

41 22

48 17

V ir

g in

ia 24

32 28

5 14

16 28

17 31

5 11

W as

hi ng

to n

31 39

25 26

18 27

33 37

44 28

21 31

W es

t V

ir g

in ia

30 26

32 45

29 42

32 34

28 46

27 45

W is

co ns

in 39

37 12

39 4

31 40

26 27

33 4

W yo

m in

g 4

4 2

44 34

5 3

6 2

42 37

N o

te : F

C I =

F is

ca l C

o n

d it

io n

In d

ex . M

aj o

r m

et h

o d

o lo

g ic

al c

h an

g e

s in

cl u

d e

ch an

g e

s to

t h

e w

ay s

e rv

ic e

-l ev

e l s

o lv

e n

cy is

c al

cu la

te d

in t

h e

2 0

15 e

d it

io n

( F

Y 2

0 13

), t

h e

ca p

p in

g o

f o

u tl

ie rs

in t

h e

2 0

17 e

d it

io n

( F

Y 2

0 15

), a

n d

t h

e e

q u

al w

e ig

h ti

n g

o f

e ac

h s

u b

in d

ex a

re a

in t

h is

e d

it io

n (

F Y

2 0

16 ).

E ac

h n

ew m

et h

o d

o lo

g ic

al c

h an

g e

w as

c ar

ri e

d in

to t

h e

n ex

t ye

ar ’s

e d

it io

n a

n d

h as

n o

w b

e e

b ac

kt ra

ck e

d t

o r

e p

ro d

u ce

t h

e ra

n ki

n g

s w

it h

a s

ta n

d ar

d iz

e d

m et

h o

d o

lo g

y ac

ro ss

a ll

ye ar

s th

at t

h is

s tu

d y

h as

b e

e n

p ro

d u

ce d

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

APPENDIX B. DATA TABLES TABLE B1. COMPONENTS OF CASH SOLVENCY: CASH, QUICK, AND CURRENT RATIOS FOR THE STATES (FISCAL YEAR 2016)

State Cash ratio Quick ratio Current ratio State Cash ratio Quick ratio Current ratio

Alabama 3.66 4.23 4.89 Montana 3.98 4.82 5.26

Alaska 17.07 17.38 17.92 Nebraska 2.95 3.86 3.95

Arizona 0.88 1.14 1.40 Nevada 1.46 2.65 2.69

Arkansas 3.17 3.86 4.14 New Hampshire 0.75 1.46 2.82

California 0.82 1.19 1.62 New Jersey 0.93 2.44 2.44

Colorado 1.32 1.93 2.11 New Mexico 2.01 2.53 2.60

Connecticut 0.42 1.00 1.05 New York 0.71 1.51 1.52

Delaware 1.34 1.95 1.98 North Carolina 1.67 2.55 2.72

Florida 4.80 5.80 5.81 North Dakota 3.23 4.59 4.63

Georgia 2.13 3.13 3.24 Ohio 3.43 4.05 4.20

Hawaii 2.22 2.77 2.91 Oklahoma 2.06 2.55 2.67

Idaho 3.57 4.36 4.66 Oregon 2.70 3.25 3.42

Illinois 0.55 0.92 1.13 Pennsylvania 0.69 1.08 1.39

Indiana 1.37 2.06 2.68 Rhode Island 1.13 1.84 2.02

Iowa 1.39 2.36 2.47 South Carolina 1.90 2.48 2.70

Kansas 0.80 1.60 1.62 South Dakota 4.76 6.63 6.78

Kentucky 0.87 1.52 1.75 Tennessee 3.03 4.12 4.17

Louisiana 1.27 2.01 2.48 Texas 1.28 1.76 2.09

Maine 0.65 1.30 2.02 Utah 1.61 3.65 3.75

Maryland 0.75 1.60 1.75 Vermont 1.62 2.46 2.50

Massachusetts 0.48 1.11 1.16 Virginia 1.55 2.23 2.31

Michigan 1.04 1.73 2.27 Washington 1.33 2.05 2.48

Minnesota 2.32 2.99 3.01 West Virginia 1.27 1.54 1.78

Mississippi 2.14 2.56 2.78 Wisconsin 0.89 1.74 1.76

Missouri 1.97 3.68 3.72 Wyoming 7.20 7.59 7.81

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

Note: Table B1 lists the underlying cash indicators for each state. As a result, it reflects Alaska’s values before they are capped. After transforming the data to cap outliers, Alaska’s cash, quick, and current ratios become 7.72, 9.81, and 9.00, respectively.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B2. COMPONENTS OF BUDGET SOLVENCY: OPERATING RATIO AND SURPLUS OR DEFICIT PER CAPITA (FISCAL YEAR 2016)

State Operating ratio Surplus or deficit

per capita ($) State Operating ratio Surplus or deficit

per capita ($)

Alabama 1.03 141.62 Montana 1.05 261.71

Alaska 0.52 −6945.82 Nebraska 0.99 −0.01

Arizona 1.05 226.94 Nevada 1.16 520.57

Arkansas 1.04 248.10 New Hampshire 1.04 412.78

California 1.04 270.78 New Jersey 0.89 −797.87

Colorado 1.01 40.33 New Mexico 0.96 −489.97

Connecticut 0.92 −692.68 New York 1.00 16.36

Delaware 0.96 −376.84 North Carolina 1.12 529.95

Florida 1.07 276.63 North Dakota 0.98 −137.47

Georgia 1.07 331.36 Ohio 1.00 63.08

Hawaii 1.05 332.18 Oklahoma 0.96 −170.68

Idaho 1.05 240.07 Oregon 1.01 −33.38

Illinois 0.92 −450.10 Pennsylvania 1.01 62.25

Indiana 1.00 −14.32 Rhode Island 1.03 225.23

Iowa 1.03 182.19 South Carolina 1.07 372.62

Kansas 0.94 −282.97 South Dakota 1.02 106.11

Kentucky 0.98 −124.77 Tennessee 1.07 289.93

Louisiana 0.96 11.04 Texas 1.03 155.48

Maine 1.04 252.21 Utah 1.08 291.33

Maryland 1.02 130.26 Vermont 1.05 412.12

Massachusetts 0.95 −490.71 Virginia 1.02 92.23

Michigan 1.03 160.46 Washington 1.04 229.47

Minnesota 1.05 312.74 West Virginia 1.01 89.07

Mississippi 1.06 323.29 Wisconsin 1.04 243.57

Missouri 1.03 107.78 Wyoming 0.93 −576.67

Source: Authors’ analysis of the FY 2006–2016 CAFRs for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B3. COMPONENTS OF LONG-RUN SOLVENCY: NET ASSET RATIO, LONG-TERM LIABILITY RATIO, AND LONG-TERM LIABILITIES PER CAPITA (FISCAL YEAR 2016)

State Net asset

ratio

Long-term liability

ratio

Long-term liability per capita ($) State

Net asset ratio

Long-term liability

ratio

Long-term liability per capita ($)

Alabama 0.01 0.31 2,118 Montana 0.22 0.20 2,247

Alaska 0.77 0.08 8,670 Nebraska 0.28 0.04 282

Arizona 0.07 0.34 2,194 Nevada 0.03 0.37 1,697

Arkansas 0.11 0.35 2,986 New Hampshire −0.02 0.50 2,555

California −0.57 0.92 5,642 New Jersey −2.98 3.88 18,928

Colorado −0.02 0.48 3,175 New Mexico 0.50 0.23 3,977

Connecticut −1.71 2.30 17,418 New York −0.24 0.58 4,605

Delaware −0.15 0.61 7,537 North Carolina 0.08 0.14 938

Florida 0.12 0.31 2,199 North Dakota 0.53 0.10 3,509

Georgia −0.01 0.47 2,302 Ohio 0.07 0.51 3,243

Hawaii −0.16 0.84 12,056 Oklahoma 0.31 0.11 609

Idaho 0.37 0.11 963 Oregon 0.17 0.41 3,283

Illinois −2.86 3.30 12,816 Pennsylvania −0.27 0.61 3,109

Indiana −0.13 0.50 2,155 Rhode Island −0.49 0.90 5,717

Iowa 0.16 0.22 1,656 South Carolina 0.17 0.23 1,311

Kansas −0.05 0.41 2,527 South Dakota 0.34 0.08 650

Kentucky −1.15 1.38 9,960 Tennessee 0.14 0.10 641

Louisiana −0.20 0.65 4,133 Texas 0.26 0.33 3,474

Maine −0.21 0.56 2,812 Utah 0.26 0.15 1,555

Maryland −0.48 0.99 7,186 Vermont −0.25 0.68 5,154

Massachusetts −1.93 2.75 11,518 Virginia −0.06 0.33 1,714

Michigan −0.10 0.45 1,883 Washington 0.02 0.64 8,169

Minnesota 0.07 0.36 2,458 West Virginia −0.12 0.43 4,194

Mississippi −0.04 0.37 3,036 Wisconsin 0.00 0.33 2,589

Missouri −0.01 0.26 1,809 Wyoming 0.74 0.10 3,989

Source: Authors’ analysis of the FY 2006–2016 CAFRs and the US Census for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B4. COMPONENTS OF SERVICE-LEVEL SOLVENCY: TAXES, REVENUES, AND EXPENSES TO TOTAL STATE PERSONAL INCOME (FISCAL YEAR 2016)

State

Taxes/ personal income

Revenues/ personal income

Expenses/ personal income State

Taxes/ personal income

Revenues/ personal income

Expenses/ personal income

Alabama 0.05 0.12 0.11 Montana 0.05 0.14 0.13

Alaska 0.00 0.14 0.26 Nebraska 0.05 0.09 0.09

Arizona 0.05 0.13 0.12 Nevada 0.04 0.09 0.08

Arkansas 0.07 0.18 0.18 New Hampshire 0.03 0.09 0.09

California 0.06 0.13 0.12 New Jersey 0.05 0.11 0.12

Colorado 0.04 0.11 0.11 New Mexico 0.07 0.23 0.24

Connecticut 0.06 0.12 0.13 New York 0.06 0.14 0.14

Delaware 0.09 0.18 0.19 North Carolina 0.06 0.11 0.10

Florida 0.04 0.09 0.09 North Dakota 0.08 0.19 0.19

Georgia 0.05 0.12 0.11 Ohio 0.05 0.12 0.12

Hawaii 0.09 0.16 0.15 Oklahoma 0.05 0.10 0.11

Idaho 0.06 0.13 0.13 Oregon 0.06 0.15 0.15

Illinois 0.05 0.10 0.11 Pennsylvania 0.05 0.12 0.12

Indiana 0.05 0.11 0.11 Rhode Island 0.06 0.15 0.14

Iowa 0.06 0.15 0.14 South Carolina 0.05 0.12 0.11

Kansas 0.05 0.10 0.10 South Dakota 0.04 0.09 0.09

Kentucky 0.07 0.15 0.16 Tennessee 0.05 0.11 0.10

Louisiana 0.04 0.12 0.13 Texas 0.04 0.11 0.11

Maine 0.06 0.14 0.13 Utah 0.06 0.10 0.09

Maryland 0.06 0.11 0.11 Vermont 0.10 0.19 0.18

Massachusetts 0.06 0.13 0.14 Virginia 0.05 0.09 0.09

Michigan 0.06 0.13 0.13 Washington 0.05 0.13 0.13

Minnesota 0.08 0.14 0.13 West Virginia 0.07 0.18 0.18

Mississippi 0.06 0.17 0.16 Wisconsin 0.06 0.13 0.13

Missouri 0.04 0.10 0.10 Wyoming 0.07 0.14 0.15

Source: Authors’ analysis of the FY 2006–2016 CAFRs and the Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B5. COMPONENTS OF TRUST FUND SOLVENCY: UNFUNDED PENSIONS AND OTHER POSTEMPLOYMENT BENEFITS AS A PERCENTAGE OF PERSONAL INCOME (FISCAL YEAR 2016)

State Pensions/

personal income OPEB/

personal income State Pensions/

personal income OPEB/

personal income

Alabama 0.46 0.05 Montana 0.53 0.01

Alaska 0.91 0.21 Nebraska 0.22 0.00

Arizona 0.40 0.00 Nevada 0.65 0.01

Arkansas 0.41 0.00 New Hampshire 0.27 0.03

California 0.54 0.05 New Jersey 0.49 0.15

Colorado 0.43 0.01 New Mexico 0.80 0.05

Connecticut 0.48 0.09 New York 0.35 0.07

Delaware 0.30 0.17 North Carolina 0.31 0.08

Florida 0.27 0.02 North Dakota 0.30 0.00

Georgia 0.40 0.04 Ohio 0.75 0.03

Hawaii 0.61 0.13 Oklahoma 0.35 0.00

Idaho 0.36 0.00 Oregon 0.65 0.00

Illinois 0.67 0.08 Pennsylvania 0.37 0.03

Indiana 0.23 0.00 Rhode Island 0.40 0.01

Iowa 0.38 0.00 South Carolina 0.46 0.05

Kansas 0.33 0.00 South Dakota 0.32 0.00

Kentucky 0.61 0.03 Tennessee 0.17 0.01

Louisiana 0.49 0.04 Texas 0.33 0.07

Maine 0.35 0.03 Utah 0.36 0.00

Maryland 0.34 0.03 Vermont 0.34 0.06

Massachusetts 0.33 0.04 Virginia 0.28 0.01

Michigan 0.42 0.04 Washington 0.34 0.04

Minnesota 0.44 0.00 West Virginia 0.41 0.04

Mississippi 0.71 0.01 Wisconsin 0.26 0.00

Missouri 0.43 0.01 Wyoming 0.49 0.01

Source: Authors’ analysis of the FY 2006–2016 CAFRs, pension and OPEB actuarial reports, and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

Note: OPEB = other postemployment benefits.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B6. STATE DEBT (FISCAL YEAR 2016)

State

Total general obligation bonds

($ thousands)

Total primary government debt

($ thousands) Personal income

($ thousands) Ratio of debt to personal income

Total primary debt per capita

($)

Alabama 722,383 5,191,977 190,791,463 0.03 1,068

Alaska 921,144 2,093,079 41,032,003 0.05 2,821

Arizona 0 9,502,194 278,924,877 0.03 1,371

Arkansas 1,518,148 3,959,545 117,572,045 0.03 1,325

California 79,043,295 112,554,735 2,197,492,012 0.05 2,868

Colorado 0 6,301,318 288,432,728 0.02 1,137

Connecticut 17,394,622 23,545,920 254,047,871 0.09 6,584

Delaware 2,118,548 3,271,448 46,362,308 0.07 3,436

Florida 10,712,000 25,174,000 944,443,033 0.03 1,221

Georgia 9,493,441 14,096,779 431,331,043 0.03 1,367

Hawaii 6,294,325 8,667,415 72,214,987 0.12 6,067

Idaho 0 1,228,927 65,823,005 0.02 730

Illinois 26,795,531 31,256,694 666,935,503 0.05 2,442

Indiana 0 1,000,258 288,486,508 0.00 151

Iowa 0 3,648,776 146,685,133 0.02 1,164

Kansas 0 7,745,489 141,112,300 0.05 2,664

Kentucky 0 7,692,612 175,258,173 0.04 1,734

Louisiana 4,610,809 12,264,745 203,591,796 0.06 2,620

Maine 464,444 1,168,260 59,005,346 0.02 877

Maryland 9,465,285 18,319,396 348,569,720 0.05 3,045

Massachusetts 21,668,296 29,569,062 443,700,515 0.07 4,341

Michigan 1,625,000 7,314,900 440,291,844 0.02 737

Minnesota 7,043,943 9,155,250 287,681,695 0.03 1,659

Mississippi 4,389,749 5,697,307 107,402,992 0.05 1,906

Missouri 208,880 3,546,970 266,406,080 0.01 582

Montana 115,500 220,753 44,188,348 0.00 212

Nebraska 0 34,780 94,661,640 0.00 18

Nevada 1,358,430 3,186,600 128,294,465 0.02 1,084

New Hampshire 889,802 1,486,235 77,848,085 0.02 1,113

New Jersey 1,991,645 42,727,114 554,267,581 0.08 4,777

New Mexico 326,755 3,497,735 80,758,305 0.04 1,681

New York 2,887,000 56,692,000 1,195,263,336 0.05 2,871

North Carolina 3,038,665 7,808,030 426,188,736 0.02 770

North Dakota 0 1,894,145 41,715,862 0.05 2,499

Ohio 9,283,156 17,689,412 521,208,626 0.03 1,523

Oklahoma 82,100 2,141,450 179,237,997 0.01 546

Oregon 5,525,430 11,083,552 184,407,086 0.06 2,708

Pennsylvania 12,517,909 16,588,566 655,506,262 0.03 1,298

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

State

Total general obligation bonds

($ thousands)

Total primary government debt

($ thousands) Personal income

($ thousands) Ratio of debt to personal income

Total primary debt per capita

($)

Rhode Island 1,051,810 2,556,297 54,486,321 0.05 2,420

South Carolina 962,196 2,856,956 195,791,444 0.01 576

South Dakota 0 522,268 41,584,285 0.01 603

Tennessee 2,124,897 2,389,853 288,531,063 0.01 359

Texas 15,060,000 50,806,000 1,327,260,948 0.04 1,823

Utah 2,585,000 5,155,000 124,319,657 0.04 1,689

Vermont 667,832 708,855 31,429,989 0.02 1,135

Virginia 601,632 6,634,016 451,911,594 0.01 789

Washington 20,518,000 25,892,000 389,858,930 0.07 3,553

West Virginia 393,089 2,030,403 68,457,129 0.03 1,109

Wisconsin 6,054,989 13,855,193 273,188,936 0.05 2,398

Wyoming 0 24,259 32,326,423 0.00 41

Source: Authors’ analysis of the FY 2006–2016 CAFRs, US Census, and Bureau of Economic Analysis Regional Eco- nomic Accounts for all 50 states.

TABLE B6. STATE DEBT (FISCAL YEAR 2016) (CONTINUED)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B7. PENSION LIABILITIES UNDER STATE DISCOUNT RATE ASSUMPTIONS (FISCAL YEAR 2016)

State Assets

($ thousands) Liabilities

($ thousands) Unfunded liability

($ thousands) Funded ratio

(percent)

Unfunded liability/personal income (percent)

Alabama 33,502,184 49,295,145 15,792,961 68 8

Alaska 14,676,677 21,045,476 6,368,799 70 16

Arizona 44,308,076 64,121,791 19,813,715 69 7

Arkansas 24,819,082 30,580,140 5,782,848 81 5

California 542,068,513 764,260,788 222,192,275 71 10

Colorado 43,404,710 74,565,792 31,161,082 58 11

Connecticut 31,370,324 66,736,317 35,365,993 47 14

Delaware 9,339,315 10,438,848 1,099,534 89 2

Florida 145,500,000 170,400,000 24,900,000 85 3

Georgia 82,676,301 110,944,246 28,267,945 75 7

Hawaii 14,998,749 27,439,234 12,440,485 55 17

Idaho 14,323,049 16,523,159 2,200,110 87 3

Illinois 118,251,464 249,337,048 131,085,584 47 20

Indiana 31,330,640 47,409,874 16,079,234 66 6

Iowa 31,960,270 38,267,347 6,307,077 84 4

Kansas 18,256,598 27,318,252 9,061,654 67 6

Kentucky 29,181,865 61,838,997 32,657,132 47 19

Louisiana 39,656,394 60,022,980 20,366,586 66 10

Maine 13,077,353 16,031,096 2,953,743 82 5

Maryland 51,955,510 72,790,542 20,835,031 71 6

Massachusetts 48,059,750 83,529,085 35,469,335 58 8

Michigan 62,215,349 100,102,842 37,887,493 62 9

Minnesota 55,956,165 73,481,336 17,525,171 76 6

Mississippi 25,685,579 42,843,536 17,157,957 60 16

Missouri 57,004,531 70,286,370 13,281,840 81 5

Montana 10,500,261 14,123,457 3,623,196 74 8

Nebraska 12,322,196 13,492,408 1,170,212 91 1

Nevada 35,896,200 48,459,200 12,563,000 74 10

New Hampshire 7,682,989 12,816,264 5,133,274 60 7

New Jersey 86,052,246 152,267,734 66,215,488 57 12

New Mexico 26,789,655 38,284,351 11,494,696 70 14

New York 297,520,200 312,171,100 14,650,900 95 1

North Carolina 92,964,638 101,529,965 8,565,327 92 2

North Dakota 4,637,285 7,135,253 2,497,968 65 6

Ohio 177,227,260 239,829,006 62,601,746 74 12

Oklahoma 29,202,211 38,772,320 9,570,109 75 5

Oregon 58,390,900 74,910,200 16,519,300 78 9

Pennsylvania 87,139,774 149,784,305 62,644,532 58 10

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

State Assets

($ thousands) Liabilities

($ thousands) Unfunded liability

($ thousands) Funded ratio

(percent)

Unfunded liability/personal income (percent)

Rhode Island 7,882,086 12,821,954 4,939,869 61 9

South Carolina 31,870,335 52,850,048 20,979,713 60 11

South Dakota 10,851,252 10,851,252 0 100 0

Tennessee 35,368,705 37,050,760 1,682,055 95 1

Texas 219,554,573 272,046,415 52,491,842 81 4

Utah 26,713,884 31,111,369 4,397,485 86 4

Vermont 4,005,175 5,976,436 1,971,260 67 6

Virginia 67,660,203 90,793,027 23,132,824 75 5

Washington 74,352,600 88,271,000 13,918,400 84 4

West Virginia 13,598,681 17,927,504 4,328,823 76 6

Wisconsin 95,396,200 95,414,000 17,800 100 0

Wyoming 7,863,264 9,937,983 2,074,719 79 6

Source: Authors’ analysis of the FY 2006–2016 CAFRs, pension actuarial reports, and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

TABLE B7. PENSION LIABILITIES UNDER STATE DISCOUNT RATE ASSUMPTIONS (FISCAL YEAR 2016)

(CONTINUED)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B8. PENSION LIABILITIES DISCOUNTED UNDER RISK-FREE DISCOUNT RATE (FISCAL YEAR 2016)

State Market value of liability

($ thousands)

Market value of unfunded liability

($ thousands) Funded ratio (percent)

Unfunded liability/ personal income

(percent)

Alabama 120,473,034 86,970,850 28 46

Alaska 52,011,300 37,334,622 28 91

Arizona 154,739,525 110,431,449 29 40

Arkansas 72,915,366 48,170,513 34 41

California 1,732,907,146 1,190,838,633 31 54

Colorado 166,217,211 122,812,501 26 43

Connecticut 153,021,586 121,651,262 21 48

Delaware 23,086,689 13,747,374 40 30

Florida 398,511,579 253,011,579 37 27

Georgia 255,444,582 172,768,281 32 40

Hawaii 59,008,684 44,009,935 25 61

Idaho 38,107,225 23,784,176 38 36

Illinois 564,045,760 445,794,296 21 67

Indiana 98,440,619 67,109,979 32 23

Iowa 88,351,824 56,391,554 36 38

Kansas 65,237,811 46,981,212 28 33

Kentucky 135,771,655 106,589,790 21 61

Louisiana 140,071,194 100,414,799 28 49

Maine 33,876,023 20,798,670 39 35

Maryland 169,051,277 117,095,767 31 34

Massachusetts 192,642,441 144,582,691 25 33

Michigan 246,290,795 184,075,446 25 42

Minnesota 181,685,664 125,729,499 31 44

Mississippi 102,313,225 76,627,646 25 71

Missouri 171,257,249 114,252,718 33 43

Montana 33,727,758 23,227,497 31 53

Nebraska 33,217,349 20,895,153 37 22

Nevada 119,817,390 83,921,190 30 65

New Hampshire 28,537,177 20,854,187 27 27

New Jersey 358,589,297 272,537,051 24 49

New Mexico 91,425,586 64,635,931 29 80

New York 719,957,636 422,437,436 41 35

North Carolina 224,528,255 131,563,618 41 31

North Dakota 17,319,233 12,681,948 27 30

Ohio 566,206,101 388,978,841 31 75

Oklahoma 92,363,644 63,161,433 32 35

Oregon 178,890,560 120,499,660 33 65

Pennsylvania 332,542,282 245,402,509 26 37

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

State Market value of liability

($ thousands)

Market value of unfunded liability

($ thousands) Funded ratio (percent)

Unfunded liability/ personal income

(percent)

Rhode Island 29,571,168 21,689,082 27 40

South Carolina 121,887,630 90,017,295 26 46

South Dakota 24,167,218 50,081,157 45 120

Tennessee 85,449,862 50,081,157 41 17

Texas 650,956,041 431,401,468 34 33

Utah 71,751,894 45,038,010 37 36

Vermont 14,674,700 10,669,525 27 34

Virginia 195,252,429 127,592,226 35 28

Washington 208,607,671 134,255,071 36 34

West Virginia 41,346,055 27,747,374 33 41

Wisconsin 166,029,890 70,633,690 57 26

Wyoming 23,732,567 15,869,303 33 49

Source: Authors’ analysis of the FY 2006–2016 CAFRs, pension actuarial reports, US Treasury daily yield curve rates, and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

TABLE B8. PENSION LIABILITIES DISCOUNTED UNDER RISK-FREE DISCOUNT RATE (FISCAL YEAR 2016) (CONTINUED)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B9. OTHER POSTEMPLOYMENT BENEFITS: RETIREE HEALTH BENEFITS (FISCAL YEAR 2016)

State

Total unfunded

OPEB liability ($ thousands)

Funded ratio

(percent)

OPEB/ personal income

(percent) State

Total unfunded OPEB liability ($ thousands)

Funded ratio

(percent)

OPEB/ personal income

(percent)

Alabama 9,478,603 13 5 Montana 458,429 0 1

Alaska 8,484,662 55 21 Nebraska n/a n/a n/a

Arizona 36,236 98 0 Nevada 1,445,333 0 1

Arkansas 124,711 0 0 New Hampshire 2,138,368 0 3

Californiaa 106,061,100 0 5 New Jersey 85,424,700 0 15

Colorado 1,845,893 17 1 New Mexico 3,805,064 11 5

Connecticut 21,887,478 1 9 New York 88,504,417 0 7

Delaware 7,729,000 4 17 North Carolina 32,467,020 4 8

Florida 20,554,898 1 2 North Dakota 87,700 53 0

Georgia 15,937,643 9 4 Ohio 15,142,634 52 3

Hawaii 9,065,926 2 13 Oklahoma 5,215 0 0

Idaho 115,982 21 0 Oregon 120,900 80 0

Illinois 51,898,621 0 8 Pennsylvania 20,724,570 1 3

Indiana 339,447 29 0 Rhode Island 644,316 18 1

Iowa 643,300 0 0 South Carolina 10,484,863 9 5

Kansas 5,657 0 0 South Dakota n/a n/a n/a

Kentucky 5,915,484 60 3 Tennessee 1,751,877 0 1

Louisiana 7,603,850 0 4 Texas 87,370,542 1 7

Maine 1,851,822 9 3 Utah 184,510 54 0

Maryland 11,789,450 2 3 Vermont 1,822,348 0 6

Massachusetts 16,322,500 4 4 Virginia 5,431,000 25 1

Michigan 17,992,900 21 4 Washington 13,750,912 0 4

Minnesota 666,638 0 0 West Virginia 3,060,099 18 4

Mississippi 709,077 0 1 Wisconsin 942,314 0 0

Missouri 3,182,984 5 1 Wyoming 243,728 0 1

Source: Authors’ analysis of the FY 2006–2016 CAFRs, OPEB actuarial reports, and Bureau of Economic Analysis Regional Economic Accounts for all 50 states.

Note: n/a = not available; OPEB = other postemployment benefits. a. California’s OPEB includes the unfunded liabilities for both the state’s OPEB plan and the California Employers’ Retiree Benefit Trust plan.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B10. PENSION PLANS (FISCAL YEAR 2016)

State Plans

Alabama Employees’ Retirement System of Alabama Teachers’ Retirement System of Alabama

Judicial Retirement Fund

Alaska

Public Employees’ Retirement System Teachers’ Retirement System Judicial Retirement System

National Guard and Naval Militia Retirement System Elected Public Officers Retirement System

Arizona

Arizona State Retirement System Public Safety Personnel Retirement Systems

Corrections Officer Retirement Plan Elected Officials’ Retirement Plan

Arkansas

Arkansas Public Employees’ Retirement System Arkansas District Judges’ Retirement System

Arkansas Teacher Retirement System Arkansas State Police Retirement System

Arkansas Judicial Retirement System Arkansas State Highway Employees’ Retirement System

California

Public Employees’ Retirement Fund Legislators’ Retirement Fund

Judges’ Retirement Fund Judges’ Retirement Fund II

California State Teachers’ Retirement System—Defined Benefit Plan California State Teachers’ Retirement System—Cash Balance Plan

California State Teachers’ Retirement System—DB Supplement University of California Retirement Plan

Colorado

Fire and Police Pension Association Fire and Police Pension Association—Hybrid Plan

State Division Trust Fund School Division Trust Fund

Local Government Division Trust Fund Judicial Division Trust Fund

Connecticut

State Employees’ Retirement System Teachers’ Retirement System Family Support Magistrates

Municipal Employees’ Retirement System Probate Judges’ and Employees’ Retirement System

Delaware

State Employees’ Plan New State Police Plan Revised Judicial Plan

Diamond State Port Corporation Plans Volunteer Fireman Pension Plans

County and Municipal Plan—General County and Municipal Plan—Police and Firefighter

Florida Florida Retirement System

Georgia

Employees’ Retirement System of Georgia Public School Employees’ Retirement System

Legislative Retirement System Georgia Judicial Retirement System

Georgia Military Pension Fund Teachers’ Retirement System

Firefighters’ Pension Fund

Hawaii Employees’ Retirement System

Police and Firefighters

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

State Plans

Idaho Public Employee Retirement System of Idaho

Firefighters’ Retirement Fund Judges’ Retirement Fund

Illinois

State Employees’ Retirement System Judges’ Retirement System

General Assembly Retirement System Teachers’ Retirement System

State Universities Retirement System Illinois Municipal Retirement System

Indiana

Public Employees’ Retirement Fund Teachers’ Retirement Fund

1977 Police Officers’ and Firefighters’ Pension and Disability Fund Judges’ Retirement System

State Excise Police, Gaming Agent, Gaming Control Officer, and Conservation Enforcement Officers’ Retirement Plan

Prosecuting Attorneys’ Retirement Fund Legislators’ Retirement System

Iowa

Iowa Public Employees’ Retirement System Judicial Retirement Fund

Peace Officers’ Retirement, Accident, and Disability System Municipal Fire and Police Retirement System

Kansas Kansas Public Employees’ Retirement System

Kansas Police and Firemen’s Retirement System Kansas Retirement System for Judges

Kentucky

Kentucky Employees’ Retirement System Teachers’ Retirement System

Judicial Retirement Plan Legislative Retirement Plan

Louisiana

Firefighters’ Retirement System Louisiana State Employees’ Retirement System

Teachers’ Retirement System of Louisiana Louisiana School Employees’ Retirement System

Louisiana State Police Retirement System

Maine

Maine Public Employees’ Retirement System Maine Judicial Retirement Program

Maine Legislative Retirement Program Maine Public Employees’ Retirement System

Consolidated Plan for Participating Local Districts

Maryland

Teachers’ Retirement System Employees’ Retirement System State Police Retirement System

Judges’ Retirement System Law Enforcement Officers’ Pension System Correctional Officers’ Retirement System

Employees’ Retirement System—Municipal Law Enforcement Officers’ Pension System—Municipal

Massachusetts State Employees’ Retirement System

Massachusetts Retirement System

Michigan

Legislative Retirement System State Police Retirement System

State Employees’ Retirement System Public School Employees’ Retirement System

Judges’ Retirement System Municipal Employees’ Retirement System of Michigan

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

TABLE B10. PENSION PLANS (FISCAL YEAR 2016) (CONTINUED)

State Plans

Minnesota

State Employees’ Retirement Fund State Patrol Retirement Fund

Correctional Employees’ Retirement Fund Statewide “specialty” retirement plans (judges, elected officials, and legislators)

General Employees’ Retirement Fund Public Employees’ Police and Fire Fund Public Employees’ Correctional Fund

Municipal Employees’ Retirement Fund Teachers’ Retirement Association

Mississippi

Public Employees’ Retirement System Mississippi Highway Safety Patrol Retirement System

Municipal Retirement System Supplemental Legislative Retirement System

Missouri

Missouri State Employees’ Plan Judicial Plan

Missouri Department of Transportation and Highway Patrol Employees’ Retirement System University of Missouri Retirement Plan

Public School Retirement System Public Education Employee Retirement System

Montana

Public Employees’ Retirement System Firefighters’ United Retirement System

Sheriffs’ Retirement System Highway Patrol Officers’ Retirement System

Game Wardens’ & Peace Officers’ Retirement System Judges’ Retirement System

Montana Municipal Police Officers’ Retirement System Volunteer Firefighters’ Compensation System

Teachers’ Retirement System

Nebraska

Nebraska School Employees’ Retirement System Nebraska Judges’ Retirement System

Nebraska State Patrol Retirement System State Employees’ Retirement Benefit Fund

County Employees’ Retirement System

Nevada Public Employees’ Retirement System

New Hampshire New Hampshire Retirement System

Judicial Retirement Plan

New Jersey

Public Employees’ Retirement System (State) Public Employees’ Retirement System (Local)

Teachers’ Pension and Annuity Fund State Police Retirement System

Judicial Retirement System Police and Firemen’s Retirement System (State) Police and Firemen’s Retirement System (Local)

New Mexico

Public Employees’ Retirement Fund Legislative Retirement Fund

Judicial Retirement Fund Magistrate Retirement Fund

Volunteer Firefighters’ Retirement Fund Educational Retirement Board

New York Employees’ Retirement System

Police and Fire Retirement System Teachers’ Retirement System

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

State Plans

North Carolina

Teachers’ and State Employees’ Retirement System Consolidated Judicial Retirement System

Legislative Retirement System Local Government Employees’ Retirement System

Firefighters’ and Rescue Squad Workers’ Pension Fund Registers of Deeds’ Supplemental Pension Fund

National Guard Pension Fund

North Dakota

Public Employees’ Retirement System Highway Patrolmen’s Retirement System

Retirement Plan for Employees of Job Service North Dakota Teachers’ Fund for Retirement

Ohio

Ohio Public Employees Retirement System School Employees’ Retirement System

State Teachers’ Retirement System Police and Fire Pension Fund

Oklahoma

Oklahoma Public Employees’ Retirement System Teachers’ Retirement System of Oklahoma

Uniform Retirement System for Justices and Judges Oklahoma Firefighters’ Pension and Retirement System

Oklahoma Police Pension and Retirement System Oklahoma Law Enforcement Retirement System

Oregon

Public Employees’ Retirement System State Employees’ Retirement System

Public School Employees’ Retirement System Municipal Retirement System

Pennsylvania State Employees’ Retirement System

Public School Employees’ Retirement System Municipal Retirement System

Rhode Island

Employees’ Retirement System of Rhode Island Teachers’ Retirement System

Municipal Employees’ Retirement System Judicial Retirement Board Trust

State Police Retirement Board Trust

South Carolina

South Carolina Retirement System Police Officers’ Retirement System

General Assembly Retirement System Judges’ and Solicitors’ Retirement System

National Guard Retirement System

South Dakota South Dakota Retirement System

Tennessee Tennessee Consolidated Retirement System

Tennessee Consolidated Retirement System—Hybrid

Texas

Employees’ Retirement System Law Enforcement and Custodial Officer Supplemental Retirement Fund

Judicial Retirement System I Judicial Retirement System II Teacher Retirement System

Municipal Retirement System County and District Retirement System Emergency Services Retirement System

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

State Plans

Utah

Noncontributory Retirement System Contributory Retirement System Public Safety Retirement System Firefighters’ Retirement System

Judges’ Retirement System Utah Governors’ and Legislators’ Retirement Plan

Tier 2 Public Employees’ Retirement System Tier 2 Public Safety and Firefighters’ Retirement System

Vermont State Employees’ Retirement System State Teachers’ Retirement System

Municipal Employees’ Retirement System

Virginia

Virginia Retirement System State Police Officers’ Retirement System Virginia Law Officers’ Retirement System

Judicial Retirement System Political Subdivisions State Employees—Teachers

Washington

Public Employees’ Retirement System Plan 1 Public Employees’ Retirement System Plan 2/3

Teachers’ Retirement System Plan 1 Teachers’ Retirement System Plan 2/3 School Employees’ Retirement System

Law Enforcement Officers’ and Fire Fighters’ Retirement Plan 1 Law Enforcement Officers’ and Fire Fighters’ Retirement Plan 2

Washington State Patrol Retirement System Public Safety Employees’ Retirement System

Judicial Retirement System

West Virginia

Public Employees’ Retirement System Deputy Sheriff Retirement System

Emergency Medical Services Retirement System Municipal Police Officers’ and Firefighters’ Retirement System

Teachers’ Retirement System Public Safety Death, Disability, and Retirement Fund

State Police Retirement System Judges’ Retirement System

Wisconsin Wisconsin Retirement System

Wyoming

Public Employees’ Pension Plan State Patrol, Game and Fish Warden, and Criminal Investigator Plan

Volunteer Firefighters’ Pension Plan Paid Firemen’s Pension Plan A Paid Firemen’s Pension Plan B

Judicial Pension Plan Law Enforcement Pension Plan

Volunteer Emergency Medical Technician Pension Plan Air Guard Firefighters’ Pension Plan

TABLE B10. PENSION PLANS (FISCAL YEAR 2016) (CONTINUED)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

APPENDIX C. STATE PROFILES

This section contains a summary of each state’s key metrics and data for fis- cal solvency, including total debt; pension liability and OPEB obligations; and underlying ratios for cash, budget, long-run, service-level, and trust fund sol- vencies. Each summary begins with a brief analysis of the state’s overall fiscal performance alongside the state’s relative rank and position for each dimension of solvency.

As noted in the study, the data and metrics can only provide a basic picture of a state’s fiscal condition. Relative ranking is not as meaningful as the under- lying fiscal indicators for a state. The metrics provided here should be used in conjunction with other data and analysis of state economic conditions and fiscal and budgetary institutions.

Key to the State Profiles

The state’s five fiscal categories have been mapped on a vertical number line on the second page of each state profile, along with its ranking for that measure. The markers represent the distance of that category from the US average. Markers that fall outside the range of +3.0 to –3.0 standard deviations are represented by up or down arrows.

The labels for each of the markers are stacked in descending order along the line. They are also color coded with the adjacent table of the state’s underly- ing ratios.

The key terms, explained on the second page, help lawmakers and others ask questions about the fiscal health of their state.

n/a = not applicable

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

SUMMARY

On the basis of its solvency in five separate categories, Alabama ranks 14th among the US states for fiscal health. Alabama has between 3.66 and 4.89 times the cash needed to cover short-term obligations. Revenues exceed expenses by 3 percent, with an improving net position of $142 per capita. In the long run, a net asset ratio of 0.01 indicates that Alabama does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 31 percent of total assets, or $2,118 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $86.97 billion, or 46 percent of state personal income. OPEB are $9.48 bil- lion, or 5 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Alabama $0.72 billion $5.19 billion $190.79 billion 2.7% $1,068

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded ratio

Alabama $15.79 billion 68% $86.97 billion 28%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Alabama $9.48 billion 13%

National average $14.51 billion 14%

1. Nebraska 2. South Dakota 3. Tennessee 4. Florida 5. Oklahoma 6. Wyoming 7. Idaho 8. Utah 9. North Carolina 10. Nevada 11. Alaska 12. New Hampshire 13. Virginia 14. Alabama 15. Missouri 16. Montana 17. Kansas 18. Georgia 19. North Dakota 20. South Carolina 21. Indiana 22. Texas 23. Ohio 24. Minnesota 25. Arkansas 26. Wisconsin 27. Arizona 28. Colorado 29. Iowa 30. Washington 31. Oregon 32. Michigan 33. Maryland 34. Maine 35. Pennsylvania 36. Mississippi 37. Louisiana 38. Hawaii 39. Vermont 40. Rhode Island 41. New York 42. California 43. West Virginia 44. Delaware 45. New Mexico 46. Kentucky 47. Massachusetts 48. New Jersey 49. Connecticut 50. Illinois

ALABAMA rank

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

distance from US average

(in standard deviations)

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Alabama 3.66 4.23 4.89 1.03 $142 0.01 0.31 $2,118

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Alabama 0.05 0.12 0.11 0.46 0.05

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Alabama ranks 6th.)

• Budget solvency measures whether a state can cover its fiscal year spending out of current revenues. Did it run a shortfall during the year? (Alabama ranks 22nd.)

• Long-run solvency measures whether a state has a hedge against large long-term liabilities. Are enough assets available to cushion the state from potential shocks or long-term fiscal risks? (Alabama ranks 19th.)

• Service-level solvency measures how high taxes, revenues, and spending are when compared to state personal income. Do states have enough “fiscal slack”? If spending commitments demand more revenues, are states in a good position to increase taxes without harming the economy? Is spending high or low relative to the tax base? (Alabama ranks 21st.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Alabama ranks 34th.)

22nd budget

solvency

34th trust fund solvency

19th long-run solvency

21st service-level

solvency

6th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

ALASKA rank

SUMMARY

On the basis of its solvency in five separate categories, Alaska ranks 11th among the US states for fiscal health. Alaska has between 17.07 and 17.92 times the cash needed to cover short-term obligations. However, much of this revenue is part of the Alaska Permanent Fund and is not readily avail- able for spending. With the fall of oil prices between FY2014 and FY2016, Alaska’s budgetary position has weakened significantly. Revenues only cover 52 percent of expenses, with a worsening net position of –$6,946 per capita. In the long run, Alaska has a net asset ratio of 0.77. Long-term liabili- ties are higher than the national average in per capita terms at $8,670 per capita, but lower than the national average when measured as a percentage of total assets. Total unfunded pension liabilities that are guaranteed to be paid are $37.33 billion, or 91 percent of state personal income. OPEB are $8.48 billion, or 21 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt

State personal income

Ratio of debt to state personal

income

Total primary debt per capita

Alaska $0.92 billion $2.09 billion $41.03 billion 5.1% $2,821

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability

Funded ratio Market value of unfunded liability

Market value of funded liability ratio

Alaska $6.37 billion 70% $37.33 billion 28%

National average $23.43 billion 73% $135.50 billion 32%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

50th budget

solvency

50th trust fund solvency

6th long-run solvency

2nd service-level

solvency

1st cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Alaska $8.48 billion 55%

National average $14.51 billion 14%

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Alaska 17.07 17.38 17.92 0.52 –$6,946 0.77 0.08 $8,670

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Alaska 0.00 0.14 0.26 0.91 0.21

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Alaska ranks 1st.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Alaska ranks 50th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Alaska ranks 50th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

SUMMARY

On the basis of its solvency in five separate categories, Arizona ranks 27th among the US states for fiscal health. Arizona has between 0.88 and 1.40 times the cash needed to cover short-term obligations, well below the US average. Revenues exceed expenses by 5 percent, with an improving net position of $227 per capita. In the long run, Arizona has a net asset ratio of 0.07. Long-term liabilities are lower than the national average, at 34 percent of total assets, or $2,194 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $110.43 billion, or 40 percent of state per- sonal income. OPEB are $0.04 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Arizona $0.00 $9.50 billion $278.92 billion 3.4% $1,371

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Arizona $19.81 billion 69% $110.43 billion 29%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Arizona $0.04 billion 98%

National average $14.51 billion 14%

ARIZONA 1. Nebraska 2. South Dakota 3. Tennessee 4. Florida 5. Oklahoma 6. Wyoming 7. Idaho 8. Utah 9. North Carolina 10. Nevada 11. Alaska 12. New Hampshire 13. Virginia 14. Alabama 15. Missouri 16. Montana 17. Kansas 18. Georgia 19. North Dakota 20. South Carolina 21. Indiana 22. Texas 23. Ohio 24. Minnesota 25. Arkansas 26. Wisconsin 27. Arizona 28. Colorado 29. Iowa 30. Washington 31. Oregon 32. Michigan 33. Maryland 34. Maine 35. Pennsylvania 36. Mississippi 37. Louisiana 38. Hawaii 39. Vermont 40. Rhode Island 41. New York 42. California 43. West Virginia 44. Delaware 45. New Mexico 46. Kentucky 47. Massachusetts 48. New Jersey 49. Connecticut 50. Illinois

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

15th budget

solvency

9th trust fund solvency

20th long-run solvency

26th service-level

solvency

46th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Arizona 0.88 1.14 1.40 1.05 $227 0.07 0.34 $2,194

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to- income ratio

OPEB-to-income ratio

Arizona 0.05 0.13 0.12 0.40 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Arizona ranks 46th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Arizona ranks 15th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Arizona ranks 9th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

ARKANSAS

SUMMARY

On the basis of its solvency in five separate categories, Arkansas ranks 25th among the US states for fiscal health. Arkansas has between 3.17 and 4.14 times the cash needed to cover short-term obligations. Revenues exceed expenses by 4 percent, with an improving net position of $248 per capita. In the long run, Arkansas has a net asset ratio of 0.11. Long-term liabilities are lower than the national average, at 35 percent of total assets, or $2,986 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $48.17 billion, or 41 percent of state personal income. OPEB are $0.12 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Arkansas $1.52 billion $3.96 billion $117.57 billion 3.4% $1,325

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Arkansas $5.78 billion 81% $48.17 billion 34%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Arkansas $0.12 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

20th budget

solvency

26th trust fund solvency

23rd long-run solvency

45th service-level

solvency

11th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Arkansas 3.17 3.86 4.14 1.04 $248 0.11 0.35 $2,986

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Arkansas 0.07 0.18 0.18 0.41 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Arkansas ranks 11th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Arkansas ranks 20th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Arkansas ranks 26th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

SUMMARY

On the basis of its solvency in five separate categories, California ranks 42nd among the US states for fiscal health. California has between 0.82 and 1.62 times the cash needed to cover short-term obligations, well below the US average. Revenues exceed expenses by 4 percent, with an improving net position of $271 per capita. In the long run, California’s negative net asset ratio of 0.57 points to the use of debt and large unfunded obligations. Long-term liabilities are higher than the national average, at 92 percent of total assets, or $5,642 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $1,190.84 billion, or 54 percent of state personal income. OPEB are $106.06 billion, or 5 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

California $79.04 billion $112.55 billion $2,197.49 billion 5.1% $2,868

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

California $222.19 billion 71% $1,190.84 billion 31%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

California $106.06 billion 0%

National average $14.51 billion 14%

CALIFORNIA 1. Nebraska 2. South Dakota 3. Tennessee 4. Florida 5. Oklahoma 6. Wyoming 7. Idaho 8. Utah 9. North Carolina 10. Nevada 11. Alaska 12. New Hampshire 13. Virginia 14. Alabama 15. Missouri 16. Montana 17. Kansas 18. Georgia 19. North Dakota 20. South Carolina 21. Indiana 22. Texas 23. Ohio 24. Minnesota 25. Arkansas 26. Wisconsin 27. Arizona 28. Colorado 29. Iowa 30. Washington 31. Oregon 32. Michigan 33. Maryland 34. Maine 35. Pennsylvania 36. Mississippi 37. Louisiana 38. Hawaii 39. Vermont 40. Rhode Island 41. New York 42. California 43. West Virginia 44. Delaware 45. New Mexico 46. Kentucky 47. Massachusetts 48. New Jersey 49. Connecticut 50. Illinois

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

17th budget

solvency

41st trust fund solvency

45th long-run solvency

28th service-level

solvency

45th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

California 0.82 1.19 1.62 1.04 $271 –0.57 0.92 $5,642

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

California 0.06 0.13 0.12 0.54 0.05

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (California ranks 45th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (California ranks 17th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (California ranks 41st.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

COLORADO

SUMMARY

On the basis of its solvency in five separate categories, Colorado ranks 28th among the US states for fiscal health. Colorado has between 1.32 and 2.11 times the cash needed to cover short-term obligations. Revenues exceed expenses by 1 percent, with an improving net position of $40 per capita. In the long run, a net asset ratio of –0.02 indicates that Colorado does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 48 percent of total assets, or $3,175 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $122.81 billion, or 43 percent of state personal income. OPEB are $1.85 bil- lion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Colorado $0.00 $6.30 billion $288.43 billion 2.2% $1,137

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Colorado $31.16 billion 58% $122.81 billion 26%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Colorado $1.85 billion 12%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

32nd budget

solvency

31st trust fund solvency

33rd long-run solvency

15th service-level

solvency

32nd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Colorado 1.32 1.93 2.11 1.01 $40 –0.02 0.48 $3,175

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Colorado 0.04 0.11 0.11 0.43 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Colorado ranks 32nd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Colorado ranks 32nd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Colorado ranks 31st.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

CONNECTICUT

SUMMARY

On the basis of its solvency in five separate categories, Connecticut ranks 49th among the US states for fiscal health. Connecticut has between 0.42 and 1.05 times the cash needed to cover short-term obligations, well below the US average. Revenues only cover 92 percent of expenses, with a worsening net position of –$693 per capita. In the long run, Connecticut’s negative net asset ratio of 1.71 points to the use of debt and large unfunded obligations. Long- term liabilities are higher than the national average, at 230 percent of total assets, or $17,418 per capita. Total unfunded pension liabilities that are guar- anteed to be paid are $121.65 billion, or 48 percent of state personal income. OPEB are $21.89 billion, or 9 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Connecticut $17.39 billion $23.55 billion $254.05 billion 9.3% $6,584

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Connecticut $35.37 billion 47% $121.65 billion 21%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Connecticut $21.89 billion 1%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

48th budget

solvency

36th trust fund solvency

47th long-run solvency

27th service-level

solvency

50th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Connecticut 0.42 1.00 1.05 0.92 –$693 –1.71 2.30 $17,418

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Connecticut 0.06 0.12 0.13 0.48 0.09

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Connecticut ranks 50th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Con- necticut ranks 48th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Connecticut ranks 36th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

DELAWARE

SUMMARY

On the basis of its solvency in five separate categories, Delaware ranks 44th among the US states for fiscal health. Delaware has between 1.34 and 1.98 times the cash needed to cover short-term obligations. Revenues only cover 96 percent of expenses, with a worsening net position of –$377 per capita. In the long run, Delaware has a net asset ratio of –0.15. Long-term liabilities are higher than the national average in per capita terms at $7,537 per capita, but slightly lower than the national average when measured as a percent- age of total assets. Total unfunded pension liabilities that are guaranteed to be paid are $13.75 billion, or 30 percent of state personal income. OPEB are $7.73 billion, or 17 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Delaware $2.12 billion $3.27 billion $46.36 billion 7.1% $3,436

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Delaware $1.10 billion 89% $13.75 billion 40%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Delaware $7.73 billion 4%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

42nd budget

solvency

11th trust fund solvency

40th long-run solvency

48th service-level

solvency

33rd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Delaware 1.34 1.95 1.98 0.96 –$377 –0.15 0.61 $7,537

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Delaware 0.09 0.18 0.19 0.30 0.17

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Delaware ranks 33rd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Delaware ranks 42nd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Delaware ranks 11th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

FLORIDA

SUMMARY

On the basis of its solvency in five separate categories, Florida ranks 4th among the US states for fiscal health. Florida has between 4.80 and 5.81 times the cash needed to cover short-term obligations, well above the US average. Revenues exceed expenses by 7 percent, with an improving net position of $277 per capita. In the long run, Florida has a net asset ratio of 0.12. Long-term liabilities are lower than the national average, at 31 percent of total assets, or $2,199 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $253.01 billion, or 27 percent of state personal income. OPEB are $20.55 billion, or 2 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Florida $10.71 billion $25.17 billion $944.44 billion 2.7% $1,221

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Florida $24.90 billion 85% $253.01 billion 37%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Florida $20.55 billion 1%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

6th budget

solvency 7th

trust fund solvency

17th long-run solvency

5th service-level

solvency

4th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Florida 4.80 5.80 5.81 1.07 $277 0.12 0.31 $2,199

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Florida 0.04 0.09 0.09 0.27 0.02

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Florida ranks 4th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Florida ranks 6th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Florida ranks 7th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

GEORGIA

SUMMARY

On the basis of its solvency in five separate categories, Georgia ranks 18th among the US states for fiscal health. Georgia has between 2.13 and 3.24 times the cash needed to cover short-term obligations. Revenues exceed expenses by 7 percent, with an improving net position of $331 per capita. In the long run, a net asset ratio of –0.01 indicates that Georgia does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 47 percent of total assets, or $2,302 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $172.77 billion, or 40 percent of state personal income. OPEB are $15.94 bil- lion, or 4 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Georgia $9.49 billion $14.10 billion $431.33 billion 3.3% $1,367

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Georgia $28.27 billion 75% $172.77 billion 32%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Georgia $15.94 billion 9%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

3rd budget

solvency

28th trust fund solvency

27th long-run solvency

19th service-level

solvency 16th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

27th

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Georgia 2.13 3.13 3.24 1.07 $331 –0.01 0.47 $2,302

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Georgia 0.05 0.12 0.11 0.40 0.04

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Georgia ranks 16th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Georgia ranks 3rd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Georgia ranks 28th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

HAWAII

SUMMARY

On the basis of its solvency in five separate categories, Hawaii ranks 38th among the US states for fiscal health. Hawaii has between 2.22 and 2.91 times the cash needed to cover short-term obligations. Revenues exceed expenses by 5 percent, with an improving net position of $332 per capita. In the long run, Hawaii has a net asset ratio of –0.16. Long-term liabilities are higher than the national average, at 84 percent of total assets, or $12,056 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $44.01 billion, or 61 percent of state personal income. OPEB are $9.07 billion, or 13 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Hawaii $6.29 billion $8.67 billion $72.21 billion 12.0% $6,067

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Hawaii $12.44 billion 55% $44.01 billion 25%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Hawaii $9.07 billion 2%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

100

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

11th budget

solvency

44th trust fund solvency

42nd long-run solvency

42nd service-level

solvency

18th cash

solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Hawaii 2.22 2.77 2.91 1.05 $332 –0.16 0.84 $12,056

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Hawaii 0.09 0.16 0.15 0.61 0.13

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Hawaii ranks 18th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Hawaii ranks 11th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Hawaii ranks 44th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

101

SUMMARY

On the basis of its solvency in five separate categories, Idaho ranks 7th among the US states for fiscal health. Idaho has between 3.57 and 4.66 times the cash needed to cover short-term obligations, well above the US average. Revenues exceed expenses by 5 percent, with an improving net position of $240 per capita. In the long run, Idaho has a net asset ratio of 0.37. Long- term liabilities are lower than the national average, at 11 percent of total assets, or $963 per capita. Total unfunded pension liabilities that are guar- anteed to be paid are $23.78 billion, or 36 percent of state personal income. OPEB are $0.12 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Idaho $0.00 $1.23 billion $65.82 billion 1.9% $730

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Idaho $2.20 billion 87% $23.78 billion 38%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Idaho $0.12 billion 21%

National average $14.51 billion 14%

IDAHO 1. Nebraska 2. South Dakota 3. Tennessee 4. Florida 5. Oklahoma 6. Wyoming 7. Idaho 8. Utah 9. North Carolina 10. Nevada 11. Alaska 12. New Hampshire 13. Virginia 14. Alabama 15. Missouri 16. Montana 17. Kansas 18. Georgia 19. North Dakota 20. South Carolina 21. Indiana 22. Texas 23. Ohio 24. Minnesota 25. Arkansas 26. Wisconsin 27. Arizona 28. Colorado 29. Iowa 30. Washington 31. Oregon 32. Michigan 33. Maryland 34. Maine 35. Pennsylvania 36. Mississippi 37. Louisiana 38. Hawaii 39. Vermont 40. Rhode Island 41. New York 42. California 43. West Virginia 44. Delaware 45. New Mexico 46. Kentucky 47. Massachusetts 48. New Jersey 49. Connecticut 50. Illinois

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

102

13th budget

solvency

21st trust fund solvency

5th long-run solvency

29th service-level

solvency

7th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

trust fund

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Idaho 3.57 4.36 4.66 1.05 $240 0.37 0.11 $963

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Idaho 0.06 0.13 0.13 0.36 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Idaho ranks 7th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Idaho ranks 13th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Idaho ranks 21st.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

103

ILLINOIS

SUMMARY

On the basis of its solvency in five separate categories, Illinois ranks 50th among the US states for fiscal health. Illinois has between 0.55 and 1.13 times the cash needed to cover short-term obligations, well below the US average. Revenues only cover 92 percent of expenses, with a worsening net position of –$450 per capita. In the long run, Illinois’s negative net asset ratio of 2.86 points to the use of debt and large unfunded obligations. Long-term liabili- ties are higher than the national average, at 330 percent of total assets, or $12,816 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $445.79 billion, or 67 percent of state personal income. OPEB are $51.90 billion, or 8 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Illinois $26.80 billion $31.26 billion $666.94 billion 4.7% $2,442

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Illinois $131.09 billion 47% $445.79 billion 21%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Illinois $51.90 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

104

46th budget

solvency

46th trust fund solvency

49th long-run solvency

14th service-level

solvency

49th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Illinois 0.55 0.92 1.13 0.92 –$450 –2.86 3.30 $12,816

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Illinois 0.05 0.10 0.11 0.67 0.08

National average

0.06 0.13 0.13 0.35 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Illinois ranks 49th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Illinois ranks 46th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Illinois ranks 46th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

105

INDIANA

SUMMARY

On the basis of its solvency in five separate categories, Indiana ranks 21st among the US states for fiscal health. Indiana has between 1.37 and 2.68 times the cash needed to cover short-term obligations. Revenues cover 100 percent of expenses, with a worsening net position of –$14 per capita. In the long run, Indiana has a net asset ratio of –0.13. Long-term liabilities are lower than the national average, at 50 percent of total assets, or $2,155 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $67.11 billion, or 23 percent of state personal income. OPEB are$0.34 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Indiana $0.00 $1.00 billion $288.49 billion 0.3% $151

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Indiana $16.08 billion 66% $67.11 billion 32%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Indiana $0.34 billion 29%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

106

36th budget

solvency

5th trust fund solvency

31st long-run solvency

18th service-level

solvency

27th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Indiana 1.37 2.06 2.68 1.00 –$14 –0.13 0.50 $2,155

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Indiana 0.05 0.11 0.11 0.23 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Indiana ranks 27th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Indiana ranks 36th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Indiana ranks 5th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

107

IOWA

SUMMARY

On the basis of its solvency in five separate categories, Iowa ranks 29th among the US states for fiscal health. Iowa has between 1.39 and 2.47 times the cash needed to cover short-term obligations. Revenues exceed expenses by 3 percent, with an improving net position of $182 per capita. In the long run, Iowa has a net asset ratio of 0.16. Long-term liabilities are lower than the national average, at 22 percent of total assets, or $1,656 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $56.39 bil- lion, or 38 percent of state personal income. OPEB are $0.64 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Iowa $0.00 $3.65 billion $146.69 billion 2.5% $1,164

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Iowa $6.31 billion 84% $56.39 billion 36%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Iowa $0.64 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

108

23rd budget

solvency

25th trust fund solvency

12th long-run solvency

41st service-level

solvency

26th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Iowa 1.39 2.36 2.47 1.03 $182 0.16 0.22 $1,656

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Iowa 0.06 0.15 0.14 0.38 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Iowa ranks 26th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Iowa ranks 23rd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Iowa ranks 25th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

109

KANSAS

SUMMARY

On the basis of its solvency in five separate categories, Kansas ranks 17th among the US states for fiscal health. Kansas has between 0.80 and 1.62 times the cash needed to cover short-term obligations, well below the US average. Revenues only cover 94 percent of expenses, with a worsening net position of –$283 per capita. In the long run, a net asset ratio of –0.05 indi- cates that Kansas does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 41 per- cent of total assets, or $2,527 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $46.98 billion, or 33 percent of state per- sonal income. OPEB are $0.01 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Kansas $0.00 $7.75 billion $141.11 billion 5.5% $2,664

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Kansas $9.06 billion 67% $46.98 billion 28%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Kansas $0.01 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

110

44th budget

solvency

2nd trust fund solvency

28th long-run solvency

10th service-level

solvency

42nd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Kansas 0.80 1.60 1.62 0.94 –$283 –0.05 0.41 $2,527

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Kansas 0.05 0.10 0.10 0.33 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Kansas ranks 42nd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Kansas ranks 44th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Kansas ranks 2nd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

111

KENTUCKY

SUMMARY

On the basis of its solvency in five separate categories, Kentucky ranks 46th among the US states for fiscal health. Kentucky has between 0.87 and 1.75 times the cash needed to cover short-term obligations, well below the US average. Revenues only cover 98 percent of expenses, with a worsening net position of –$125 per capita. In the long run, Kentucky’s negative net asset ratio of 1.15 points to the use of debt and large unfunded obligations. Long- term liabilities are higher than the national average, at 138 percent of total assets, or $9,960 per capita. Total unfunded pension liabilities that are guar- anteed to be paid are $106.59 billion, or 61 percent of state personal income. OPEB are $5.92 billion, or 3 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Kentucky $0.00 $7.69 billion $175.26 billion 4.4% $1,734

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Kentucky $32.66 billion 47% $106.59 billion 21%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Kentucky $5.92 billion 60%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

112

39th budget

solvency

43rd trust fund solvency

46th long-run solvency

43rd service-level

solvency

40th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Kentucky 0.87 1.52 1.75 0.98 –$125 –1.15 1.38 $9,960

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Kentucky 0.07 0.15 0.16 0.61 0.03

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Kentucky ranks 40th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Kentucky ranks 39th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Kentucky ranks 43rd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

113

LOUISIANA

SUMMARY

On the basis of its solvency in five separate categories, Louisiana ranks 37th among the US states for fiscal health. Louisiana has between 1.27 and 2.48 times the cash needed to cover short-term obligations. Revenues only cover 96 percent of expenses, with an improving net position of $11 per capita. In the long run, Louisiana’s negative net asset ratio of 0.2 points to the use of debt and unfunded obligations. Louisiana’s long-term liabilities are at about the same level of the US national average, at 65 percent of total assets, or $4,133 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $100.41 billion, or 49 percent of state personal income. OPEB are $7.60 billion, or 4 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Louisiana $4.61 billion $12.26 billion $203.59 billion 6.0% $2,620

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Louisiana $20.37 billion 66% $100.41 billion 28%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Louisiana $7.60 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

114

40th budget

solvency

39th trust fund solvency

38th long-run solvency

24th service-level

solvency

31st cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Louisiana 1.27 2.01 2.48 0.96 $11 –0.20 0.65 $4,133

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Louisiana 0.04 0.12 0.13 0.49 0.04

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Louisiana ranks 31st.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Louisiana ranks 40th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Louisiana ranks 39th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

115

MAINE

SUMMARY

On the basis of its solvency in five separate categories, Maine ranks 34th among the US states for fiscal health. Maine has between 0.65 and 2.02 times the cash needed to cover short-term obligations. Revenues exceed expenses by 4 percent, with an improving net position of $252 per capita. In the long run, Maine’s negative net asset ratio of 0.21 points to the use of debt and unfunded obligations. Long-term liabilities are lower than the national average, at 56 percent of total assets, or $2,812 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $20.80 billion, or 35 percent of state personal income. OPEB are $1.85 billion, or 3 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Maine $0.46 billion $1.17 billion $59.01 billion 2.0% $877

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Maine $2.95 billion 82% $20.80 billion 39%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Maine $1.85 billion 9%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

116

16th budget

solvency

22nd trust fund solvency

35th long-run solvency

33rd service-level

solvency

43rd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

trust fund

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Maine 0.65 1.30 2.02 1.04 $252 –0.21 0.56 $2,812

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Maine 0.06 0.14 0.13 0.35 0.03

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Maine ranks 43rd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Maine ranks 16th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Maine ranks 22nd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

117

MARYLAND

SUMMARY

On the basis of its solvency in five separate categories, Maryland ranks 33rd among the US states for fiscal health. Maryland has between 0.75 and 1.75 times the cash needed to cover short-term obligations, below the US aver- age. Revenues exceed expenses by 2 percent, with an improving net position of $130 per capita. In the long run, Maryland’s negative net asset ratio of 0.48 points to the use of debt and unfunded obligations. Long-term liabilities are higher than the national average, at 99 percent of total assets, or $7,186 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $117.10 billion, or 34 percent of state personal income. OPEB are $11.79 bil- lion, or 3 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Maryland $9.47 billion $18.32 billion $348.57 billion 5.3% $3,045

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Maryland $20.84 billion 71% $117.10 billion 31%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Maryland $11.79 billion 2%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

118

27th budget

solvency 17th

trust fund solvency

44th long-run solvency

17th service-level

solvency

41st cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Maryland 0.75 1.60 1.75 1.02 $130 –0.48 0.99 $7,186

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Maryland 0.06 0.11 0.11 0.34 0.03

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Maryland ranks 41st.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Maryland ranks 27th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Maryland ranks 17th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

119

MASSACHUSETTS

SUMMARY

On the basis of its solvency in five separate categories, Massachusetts ranks 47th among the US states for fiscal health. Massachusetts has between 0.48 and 1.16 times the cash needed to cover short-term obligations, well below the US average. Revenues only cover 95 percent of expenses, with a worsening net position of –$491 per capita. In the long run, Massachusetts’s negative net asset ratio of 1.93 points to the use of debt and unfunded obligations. Long- term liabilities are higher than the national average, at 275 percent of total assets, or $11,518 per capita. Total unfunded pension liabilities that are guar- anteed to be paid are $144.58 billion, or 33 percent of state personal income. OPEB are $16.32 billion, or 4 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Massachusetts $21.67 billion $29.57 billion $443.70 billion 6.7% $4,341

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Massachusetts $35.47 billion 58% $144.58 billion 25%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Massachusetts $16.32 billion 4%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

120

45th budget

solvency

16th trust fund solvency

48th long-run solvency

35th service-level

solvency

48th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio

Quick ratio

Current ratio

Operating ratio

Surplus (or deficit) per

capita Net asset

ratio

Long-term liability

ratio

Long-term liability

per capita

Massachusetts 0.48 1.11 1.16 0.95 –$491 –1.93 2.75 $11,518

National average 2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to- income ratio

OPEB-to-income ratio

Massachusetts 0.06 0.13 0.14 0.33 0.04

National average 0.06 0.13 0.13 0.35 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Massachusetts ranks 48th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Massa- chusetts ranks 45th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Massachusetts ranks 16th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

121

MICHIGAN

SUMMARY

On the basis of its solvency in five separate categories, Michigan ranks 32nd among the US states for fiscal health. Michigan has between 1.04 and 2.27 times the cash needed to cover short-term obligations. Revenues exceed expenses by 3 percent, with an improving net position of $160 per capita. In the long run, Michigan has a net asset ratio of –0.1. Long-term liabilities are lower than the national average, at 45 percent of total assets, or $1,883 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $184.08 billion, or 42 percent of state personal income. OPEB are $17.99 billion, or 4 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Michigan $1.63 billion $7.31 billion $440.29 billion 1.7% $737

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Michigan $37.89 billion 62% $184.08 billion 25%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Michigan $17.99 billion 21%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

122

25th budget

solvency

30th trust fund solvency

26th long-run solvency

31st service-level

solvency

35th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

26th

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Michigan 1.04 1.73 2.27 1.03 $160 –0.10 0.45 $1,883

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Michigan 0.06 0.13 0.13 0.42 0.04

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Michigan ranks 35th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Michigan ranks 25th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Michigan ranks 30th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

123

MINNESOTA

SUMMARY

On the basis of its solvency in five separate categories, Minnesota ranks 24th among the US states for fiscal health. Minnesota has between 2.32 and 3.01 times the cash needed to cover short-term obligations. Revenues exceed expenses by 5 percent, with an improving net position of $313 per capita. In the long run, Minnesota has a net asset ratio of 0.07. Long-term liabilities are lower than the national average, at 36 percent of total assets, or $2,458 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $125.73 billion, or 44 percent of state personal income. OPEB are $0.67 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Minnesota $7.04 billion $9.16 billion $287.68 billion 3.2% $1,659

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Minnesota $17.53 billion 76% $125.73 billion 31%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Minnesota $0.67 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

124

12th budget

solvency

32nd trust fund solvency

22nd long-run solvency

36th service-level

solvency

17th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

22nd

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Minnesota 2.32 2.99 3.01 1.05 $313 0.07 0.36 $2,458

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to- income ratio

OPEB-to-income ratio

Minnesota 0.08 0.14 0.13 0.44 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Minnesota ranks 17th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Minne- sota ranks 12th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Minnesota ranks 32nd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

125

MISSISSIPPI

SUMMARY

On the basis of its solvency in five separate categories, Mississippi ranks 36th among the US states for fiscal health. Mississippi has between 2.14 and 2.78 times the cash needed to cover short-term obligations. Revenues exceed expenses by 6 percent, with an improving net position of $323 per capita. In the long run, a net asset ratio of –0.04 indicates that Mississippi does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 37 percent of total assets, or $3,036 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $76.63 billion, or 71 percent of state personal income. OPEB are $0.71 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Mississippi $4.39 billion $5.70 billion $107.40 billion 5.3% $1,906

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Mississippi $17.16 billion 60% $76.63 billion 25%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Mississippi $0.71 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

126

8th budget

solvency

47th trust fund solvency

29th long-run solvency

44th service-level

solvency

19th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Mississippi 2.14 2.56 2.78 1.06 $323 –0.04 0.37 $3,036

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Mississippi 0.06 0.17 0.16 0.71 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Mississippi ranks 19th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Missis- sippi ranks 8th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Mississippi ranks 47th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

127

MISSOURI

SUMMARY

On the basis of its solvency in five separate categories, Missouri ranks 15th among the US states for fiscal health. Missouri has between 1.97 and 3.72 times the cash needed to cover short-term obligations. Revenues exceed expenses by 3 percent, with an improving net position of $108 per capita. In the long run, a net asset ratio of –0.01 indicates that Missouri does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 26 percent of total assets, or $1,809 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $114.25 billion, or 43 percent of state personal income. OPEB are $3.18 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Missouri $0.21 billion $3.55 billion $266.41 billion 1.3% $582

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Missouri $13.28 billion 81% $114.25 billion 33%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Missouri $3.18 billion 5%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

128

26th budget

solvency

33rd trust fund solvency

15th long-run solvency

8th service-level

solvency

14th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Missouri 1.97 3.68 3.72 1.03 $108 –0.01 0.26 $1,809

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Missouri 0.04 0.10 0.10 0.43 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Missouri ranks 14th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Missouri ranks 26th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Missouri ranks 33rd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

129

MONTANA

SUMMARY

On the basis of its solvency in five separate categories, Montana ranks 16th among the US states for fiscal health. Montana has between 3.98 and 5.26 times the cash needed to cover short-term obligations, well above the US average. Revenues exceed expenses by 5 percent, with an improving net position of $262 per capita. In the long run, Montana has a net asset ratio of 0.22. Long-term liabilities are lower than the national average, at 20 per- cent of total assets, or $2,247 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $23.23 billion, or 53 percent of state per- sonal income. OPEB are $0.46 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Montana $0.12 billion $0.22 billion $44.19 billion 0.5% $212

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Montana $3.62 billion 74% $23.23 billion 31%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Montana $0.46 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

130

14th budget

solvency

40th trust fund solvency

13th long-run solvency

34th service-level

solvency

5th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Montana 3.98 4.82 5.26 1.05 $262 0.22 0.20 $2,247

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Montana 0.05 0.14 0.13 0.53 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Montana ranks 5th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Montana ranks 14th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities, OPEB liabilities, and state debt compared to the state personal income? (Montana ranks 40th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

131

NEBRASKA

SUMMARY

On the basis of its solvency in five separate categories, Nebraska ranks 1st among the US states for fiscal health. Nebraska has between 2.95 and 3.95 times the cash needed to cover short-term obligations, above the US aver- age. Revenues only cover 99 percent of expenses, and its net position is unchanged from the previous year. In the long run, Nebraska has a net asset ratio of 0.28. Long-term liabilities are lower than the national average, at 4 percent of total assets, or $282 per capita. Total unfunded pension liabili- ties that are guaranteed to be paid are $20.90 billion, or 22 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Nebraska $0.00 $0.03 billion $94.66 billion 0.0% $18

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Nebraska $1.17 billion 91% $20.90 billion 37%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Nebraska n/a* n/a*

National average $14.51 billion 14%

* Nebraska does not report an OPEB liability.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

132

37th budget

solvency

4th trust fund solvency

1st long-run solvency

7th service-level

solvency

12th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Nebraska 2.95 3.86 3.95 0.99 $0 0.28 0.04 $282

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Nebraska 0.05 0.09 0.09 0.22 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Nebraska ranks 12th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Nebraska ranks 37th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Nebraska ranks 4th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

133

NEVADA

SUMMARY

On the basis of its solvency in five separate categories, Nevada ranks 10th among the US states for fiscal health. Nevada has between 1.46 and 2.69 times the cash needed to cover short-term obligations. Revenues exceed expenses by 16 percent, with an improving net position of $521 per capita. In the long run, a net asset ratio of 0.03 indicates that Nevada does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 37 percent of total assets, or $1,697 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $83.92 billion, or 65 percent of state personal income. OPEB are $1.45 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Nevada $1.36 billion $3.19 billion $128.29 billion 2.5% $1,084

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Nevada $12.56 billion 74% $83.92 billion 30%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Nevada $1.45 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

134

1st budget

solvency

45th trust fund solvency

16th long-run solvency

1st service-level

solvency

24th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Nevada 1.46 2.65 2.69 1.16 $521 0.03 0.37 $1,697

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Nevada 0.04 0.09 0.08 0.65 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Nevada ranks 24th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Nevada ranks 1st.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Nevada ranks 45th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

135

NEW HAMPSHIRE

SUMMARY

On the basis of its solvency in five separate categories, New Hampshire ranks 12th among the US states for fiscal health. New Hampshire has between 0.75 and 2.82 times the cash needed to cover short-term obligations. Revenues exceed expenses by 4 percent, with an improving net position of $413 per capita. In the long run, a net asset ratio of –0.02 indicates that New Hamp- shire does not have any assets remaining after debts have been paid. Long- term liabilities are lower than the national average, at 50 percent of total assets, or $2,555 per capita. Total unfunded pension liabilities are $20.85 billion, or 27 percent of state personal income. OPEB are $2.14 billion, or 3 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

New Hampshire $0.89 billion $1.49 billion $77.85 billion 1.9% $1,113

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

New Hampshire $5.13 billion 60% $20.85 billion 27%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

New Hampshire $2.14 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

136

10th budget

solvency

8th trust fund solvency

30th long-run solvency

3rd service-level

solvency

36th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

New Hampshire

0.75 1.46 2.82 1.04 $413 –0.02 0.50 $2,555

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

New Hampshire

0.03 0.09 0.09 0.27 0.03

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (New Hampshire ranks 36th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (New Hampshire ranks 10th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (New Hampshire ranks 8th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

137

NEW JERSEY

SUMMARY

On the basis of its solvency in five separate categories, New Jersey ranks 48th among the US states for fiscal health. New Jersey has between 0.93 and 2.44 times the cash needed to cover short-term obligations. Revenues only cover 89 percent of expenses, with a worsening net position of –$798 per capita. In the long run, New Jersey’s negative net asset ratio of 2.98 points to the use of debt and large unfunded obligations. Long-term liabilities are higher than the national average, at 388 percent of total assets, or $18,928 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $272.54 billion, or 49 percent of state personal income. OPEB are $85.42 billion, or 15 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

New Jersey $1.99 billion $42.73 billion $554.27 billion 7.7% $4,777

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

New Jersey $66.22 billion 57% $272.54 billion 24%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

New Jersey $85.42 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

138

49th budget

solvency

38th trust fund solvency

50th long-run solvency

20th service-level

solvency

30th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

New Jersey

0.93 2.44 2.44 0.89 –$798 –2.98 3.88 $18,928

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

New Jersey

0.05 0.11 0.12 0.49 0.15

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (New Jersey ranks 30th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (New Jersey ranks 49th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (New Jersey ranks 38th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

139

NEW MEXICO

SUMMARY

On the basis of its solvency in five separate categories, New Mexico ranks 45th among the US states for fiscal health. New Mexico has between 2.01 and 2.60 times the cash needed to cover short-term obligations. Revenues only cover 96 percent of expenses, with a worsening net position of –$490 per capita. In the long run, New Mexico has a net asset ratio of 0.5. Long-term liabilities are lower than the national average, at 23 percent of total assets, or $3,977 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $64.64 billion, or 80 percent of state personal income. OPEB are $3.81 billion, or 5 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

New Mexico $0.33 billion $3.50 billion $80.76 billion 4.3% $1,681

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

New Mexico $11.49 billion 70% $64.64 billion 29%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

New Mexico $3.81 billion 11%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

140

43rd budget

solvency

49th trust fund solvency

14th long-run solvency

50th service-level

solvency

21st cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

New Mexico

2.01 2.53 2.60 0.96 –$490 0.50 0.23 $3,977

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

New Mexico

0.07 0.23 0.24 0.80 0.05

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (New Mexico ranks 21st.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (New Mexico ranks 43rd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (New Mexico ranks 49th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

141

NEW YORK

SUMMARY

On the basis of its solvency in five separate categories, New York ranks 41st among the US states for fiscal health. New York has between 0.71 and 1.52 times the cash needed to cover short-term obligations, well below the US average. Revenues match expenses, with an improving net position of $16 per capita. In the long run, New York’s negative net asset ratio of 0.24 points to the use of debt and unfunded obligations. Long-term liabilities are 58 percent of total assets, lower than the national average. In per capita terms, long-term liabilities are larger than the national average at $4,605. Total unfunded pension liabilities that are guaranteed to be paid are $422.44 bil- lion, or 35 percent of state personal income. OPEB are $88.50 billion, or 7 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

New York $2.89 billion $56.69 billion $1,195.26 billion 4.7% $2,871

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

New York $14.65 billion 95% $422.44 billion 41%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

New York $88.50 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

142

35th budget

solvency 23rd

trust fund solvency

39th long-run solvency

38th service-level

solvency

44th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

New York 0.71 1.51 1.52 1.00 $16 –0.24 0.58 $4,605

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

New York 0.06 0.14 0.14 0.35 0.07

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (New York ranks 44th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (New York ranks 35th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities, OPEB liabilities, and state debt compared to the state personal income? (New York ranks 23rd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

143

NORTH CAROLINA

SUMMARY

On the basis of its solvency in five separate categories, North Carolina ranks 9th among the US states for fiscal health. North Carolina has between 1.67 and 2.72 times the cash needed to cover short-term obligations. Revenues exceed expenses by 12 percent, with an improving net position of $530 per capita. In the long run, North Carolina has a net asset ratio of 0.08. Long- term liabilities are lower than the national average, at 14 percent of total assets, or $938 per capita. Total unfunded pension liabilities that are guar- anteed to be paid are $131.56 billion, or 31 percent of state personal income. OPEB are $32.47 billion, or 8 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

North Carolina $3.04 billion $7.81 billion $426.19 billion 1.8% $770

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

North Carolina $8.57 billion 92% $131.56 billion 41%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

North Carolina $32.47 billion 4%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

144

2nd budget

solvency

14th trust fund solvency

8th long-run solvency

16th service-level

solvency

23rd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

North Carolina

1.67 2.55 2.72 1.12 $530 0.08 0.14 $938

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

North Carolina

0.06 0.11 0.10 0.31 0.08

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (North Carolina ranks 23rd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (North Carolina ranks 2nd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (North Carolina ranks 14th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

145

NORTH DAKOTA

SUMMARY

On the basis of its solvency in five separate categories, North Dakota ranks 19th among the US states for fiscal health. North Dakota has between 3.23 and 4.63 times the cash needed to cover short-term obligations, well above the US average. Revenues only cover 98 percent of expenses, with a worsen- ing net position of –$137 per capita. In the long run, North Dakota has a net asset ratio of 0.53. Long-term liabilities are lower than the national average, at 10 percent of total assets, or $3,509 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $12.68 billion, or 30 percent of state personal income. OPEB are $0.09 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

North Dakota $0.00 $1.89 billion $41.72 billion 4.5% $2,499

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

North Dakota $2.50 billion 65% $12.68 billion 27%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

North Dakota $0.09 billion 53%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

146

38th budget

solvency

12th trust fund solvency

9th long-run solvency

49th service-level

solvency

8th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

North Dakota

3.23 4.59 4.63 0.98 –$137 0.53 0.10 $3,509

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

North Dakota

0.08 0.19 0.19 0.30 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (North Dakota ranks 8th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (North Dakota ranks 38th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (North Dakota ranks 12th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

147

OHIO rank

SUMMARY

On the basis of its solvency in five separate categories, Ohio ranks 23rd among the US states for fiscal health. Ohio has between 3.43 and 4.20 times the cash needed to cover short-term obligations, well above the US average. Revenues match expenses, with an improving net position of $63 per capita. In the long run, Ohio has a net asset ratio of 0.07. Long-term liabilities are lower than the national average, at 51 percent of total assets, or $3,243 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $388.98 billion, or 75 percent of state personal income. OPEB are $15.14 bil- lion, or 3 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Ohio $9.28 billion $17.69 billion $521.21 billion 3.4% $1,523

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Ohio $62.60 billion 74% $388.98 billion 31%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Ohio $15.14 billion 52%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

148

33rd budget

solvency

48th trust fund solvency

32nd long-run solvency

25th service-level

solvency

9th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Ohio 3.43 4.05 4.20 1.00 $63 0.07 0.51 $3,243

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Ohio 0.05 0.12 0.12 0.75 0.03

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Ohio ranks 9th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Ohio ranks 33rd.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Ohio ranks 48th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

149

OKLAHOMA

SUMMARY

On the basis of its solvency in five separate categories, Oklahoma ranks 5th among the US states for fiscal health. Oklahoma has between 2.06 and 2.67 times the cash needed to cover short-term obligations. Revenues only cover 96 percent of expenses, with a worsening net position of –$171 per capita. In the long run, Oklahoma has a net asset ratio of 0.31. Long-term liabilities are lower than the national average, at 11 percent of total assets, or $609 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $63.16 billion, or 35 percent of state personal income. OPEB are $0.01 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Oklahoma $0.08 billion $2.14 billion $179.24 billion 1.2% $546

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Oklahoma $9.57 billion 75% $63.16 billion 32%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Oklahoma $0.01 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

150

41st budget

solvency

1st trust fund solvency

3rd long-run solvency

11th service-level

solvency

20th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Oklahoma 2.06 2.55 2.67 0.96 –$171 0.31 0.11 $609

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Oklahoma 0.05 0.10 0.11 0.35 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Oklahoma ranks 20th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Okla- homa ranks 41st.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Oklahoma ranks 1st.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

151

OREGON

SUMMARY

On the basis of its solvency in five separate categories, Oregon ranks 31st among the US states for fiscal health. Oregon has between 2.70 and 3.42 times the cash needed to cover short-term obligations. Revenues exceed expenses by 1 percent, with a worsening net position of –$33 per capita. In the long run, Oregon has a net asset ratio of 0.17. Long-term liabilities are lower than the national average, at 41 percent of total assets, or $3,283 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $120.50 billion, or 65 percent of state personal income. OPEB are $0.12 bil- lion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Oregon $5.53 billion $11.08 billion $184.41 billion 6.0% $2,708

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Oregon $16.52 billion 78% $120.50 billion 33%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Oregon $0.12 billion 80%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

152

42nd trust fund solvency

25th long-run solvency

40th service-level

solvency

13th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

34th budget

solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Oregon 2.70 3.25 3.42 1.01 –$33 0.17 0.41 $3,283

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Oregon 0.06 0.15 0.15 0.65 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Oregon ranks 13th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Oregon ranks 34th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Oregon ranks 42nd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

153

PENNSYLVANIA

SUMMARY

On the basis of its solvency in five separate categories, Pennsylvania ranks 35th among the US states for fiscal health. Pennsylvania has between 0.69 and 1.39 times the cash needed to cover short-term obligations, well below the US average. Revenues exceed expenses by 1 percent, with an improving net position of $62 per capita. In the long run, Pennsylvania’s negative net asset ratio of 0.27 points to the use of debt and large unfunded obligations. Long-term liabilities are lower than the national average, at 61 percent of total assets, or $3,109 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $245.40 billion, or 37 percent of state personal income. OPEB are $20.72 billion, or 3 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Pennsylvania $12.52 billion $16.59 billion $655.51 billion 2.5% $1,298

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Pennsylvania $62.64 billion 58% $245.40 billion 26%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Pennsylvania $20.72 billion 1%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

154

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

31st budget

solvency 24th

trust fund solvency

37th long-run solvency

23rd service-level

solvency

47th cash

solvency

UNDERLYING RATIOS

Cash ratio Quick ratio

Current ratio

Operating ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Pennsylvania 0.69 1.08 1.39 1.01 $62 –0.27 0.61 $3,109

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to- income ratio

OPEB-to-income ratio

Pennsylvania 0.05 0.12 0.12 0.37 0.03

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Pennsylvania ranks 47th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Pennsyl- vania ranks 31st.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Pennsylvania ranks 24th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

155

RHODE ISLAND

SUMMARY

On the basis of its solvency in five separate categories, Rhode Island ranks 40th among the US states for fiscal health. Rhode Island has between 1.13 and 2.02 times the cash needed to cover short-term obligations. Revenues exceed expenses by 3 percent, with an improving net position of $225 per capita. In the long run, Rhode Island’s negative net asset ratio of 0.49 points to the use of debt and unfunded obligations. Long-term liabilities are higher than the national average, at 90 percent of total assets, or $5,717 per capita. Total unfunded pen- sion liabilities that are guaranteed to be paid are $21.69 billion, or 40 percent of state personal income. OPEB are $0.64 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Rhode Island $1.05 billion $2.56 billion $54.49 billion 4.7% $2,420

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Rhode Island $4.94 billion 61% $21.69 billion 27%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Rhode Island $0.64 billion 18%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

156

21st budget

solvency

27th trust fund solvency

43rd long-run solvency

39th service-level

solvency

37th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Rhode Island

1.13 1.84 2.02 1.03 $225 –0.49 0.90 $5,717

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Rhode Island

0.06 0.15 0.14 0.40 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Rhode Island ranks 37th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Rhode Island ranks 21st.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Rhode Island ranks 27th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

157

SOUTH CAROLINA

SUMMARY

On the basis of its solvency in five separate categories, South Carolina ranks 20th among the US states for fiscal health. South Carolina has between 1.90 and 2.70 times the cash needed to cover short-term obligations. Revenues exceed expenses by 7 percent, with an improving net position of $373 per capita. In the long run, South Carolina has a net asset ratio of 0.17. Long-term liabilities are lower than the national average, at 23 percent of total assets, or $1,311 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $90.02 billion, or 46 percent of state personal income. OPEB are $10.48 billion, or 5 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

South Carolina $0.96 billion $2.86 billion $195.79 billion 1.5% $576

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

South Carolina $20.98 billion 60% $90.02 billion 26%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

South Carolina $10.48 billion 9%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

158

5th budget

solvency

35th trust fund solvency

11th long-run solvency

22nd service-level

solvency

22nd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

South Carolina

1.90 2.48 2.70 1.07 $373 0.17 0.23 $1,311

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

South Carolina

0.05 0.12 0.11 0.46 0.05

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (South Carolina ranks 22nd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (South Carolina ranks 5th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (South Carolina ranks 35th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

159

SOUTH DAKOTA

SUMMARY

On the basis of its solvency in five separate categories, South Dakota ranks 2nd among the US states for fiscal health. South Dakota has between 4.76 and 6.78 times the cash needed to cover short-term obligations, well above the US average. Revenues exceed expenses by 2 percent, with an improving net position of $106 per capita. In the long run, South Dakota has a net asset ratio of 0.34. Long-term liabilities are lower than the national average, at 8 percent of total assets, or $650 per capita. Total unfunded pension liabili- ties that are guaranteed to be paid are $13.32 billion, or 32 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

South Dakota $0.00 $0.52 billion $41.58 billion 1.3% $603

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

South Dakota $0.00 100% $13.32 billion 45%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

South Dakota n/a* n/a*

National average $14.51 billion 14%

* South Dakota does not report an OPEB liability.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

160

28th budget

solvency

13th trust fund solvency

2nd long-run solvency

6th service-level

solvency

3rd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

South Dakota

4.76 6.63 6.78 1.02 $106 0.34 0.08 $650

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

South Dakota

0.04 0.09 0.09 0.32 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (South Dakota ranks 3rd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (South Dakota ranks 28th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (South Dakota ranks 13th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

161

TENNESSEE

SUMMARY

On the basis of its solvency in five separate categories, Tennessee ranks 3rd among the US states for fiscal health. Tennessee has between 3.03 and 4.17 times the cash needed to cover short-term obligations. Revenues exceed expenses by 7 percent, with an improving net position of $290 per capita. In the long run, Tennessee has a net asset ratio of 0.14. Long-term liabilities are lower than the national average, at 10 percent of total assets, or $641 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $50.08 bil- lion, or 17 percent of state personal income. OPEB are $1.75 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Tennessee $2.12 billion $2.39 billion $288.53 billion 0.8% $359

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Tennessee $1.68 billion 95% $50.08 billion 41%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Tennessee $1.75 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

162

7th budget

solvency

3rd trust fund solvency

4th long-run solvency

12th service-level

solvency

10th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Tennessee 3.03 4.12 4.17 1.07 $290 0.14 0.10 $641

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Tennessee 0.05 0.11 0.10 0.17 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Tennessee ranks 10th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Tennes- see ranks 7th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Tennessee ranks 3rd.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

163

TEXAS

SUMMARY

On the basis of its solvency in five separate categories, Texas ranks 22nd among the US states for fiscal health. Texas has between 1.28 and 2.09 times the cash needed to cover short-term obligations. Revenues exceed expenses by 3 percent, with an improving net position of $155 per capita. In the long run, Texas has a net asset ratio of 0.26. Long-term liabilities are lower than the national average, at 33 percent of total assets, or $3,474 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $431.40 bil- lion, or 33 percent of state personal income. OPEB are $87.37 billion, or 7 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Texas $15.06 billion $50.81 billion $1,327.26 billion 3.8% $1,823

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Texas $52.49 billion 81% $431.40 billion 34%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Texas $87.37 billion 1%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

164

24th budget

solvency

15th trust fund solvency

13th service-level

solvency

34th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

21st long-run solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Texas 1.28 1.76 2.09 1.03 $155 0.26 0.33 $3,474

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Texas 0.04 0.11 0.11 0.33 0.07

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Texas ranks 34th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Texas ranks 24th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Texas ranks 15th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

165

UTAH

SUMMARY

On the basis of its solvency in five separate categories, Utah ranks 8th among the US states for fiscal health. Utah has between 1.61 and 3.75 times the cash needed to cover short-term obligations. Revenues exceed expenses by 8 per- cent, with an improving net position of $291 per capita. In the long run, Utah has a net asset ratio of 0.26. Long-term liabilities are lower than the national average, at 15 percent of total assets, or $1,555 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $45.04 billion, or 36 percent of state personal income. OPEB are $0.18 billion, or less than 1 per- cent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Utah $2.59 billion $5.16 billion $124.32 billion 4.1% $1,689

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Utah $4.40 billion 86% $45.04 billion 37%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Utah $0.18 billion 54%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

166

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

4th budget

solvency

10th long-run solvency

9th service-level

solvency

15th cash

solvency

20th trust fund solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Utah 1.61 3.65 3.75 1.08 $291 0.26 0.15 $1,555

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Utah 0.06 0.10 0.09 0.36 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Utah ranks 15th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Utah ranks 4th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Utah ranks 20th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

167

VERMONT

SUMMARY

On the basis of its solvency in five separate categories, Vermont ranks 39th among the US states for fiscal health. Vermont has between 1.62 and 2.50 times the cash needed to cover short-term obligations. Revenues exceed expenses by 5 percent, with an improving net position of $412 per capita. In the long run, Vermont’s negative net asset ratio of 0.25 points to the use of debt and unfunded obligations. Long-term liabilities are higher than the national average, at 68 percent of total assets, or $5,154 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $10.67 billion, or 34 percent of state personal income. OPEB are $1.82 billion, or 6 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Vermont $0.67 billion $0.71 billion $31.43 billion 2.3% $1,135

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Vermont $1.97 billion 67% $10.67 billion 27%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Vermont $1.82 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

168

9th budget

solvency

18th trust fund solvency

41st long-run solvency

47th service-level

solvency

25th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Vermont 1.62 2.46 2.50 1.05 $412 –0.25 0.68 $5,154

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Vermont 0.10 0.19 0.18 0.34 0.06

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Vermont ranks 25th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Vermont ranks 9th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Vermont ranks 18th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

169

VIRGINIA

SUMMARY

On the basis of its solvency in five separate categories, Virginia ranks 13th among the US states for fiscal health. Virginia has between 1.55 and 2.31 times the cash needed to cover short-term obligations. Revenues exceed expenses by 2 percent, with an improving net position of $92 per capita. In the long run, Virginia has a net asset ratio of –0.06. Long-term liabilities are lower than the national average, at 33 percent of total assets, or $1,714 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $127.59 billion, or 28 percent of state personal income. OPEB are $5.43 bil- lion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Virginia $0.60 billion $6.63 billion $451.91 billion 1.5% $789

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Virginia $23.13 billion 75% $127.59 billion 35%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Virginia $5.43 billion 25%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

170

29th budget

solvency

10th trust fund solvency

18th long-run solvency

4th service-level

solvency

28th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Virginia 1.55 2.23 2.31 1.02 $92 –0.06 0.33 $1,714

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Virginia 0.05 0.09 0.09 0.28 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Virginia ranks 28th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Virginia ranks 29th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Virginia ranks 10th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

171

WASHINGTON

SUMMARY

On the basis of its solvency in five separate categories, Washington ranks 30th among the US states for fiscal health. Washington has between 1.33 and 2.48 times the cash needed to cover short-term obligations. Revenues exceed expenses by 4 percent, with an improving net position of $229 per capita. In the long run, a net asset ratio of 0.02 indicates that Washington does not have any assets remaining after debts have been paid. Long-term liabilities are higher than the national average, at 64 percent of total assets, or $8,169 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $134.26 billion, or 34 percent of state personal income. OPEB are $13.75 billion, or 4 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Washington $20.52 billion $25.89 billion $389.86 billion 6.6% $3,553

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Washington $13.92 billion 84% $134.26 billion 36%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Washington $13.75 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

172

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0

19th budget

solvency

19th trust fund solvency

36th long-run solvency

30th service-level

solvency

29th cash

solvency

UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Washington 1.33 2.05 2.48 1.04 $229 0.02 0.64 $8,169

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Washington 0.05 0.13 0.13 0.34 0.04

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Washington ranks 29th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Wash- ington ranks 19th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Washington ranks 19th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

173

WEST VIRGINIA

SUMMARY

On the basis of its solvency in five separate categories, West Virginia ranks 43rd among the US states for fiscal health. West Virginia has between 1.27 and 1.78 times the cash needed to cover short-term obligations. Revenues exceed expenses by 1 percent, with an improving net position of $89 per capita. In the long run, West Virginia has a net asset ratio of –0.12. Long-term liabilities are lower than the national average, at 43 percent of total assets, or $4,194 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $27.75 billion, or 41 percent of state personal income. OPEB are $3.06 billion, or 4 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

West Virginia $0.39 billion $2.03 billion $68.46 billion 3.0% $1,109

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

West Virginia $4.33 billion 76% $27.75 billion 33%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

West Virginia $3.06 billion 18%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

174

30th budget

solvency

29th trust fund solvency

34th long-run solvency

46th service-level

solvency

38th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

West Virginia

1.27 1.54 1.78 1.01 $89 –0.12 0.43 $4,194

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

West Virginia

0.07 0.18 0.18 0.41 0.04

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (West Virginia ranks 38th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (West Virginia ranks 30th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (West Virginia ranks 29th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

175

WISCONSIN

SUMMARY

On the basis of its solvency in five separate categories, Wisconsin ranks 26th among the US states for fiscal health. Wisconsin has between 0.89 and 1.76 times the cash needed to cover short-term obligations, well below the US average. Revenues exceed expenses by 4 percent, with an improving net position of $244 per capita. In the long run, a net asset ratio of 0 indicates that Wisconsin does not have any assets remaining after debts have been paid. Long-term liabilities are lower than the national average, at 33 per- cent of total assets, or $2,589 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $70.63 billion, or 26 percent of state per- sonal income. OPEB are $0.94 billion, or less than 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Wisconsin $6.05 billion $13.86 billion $273.19 billion 5.1% $2,398

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Wisconsin $0.02 billion 100% $70.63 billion 57%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Wisconsin $0.94 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

176

18th budget

solvency

6th trust fund solvency

24th long-run solvency

32nd service-level

solvency

39th cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Wisconsin 0.89 1.74 1.76 1.04 $244 0.00 0.33 $2,589

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Wisconsin 0.06 0.13 0.13 0.26 0.00

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Wisconsin ranks 39th.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Wiscon- sin ranks 18th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Wisconsin ranks 6th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

177

WYOMING

SUMMARY

On the basis of its solvency in five separate categories, Wyoming ranks 6th among the US states for fiscal health. Wyoming has between 7.20 and 7.81 times the cash needed to cover short-term obligations, well above the US average. Revenues only cover 93 percent of expenses, with a worsening net position of –$577 per capita. In the long run, Wyoming has a net asset ratio of 0.74. Long-term liabilities are lower than the national average, at 10 per- cent of total assets, or $3,989 per capita. Total unfunded pension liabilities that are guaranteed to be paid are $15.87 billion, or 49 percent of state per- sonal income. OPEB are $0.24 billion, or 1 percent of state personal income.

2016 TOTAL LONG-TERM OBLIGATIONS STATE DEBT

General obligation

bonds

Total primary government

debt State personal

income

Ratio of debt to state personal

income Total primary

debt per capita

Wyoming $0.00 $0.02 billion $32.33 billion 0.1% $41

National average $5.85 billion $12.65 billion $319.33 billion 3.7% $1,830

PENSION LIABILITY

Unfunded pension liability Funded ratio

Market value of unfunded liability

Market value of funded liability ratio

Wyoming $2.07 billion 79% $15.87 billion 33%

National average $23.43 billion 73% $135.50 billion 32%

OTHER POSTEMPLOYMENT BENEFITS (OPEB)

Total unfunded OPEB Funded ratio

Wyoming $0.24 billion 0%

National average $14.51 billion 14%

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

178

47th budget

solvency

37th trust fund solvency

7th long-run solvency

37th service-level

solvency

2nd cash

solvency

–3.0

–2.0

–1.0

US avg

1.0

2.0

3.0 UNDERLYING RATIOS

Cash ratio Quick ratio Current

ratio Operating

ratio

Surplus (or deficit) per capita

Net asset ratio

Long-term liability

ratio

Long-term liability

per capita

Wyoming 7.20 7.59 7.81 0.93 –$577 0.74 0.10 $3,989

National average

2.22 2.99 3.22 1.01 –$72 –0.17 0.63 $4,387

Tax-to-income ratio

Revenue-to- income ratio

Expenses-to- income ratio

Pension-to-income ratio

OPEB-to-income ratio

Wyoming 0.07 0.14 0.15 0.49 0.01

National average

0.06 0.13 0.13 0.43 0.04

KEY TERMS

• Cash solvency measures whether a state has enough cash to cover its short- term bills, which include accounts payable, vouchers, warrants, and short- term debt. (Wyoming ranks 2nd.)

• Budget solvency measures whether a state can cover its fiscal year spend- ing using current revenues. Did it run a shortfall during the year? (Wyoming ranks 47th.)

• Trust fund solvency measures how much debt a state has. How large are unfunded pension liabilities and OPEB liabilities compared to the state per- sonal income? (Wyoming ranks 37th.)

distance from US average

(in standard deviations)

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

179

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

ACKNOWLEDGMENTS

Alice Calder, Danielle Barden, Lourdes Bautista, Ross Brady, Joe Conway, Michael Leahy, Justin Leventhal, and Nick Zaiac provided research assistance for this paper.

ABOUT THE AUTHORS

Eileen Norcross, MA, is the vice president of policy research and a senior research fellow at the Mercatus Center at George Mason University. Her research includes state and local finances, public sector pensions, fiscal federalism, and institutional analyses of state and local economies.

Olivia Gonzalez is a research associate at the Mercatus Center at George Mason University. Her research focuses on public finance issues and economic development, specializing in state fiscal health and public sector pensions. She received her BS in economics from George Mason University, where she is cur- rently a PhD student.

M E R C AT U S C E N T E R AT G E O R G E M A S O N U N I V E R S I T Y

180

ABOUT THE MERCATUS CENTER AT GEORGE MASON UNIVERSITY

The Mercatus Center at George Mason University is the world’s premier university source for market-oriented ideas—bridging the gap between academic ideas and real-world problems.

A university-based research center, Mercatus advances knowledge about how markets work to improve people’s lives by training graduate students, con- ducting research, and applying economics to offer solutions to society’s most pressing problems.

Our mission is to generate knowledge and understanding of the institu- tions that affect the freedom to prosper and to find sustainable solutions that overcome the barriers preventing individuals from living free, prosperous, and peaceful lives.

Founded in 1980, the Mercatus Center is located on George Mason Univer- sity’s Arlington and Fairfax campuses.

Introduction
1. RANKING THE STATES

Cash Solvency Rankings
Budget Solvency Rankings
Long-Run Solvency Rankings
Service-Level Solvency Rankings
Trust Fund Solvency Rankings
Overall Ranking of the States

2. FISCAL CONDITION TRENDS

National Trends

Cash solvency.
Budget solvency.
Long-run solvency.
Service-level solvency.
Trust fund solvency.

Fiscal Implications of Heavy Reliance on Oil Tax Revenues
Fiscal Implications of Major Tax Reforms
States with Pension Problems
States with Consistently Strong Fiscal Performance
States with Consistently Weak Fiscal Performance

3. CONCLUSION
APPENDIX A. RANKING METHODOLOGY
APPENDIX B. DATA TABLES
APPENDIX C. STATE PROFILES

__MACOSX/Economics Resources/._State Fiscal Rankings in 2018 - NORCROSS File.pdf

Economics Resources/Glossary - GRUBER File.pdf

__MACOSX/Economics Resources/._Glossary - GRUBER File.pdf

Economics Resources/Education - GRUBER File.pdf

__MACOSX/Economics Resources/._Education - GRUBER File.pdf

Economics Resources/Taxes on Risk Taking and Wealth - GRUBER File.pdf

__MACOSX/Economics Resources/._Taxes on Risk Taking and Wealth - GRUBER File.pdf

Economics Resources/Social Security - GRUBER File.pdf

__MACOSX/Economics Resources/._Social Security - GRUBER File.pdf

Economics Resources/State and Local Government Expenditures - GRUBER File.pdf

__MACOSX/Economics Resources/._State and Local Government Expenditures - GRUBER File.pdf

Economics Resources/Political Economy of Fiscal Institutions - von Hagen File.pdf

__MACOSX/Economics Resources/._Political Economy of Fiscal Institutions - von Hagen File.pdf

Economics Resources/Equity Implications of Taxation - GRUBER File.pdf

__MACOSX/Economics Resources/._Equity Implications of Taxation - GRUBER File.pdf

Economics Resources/Earmarking - Jackson.pdf

Tax earmarking, party politics and gubernatorial veto: theory and evidence from US states Author(s): Jeremy Jackson Source: Public Choice, Vol. 155, No. 1/2 (April 2013), pp. 1-18 Published by: Springer Stable URL: https://www.jstor.org/stable/23355150 Accessed: 18-11-2018 17:18 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Public Choice

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 DOI 10.1007/s 11127-011 -9822-y

Tax earmarking, party politics and gubernatorial veto: theory and evidence from US states

Jeremy Jackson

Received: 29 April 2010 / Accepted: 29 June 2011 / Published online: 20 July 2011 © Springer Science+Business Media, LLC 2011

Abstract This paper provides a theory of earmarking based on the relative power of a leg islature and executive. The politically powerful use earmarking as a means of resolving uncertainty and insulating preferred policy from the reach of future government. Tax rev enue will be earmarked more often when political power is unified under one party or when a party has the legislative majority needed to overturn a gubernatorial veto. An empirical test of the theoretical predictions are conducted using a panel of data for US states. A state with a legislature controlled by a single party with a large enough majority to overturn a gu bernatorial veto will earmark 5% more of its tax revenue than other states and a state with a

unified government will earmark 6.5% more. Together these explain 18.5% of the observed decrease in the percentage of state tax revenues earmarked from 1954 to 1997.

Keywords Earmarking • Veto • Spatial autocorrelation • Separation of powers

JEL Classification D72 • D78 • H41 H71

1 Introduction

Earmarking is a term that has two distinct meanings in the economics and political science literatures. It is used by the popular media and by some political scientists in reference to pork-barrel spending: spending targeted to benefit some interest group or jurisdiction. This is different from the definition of earmarking in the public finance literature which is addressed in this paper. I borrow the definition of earmarking as used in Pérez and Snell (1995). "Earmarking means designating some or all of the collections from a specific tax for a specific expenditure, with the intention that the designation will continue into the fu

ture." Earmarked tax revenues bypass the normal procedure in which tax revenue is pooled in a general fund and then allocated among various government spending programs. Tax revenues that are earmarked are directed away from the general fund and are not subject to

J. Jackson (IS1)

North Dakota State University, Richard H. Barry Hall, 811 2nd Ave. N. Fargo, ND 58108-6050, USA e-mail: [email protected]

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

2 Public Choice (2013) 155:1-18

the continued review process inherent in general fund appropriations. In the public finance view, tax earmarking has dynamic implications. The accounts to which earmarked tax rev enues are dedicated outlive the current fiscal year; without a change in policy the earmarked

money will continue to be dedicated to its purpose into the future. Thus, tax earmarking determines the allocation of funds over relatively long time horizons.1

Tax earmarking is a prevalent practice among US state governments. In 1954 the average state earmarked 51 % of its tax revenue to specific uses. The percentage earmarked dropped

to 27.5% in 1979 and has remained relatively flat in the years following. Table 1 shows earmarking levels for the 50 states as reported in Fiscal Planning Services (2000).

The dominant use of earmarked funds by states is for transportation: the building of highways and bridges, their maintenance, and public transit. It is not surprising, then, that the largest category of tax that is earmarked is the fuel tax and highway user fees.2 There are few examples of earmarked taxes available at the national level. The lone example is the federal payroll tax which is earmarked to fund social security and Medicare. Most of what gets termed as earmarking at the national level falls into the political science definition of earmarking which refers to pork barrel politics.

There are several reasons often cited in support of tax earmarking and against it. Those who support earmarking make the case that it can enforce a cost-benefit principle when those who pay the earmarked tax and the benefactors of that tax are one and the same. An earmarked tax on a product or service that is highly complementary in consumption to the government service provided may be viewed as a user-charge of sorts.3 This alleviates some of the typical efficiency problems, such as freeriding, associated with the provision of public goods. However, user charges can be employed only when the users of a particular service can be identified and charged. The most prominent tax which meets this criterion are

gasoline taxes, which often are earmarked for road maintenance and construction. It is very difficult to make such a case for any other tax-expenditure bundle. Many states earmark a portion of their income tax revenues and even more earmark a portion of various sales taxes

and registration fees. These taxes get earmarked for items such as: debt service, local gov ernment, education, healthcare/welfare and a variety of miscellaneous items, see Pérez and Snell (1995). Clearly, no cost-benefit principle is at play with these earmarked tax dollars.

Another commonly cited reason in support of tax earmarking is that earmarking stabilizes

government finances, thus serving as a means of controlling debt. Tax earmarking creates a link between revenue and expenditure; while this link may bring about a rigidity in the budget it can serve to limit current expenditure to current revenue thus reducing the amount

of debt financing needed to fund projects.

Earmarking can also serve as a tool to gain public support for new taxes. Voters are assured, through the earmarks, that revenues generated by a new tax will not be diverted to some other expenditure seen as less worthy by the electorate. However, if constituents believe that an earmark will cause general funds to be diverted away from the project in question, this justification loses its appeal.

There are also arguments suggesting that earmarking is not a beneficial practice. The most widely stated is that earmarking reduces the flexibility, freedom and oversight of public

expenditures. Earmarking may limit the ability of the government to respond to changing

'Tax earmarking is also often referred to as tax hypothecation.

2For detailed tables listing the purpose and source of earmarked taxes I refer you to tables 19 and 5 respec tively in Fiscal Planning Services (2000).

3 For discussion of the user charge principle for earmarking, see Lee and Wagner (1991).

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18

Table 1 State earmarking of revenues

1954 1963 1979 1984 1988 1993

89 87

47 51 31 29 32 30

41 36 21 18 17 13

42 28 12 13 12 19

75 51 17 25 18 20

26 23 0 1 12 10

0 3 0 5 7 6

40 39 28 28 26 28

29 22 II 9 8 6

7 5 5 6 5

51 44 38 32 25 21

39 43 14 18 21 32

49 39 43 33 30 26

51 44 19 13 21 22

77 66 29 25 21 25

46 29 16 4

85 87 5 4 9 15

46 39 19 20 17 12

47 40 34 24 20 17

56 54 41 40 39

67 57 38 39 35 39

73 74 12 13 14 16

40 37 30 26 26

57 40 20 29 30 27

61 53 55 60 65 64

55 53 41 29 22 21

55 35 34 52 49 57

53 54 31 24 24 14

7 2 25 39 36 39

80 31 36 44 47 40

13 10 0 6 8

38 30 20 8 14 19

73 43 29 21 22 22

48 48 21 18 19 17

62 59 43 24 21

47 36 23 19 23 21

41 63 15 15 14 11

6 4 0 1 5 5

69 62 56 55 44 17

59 54 33 32 27 47

72 77 60 61 66 60

(continued on the next page)

^ Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

4 Public Choice (2013) 155:1-18

Table 1 (Continued)

State 1954 1963 1979 1984 1988 1993 1997

Texas 81 66 54 20 24 21 14

Utah 71 62 52 48 55 54

Vermont 42 39 23 23 12 13 15

Virginia 39 32 27 24 25 25 23

Washington 35 30 29 26 29 30 26

West Virginia 57 39 21 21 20 19 21

Wisconsin 63 61 12 12 9 8

Wyoming 61 64 54 69 17 47

Average 51.27 43.06 27.50 26.60 25.07 24.50 24.06

Sources:

1954, 1963: Earmarked Slate Taxes, Tax Foundation

1979: March 19, 1980 Memo, Montana Office of the Legislative Fiscal Analyst

1984, 1988, 1993: Earmarking State Taxes, National Conference of State Legislatures

1997: Dedicated Tax Revenues: A Fifty-State Report, Fiscal Planning Services, Inc

economic conditions. This criticism drives at the heart of earmarking as earmarking does not allow either the legislature or the governor to weigh the relative merits of state programs

relative to the revenue available at every budget cycle. For this reason earmarking is seen by its critics as an inefficient method of allocating funds to projects regardless of the benefits it

may entail. While "a significant amount of research has been devoted to the effects of earmarking on

public good provision or education (many earmarked taxes are targeted toward education), surprisingly little attention has been given to the reason why earmarking might occur in the

first place. Although there is theoretical literature that tries to explain earmarking, it too is small and incomplete. There is an even greater lack of empirical research into the causes of earmarking. This paper provides both a theory of earmarking and empirical support for the theory.

A recent trend in political economy has been to focus on the role of the separation of powers such as between a legislature and an executive. It is inappropriate to focus analysis on just legislative decision-making when the executive branch also plays a role in policy for

mation and implementation. This is the case made by Figueiredo et al. (2000). Recently the separation-of-powers approach was applied to the study of the budget process by Grossman and Helpman (2008).

In this paper I describe a theory of earmarking based on the separation-of-powers to policymaking. A legislature chooses a policy bill to send to the governor who can either sign the bill or veto it. The legislature may overturn the veto if the requirements to do so are met. Legislatures controlled by a party with unilateral power to overturn gubernatorial vetoes and unified (one party) governments will tend to implement earmarked policies more than governments with less concentrated party control. This prediction is tested empirically using a panel of data on US state tax earmarking behavior. The econometric findings show that the legislative ability of a party to overturn a gubernatorial veto causes an increase in a state's percentage of revenues earmarked by five percentage points and a unified government leads to an increase of 6.5 percentage points. This suggests that 18.5% of the decrease in the

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 5

percentage of state revenues earmarked from 1954 to 1997 can be explained by the reduction in concentrated political control.

The paper is organized as follows. Section 2 reviews the relevant literature while Sect. 3 presents the theory to be tested. Section 4 outlines the econometric strategy and data with the results presented in Sect. 5. Section 6 concludes.

2 Literature review

The seminal paper on earmarked taxation is Buchanan (1963). Buchanan defines earmark ing "as the practice of designating or dedicating specific revenues to the financing of spe cific public services." Buchanan uses a median voter approach to analyze how earmarking and general fund financing may have differing implications for public spending. Buchanan (1963) and the succeeding literature of Goetz (1968), Goetz and McKnew (1972), Brown ing (1975), and more recently Athanassakos (1990) present analysis of the implications of earmarking but have little to say regarding the decision to earmark itself.

The more recent theoretical contributions all take a game theoretic approach. Most re cently, Jackson (2011) formulates a legislative bargaining model in which all available rev enue is spent via earmarking to the neglect of a general fund. Jackson (2011 ) explicitly mod els earmarking as it precedes general fund appropriations decisions. A legislator proposing an earmark has the incentive and ability to compensate other legislators enough to secure a winning coalition for his or her earmarking proposal. This earmarking proposal compen sates the coalition members for the opportunities they forgo in general fund bargaining. That paper shows that when institutions allow earmarking to occur then, in the absence of any frictions, full earmarking will occur in equilibrium.

The idea that earmarking may present a possible solution to an agency problem is ex plored by both Dhillon and Perroni (2001) and Bös (2000). While both papers show that earmarking can act as a mechanism that mitigates principle-agent costs, neither model is particularly realistic in its treatment of earmarking. In particular, Dhillon and Perroni (2001 ) do not model the public choice process involved in public good provision and while Bös (2000) does address aspects of public choice, his model does not consider the legislative body (parliament) as strategic in itself.

Brett and Keen (2000) present a model that proposes a compelling rational for earmark ing. In their model, incumbent politicians are able to restrict the behavior of their successors

by earmarking funds for preferred expenditures such as environmental protection. This is done when incumbents believe their re-election prospects are sufficiently low. Brett and Keen (2000) also show that earmarking can be used to mitigate the negative reputation ef fects of implementing a new tax.

A number of papers have also examined earmarking empirically. Novarro (2004) tests the hypothesis of Brett and Keen (2000) using data on earmarking of revenues for envi ronmental policies by Democratic legislative majorities in US states who proceeded to lose control of the legislature in the next election. Novarro (2004) finds no evidence of the type of strategic behavior described by Brett and Keen (2000). In fact she finds no evidence that

Democrats earmark strategically at all. This non-finding is perhaps less of a critique of the Brett and Keen model but more due to a failure in the empirical strategy to adequately fit the assumptions of the model. Deran (1965) and Dye and McGuire (1992), along with a large literature, explore earmarking and its effect on education spending. Novarro (2002) and Evans and Zhang (2007) look at the effect earmarking lottery profits has on education spending. Landry and Price (2007) study the effect earmarking lottery profits to a public

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

6 Public Choice (2013) 155:1-18

good (education) has on lottery play. The majority consensus is that earmarking does not increase public good (education) expenditure. Instead it allows general fund dollars to be diverted to other uses. There is little, if any, literature that actually explores the determinants

of the decision to earmark empirically.

A separate but relevant literature on policy insulation has also recently appeared. This literature (Moe 1989, 1990, 1991; McCubbins et al. 1987, 1989) is predominately a non formal literature that relates political (electoral) uncertainty to bureaucratic constraints and

efficiency. The politically powerful implement a bureaucratic and organizational structure

that protects their favored policy from those who will hold office in the future. Given this

description I interpret tax earmarking as a form of policy insulation. This theory was then

formalized by Figueiredo (2002) in what he terms the "insulation game."4 This game pre dicts that those who are electorally weak will be most likely to insulate their policy in the

event that they are able to gain momentary control of the government. There are two major

drawbacks to this theory. First, it assumes that insulated policies will proceed forever. An

extension of the game that allows earlier legislation, including insulating mechanisms, to

be repealed is hinted at but not rigorously analyzed. Second, the game doesn't consider a rich institutional structure. Bureaucratic structure in practice is centered on the separation

of powers between branches, yet this model examines only one branch of government, con trolled by one of two parties probabilistically. Figueiredo (2003) then tests these predictions

empirically, finding evidence that electorally weak groups will insulate policies when they

do have the power to do so. These results are consistent with the conclusion of this paper.

The most recent contribution to the earmarking literature is Anesi (2006). Anesi develops

a two-period model in which the incumbent party in time period 1 may choose to earmark

some or all tax revenue to one of two public goods.5 If revenue is earmarked it constrains

policy choice, i.e., the bundle of public goods chosen by the party elected in time period two. The party with the stronger preference for g \ has incentive to earmark revenue when

it is the incumbent in an attempt to constrain the other party in the event that the election

is lost in period two. This is similar to the intuition gained from Brett and Keen (2000) and Jackson (2011). Anesi's main contribution is that he considers earmarking incentives with

endogenous elections. With endogenous elections there are parameter ranges in which the

incumbent party may not have an incentive to earmark. By earmarking revenue a party can

constrain its opponent in such a way that any pre-election advantage is lost. That is, the majority of voters may have preferences more in line with the incumbent (they, too prefer

higher levels of gi) so that when the incumbent earmarks to gi the policies that will be implemented by either party become more closely aligned. If the incumbent has a large electoral advantage initially the act of earmarking may serve only to reduce it. Thus, with

endogenous elections there isn't always an incentive to earmark in order to constrain the future office holder. Even further, if the incumbent party prefers a lower level of gi, it may

find it in its interest to earmark revenue to gi in an attempt to improve its chances of winning

the election in time period two. Such an incumbent may find it optimal to constrain itself to

a policy it doesn't prefer in order to increase the probability of reelection. While this theory

is compelling there are no direct empirical predictions to be tested.

4This game is a modification of the reciprocity game introduced by Calvert ( 1989).

5 In the paper there are two public goods, #1 and #2 but tax revenue may be earmarked only for public good si.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 7

3 Theory: earmarking and the budget process

Earmarking refers to the dedication of specific tax revenues to specific expenditures on an ongoing basis: tying a tax to an expenditure. Earmarking of this type comes about in two predominate ways at the US state level. First, a state's constitution can stipulate how cer tain tax revenues must be spent. Secondly, a statute may be passed and signed into law that earmarks specific tax dollars to a specific expenditure. Any earmarking accomplished by constitutional means is both difficult to establish and difficult to abolish. However, earmark

ing is much more easily accomplished by statute. State level tax and expenditure decisions are made jointly by the legislative and execu

tive branches of state government. Legislatures submit bills that establish laws, taxes, and specify expenditure plans. The governor can either sign a bill with the bill then proceeding to become law, or veto a bill (if the state's laws permit a veto). Following a veto of a bill the legislative assembly may have the ability (according to state rules) to overturn the governor's

veto and cause a vetoed bill to become law. Overriding a gubernatorial veto usually requires the legislature to garner a supermajority in favor of the override. How large a supermajority is required to override a gubernatorial veto varies from state to state.

This process and procedure has strong implications for the patterns of earmarking be havior that should be observed across US states. The legislature can send bills to the gov ernor under two formats. First, it can just spend money in standard appropriations bills that disperse dollars from the general budgetary fund. Alternatively, the legislature could simul taneously specify an expenditure and a tax revenue source to fund it. This is an earmark. Such an earmark requires that all (or a portion of) tax revenue from the specified source (such as a gasoline tax) go into a fund that can only be spent on a specific expenditure (such as transportation infrastructure). Using earmarking to finance public expenditures creates a dedicated account such that all tax revenues from the earmarked tax go into that account and

can be used only for expenditures for which the account is dedicated.

Earmarking, by definition, is a means of secure funding for a project on into the future.

When tax earmarks are created they are in place for the life of the tax or until another bill either abolishes the tax, the earmark, or the account to which the revenue is earmarked. If a

project is funded by a general fund appropriations bill then its funding in the future can be

stopped merely by not allocating any funds in future appropriations bills to that project. But,

if a project is funded through earmarks the project is more difficult to cancel. The project won't go away through inaction as is the case for a project financed through the general fund; new legislation must be passed to eliminate either the tax or the earmark. Earmarking is thus an effective legislative tool that can be used to insulate preferred policies from future

policy decisions by an unknown government.

Consider a legislature dominated by two parties; each party has a set of expenditures that it favors. For expositional ease call one party the fire party and the other the police party.6 The fire party believes that it is optimal to spend a lot of money on fire protection and little money on police protection while the police party believes society is best served

by spending a lot of money on police protection and little money on fire protection. Both believe that money should be spent on both fire and police services they just differ on the relative levels of provision.

If the fire party currently holds a majority in the legislature and the governor is also in the fire party, then clearly a lot of money will be spent on fire protection and little on police

6I borrow the example of earmarking taxes for police arid fire protection from Buchanan (1963).

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

8 Public Choice (2013) 155:1-18

protection in the current time period. If the fire party is confident that it will maintain control

of the legislature and the governorship on into the future then there is little incentive to fund

fire and police protection with tax earmarks. Tax earmarks are unlikely to alter the funding for fire and police protection in the future and will only create a rigid budget that can't adapt to future needs. If, however, the fire party perceives that there is a chance that it could

lose control of either the legislature or the governorship then the fire party may find it in its

interest to create tax earmarks to fund its preferred quantity of fire and police protection.

The fire party can establish tax earmarks if it has control over both the legislature and the

governorship, that is, when the government is unified or if the legislature is controlled by a large enough majority to overturn any veto of a governor who belongs to the police party. The fire party will be unable to establish tax earmarks favoring fire protection if it does not have control of a unified government, or if it has a legislative majority but it is not large enough to overturn a veto by a governor from the police party.

Because all governments face electoral uncertainty over future electoral outcomes, it makes sense that a government will tend to earmark policies whenever it can in order to avert this uncertainty. This is consistent with the results of Jackson (2011), who shows that in a frictionless world where policy is the result of legislative bargaining among self-interested legislators, all revenue will be earmarked. The contention of this paper diverges from Jack son (2011) most significantly in that differing party preferences and the separation of powers among government branches creates a friction that prevents earmarking from occurring all the time. Frictions are overcome whenever the government is unified or the legislature has the power to overturn the governor's veto, making the power separation irrelevant.

In summary, we should observe an increase in the use of earmarking as a means of financ ing public spending when (1) one party controls both the state legislature and governorship and (2) the legislature is controlled by a large enough majority to overturn the veto of a governor from the opposing party. The incentive to earmark is reduced whenever the party in control faces little electoral uncertainty.

4 Data and econometric model

Theory suggests that tax dollars will be earmarked to specific uses when a legislature has the power to overturn gubernatorial vetoes and when the government is united under one party's

control. The policy insulation literature suggests that parties will have greater incentive to earmark (insulate) when electoral uncertainty is high. I now test these theoretical predictions using an econometric model.

The dependent variable for the study is the percentage of tax revenues that are earmarked

for specific uses in US states. The data on state-level earmarking were obtained from Fiscal Planning Services (2000), which contains data for the years 1954, 1963, 1979, 1984, 1988, 1993 and 1997. The report makes use of data from a series of surveys given to state budget officials; the data and their sources can be seen in Table 1. It is important to note that the data

were constructed using a strict definition of earmarking consistent with the public finance perspective. The data refer only to those tax dollars that are designated to a specific purpose on a continuing basis.

To test the theoretical prediction empirically I create two variables: V O and same. The variable V O is a dummy equal to one if the state's house and senate are both controlled by a large enough majority of the same party to overturn a gubernatorial veto. The variable same is a dummy variable set to one when a state's governorship, house, and senate are controlled by the same party. If earmarking plays the insulative role that theory suggests then these two

variables should have positive and statistically significant coefficients.

â Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 9

I also test the prediction of the policy insulation literature that earmarking should occur more often when electoral uncertainty is high and less often when electoral uncertainty is low. To control for electoral uncertainty I create two variables, senper and houper, which give the percentage of seats held by the majority party in the state senate and house, respec tively. If a party currently holds a large share of the seats in the senate (house) then it should

expect to maintain control in the future: losing control would require losing a large number of seats. Regardless of which party is in the majority, the percentage of seats held by the majority party can serve as a proxy for electoral uncertainty. Both senper and houper can take a value ranging between 0.5 and 1 with low values being associated with higher levels of electoral uncertainty and higher values with lower levels of electoral uncertainty. The ory suggests that both of these variables should have negative coefficients and to the degree

that electoral uncertainty versus party control (as captured by V 0 and same) matters for earmarking decisions the coefficients should be statistically significant.

Data for the construction of V O, same, senper, and houper were obtained from The Book of the States.7 The Book of the States reports the number of legislators from each party in both the house and senate, along with the party affiliation of the governor for all US states. There also are tables which describe institutional differences across the states.

In particular there is a table that lists the governors with veto power and the size of major ity vote required for the legislature to overturn the governor's veto. Two states, Nebraska and Minnesota, provide some difficulty in constructing the variables. Nebraska has a uni cameral system with non-partisan elections, meaning that information is available on party

affiliations. Therefore, for Nebraska V O is set to zero for all time periods. Minnesota also held non-partisan elections until 1976, VO therefore is set to zero for Minnesota in both 1954 and 1963. It is less clear what to do with the variable same in these instances. I show

empirical results both for setting same equal to one and equal to zero for Nebraska and the relevant years for Minnesota. I must also address the same data points for senper and houper ; I set each of these to 0.5 demonstrating a high level of electoral uncertainty (from the party perspective) when elections are non-partisan.

Does the ability of a legislature to overturn a gubernatorial veto, a unified government, or both cause increased usage of earmarking at the state level? I estimate the effects of V O and same on earmarking while accounting for state level and year specific fixed effects. It is also widely known that state level policy variables tend to be subject to spatial autocorrelation.8 Therefore, the econometric specification also controls for spatial autocorrelation in the data

with both a spatial lag and a spatial error term. I confirm the need for spatial techniques by conducting both the Moran-I test and the Lagrange Multiplier spatial diagnostic tests introduced by Anselin et al. (1996).9 For a detailed discussion of spatial econometrics I refer the reader to Anselin (1988).10

A spatial lag term controls for the potential influence of the policy of neighbor states on a state's own policy. This influence can come from a number of potential avenues as sum marized in Revelli (2005). Yardstick competition describes mimicking policy behavior; a

7 The Book of the States is published by the Council of State Governments. It was previously published on a biannual basis but is now available annually.

8For examples and discussion see Besley and Case (1995), Case and Rosen (1993), Brueckner (2003), Brown and Rork (2005), and LeSage and Dominguez (2010).

9 In the interest of brevity I have omitted the spatial diagnostic tests but can provide these results along with the OLS results used to compute them upon request.

l0The spatial econometric methods I use requires a balanced panel. Therefore, for years in which earmarking data are unavailable, a linearly interpolated value is used.

^ Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

10 Public Choice (2013) 155:1-18

state may try to replicate the policy of others because it views those states' policies as being successful. Fiscal competition between states can also cause the earmarking policy of one state to affect another as mobile residents cross state borders in response to policy differ ences. More direct policy spillovers also are possible. For example, investment in highways and bridges in one state may prompt a neighboring state to invest complementarily; if both are funded through earmarked taxes, then one state's earmarking policy can directly affect

its neighbor's. All of these possible avenues of influence are implemented empirically in the

same way using a spatial lag. The spatial lag term is created by premultiplying the dependent variable vector, y<, by

a weighting matrix, Wt. The time specific weighting matrix, W,, assigns a weight of zero to state i and then averages the value of the independent variable for neighboring states. The specification of a weighting matrix is highly dependent on the definition of neighbor. The most commonly used weighting matrix is the contiguity matrix that treats those states

sharing a geographic border as neighbors and weights them equally." It is also possible that a state will consider other states to be their neighbors when they are similar in some demographic or fiscal dimension.12 While it is possible to estimate W, econometrically, doing so is quite difficult.13 Therefore the empirical analysis proceeds by imposing a variety of weight matrices on the estimation. I estimate results using the standard contiguity based

weight matrix (Wco„,) and weight matrices based on similarity in indebtedness (Wjelj, pill>

and WjebiPi).u The construction of weight matrices is addressed in Appendix. The estimation equation including the spatial lag is written in (1 ). The spatial lag param

eter is p, ft the coefficient for V O, ft is the coefficient for same, ft is the coefficient for senper, ft is the coefficient for houper, ß a vector of parameters for the control variables in vector x„, y, is the state fixed effect, y, the time effect, and e„ is the error term.

y it = P W,y, +ft V On + ft same,, + ß3senperit + ß^hourper,, + ß\„ + y,- + y, + (1)

The presence of a spatial lag creates an endogeneity problem as the dependent variable appears on both sides of the regression equation. Therefore the OLS estimators will be both biased and inconsistent.

In addition to a spatial lag it is possible that the error term, e„, is subject to spatial dependence. This spatial dependence takes the form specified in (2) where e, is a vector of all states' errors in year t.

€it = XW,€, + vit (2)

The spatial dependence of e„ comes about when there are omitted variables that are spatially

dependent. If the spatial dependence is present and a correction is not made then the OLS estimators will not be consistent.

The presence of both spatial lag and spatial error dependence in (1) and (2) results in both biased and inconsistent OLS estimates. These issues are accounted for by using well

' ' Using a contiguity matrix the weighting matrix will be the same regardless of the time period.

12Demographic and fiscal characteristics change over time; therefore a weighting matrix defined over these dimensions may change over time as well.

' For an example, see Brett et al. (2002).

I4I also ran regression analysis using spatial matrices based on similarities in population, debt outstanding, personal income, and road miles. The results from these regressions are consistent with those reported in this paper. They are omitted for brevity but available upon request.

<£} Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18

Table 2 Summary statistics

1954 1963 1979 1984 1988 1993 1997

Ear 0.5127 0.4458 0.2921 0.2756 0.2668 0.2525 0.2473

(0.2020) (0.1959) (0.1789) (0.1839) (0.1717) (0.1707) (0.1770)

VO 0.6458 0.4583 0.4375 0.3750 0.3542 0.2708 0.2292

(0.4833) (0.5035) (0.5013) (0.4892) (0.4833) (0.4491) (0.4247)

same 0.8125 0.6667 0.5417 0.5833 0.3750 0.3958 0.3750

(0.3944) (0.4764) (0.5035) (0.4982) (0.4892) (0.4942) (0.4892)

senper 0.7877 0.7330 0.7058 0.6882 0.6647 0.6495 0.6008

(0.1586) (0.1646) (0.1457) (0.1377) (0.1146) (0.1229) (0.0785)

houper 0.7607 0.7116 0.6769 0.6795 0.6571 0.6359 0.6106

(0.1658) (0.1620) (0.1404) (0.1301) (0.1159) (0.1045) (0.0979) *

nm 39.4625 59.5938 180.4292 159.7792 126.2458 78.4788 37.1449

(487.242) (506.464) (591.140) (633.795) (669.169) (506.873) (420.814)

gov 0.3958 0.6667 0.6458 0.6667 0.5000 0.5833 0.3125

(0.4942) (0.4764) (0.4833) (0.4764) (0.5053) (0.4982) (0.4684)

sen 0.3913 0.5870 0.7234 0.7021 0.6809 0.7234 0.4894

(0.4934) (0.4978) (0.4522) (0.4623) (0.4712) (0.4522) (0.5053)

hou 0.3913 0.5870 0.7021 0.7872 0.7021 0.7234 0.5319

(0.4934) (0.4978) (0.4623) (0.4137) (0.4623) (0.4522) (0.5044)

sengov 0.8542 0.7083 0.6250 0.6250 0.4583 0.5417 0.4375

(0.3567) (0.4965) (0.4892) (0.4892) (0.5035) (0.5035) (0.5013)

hougov 0.9167 0.8750 0.8958 0.9167 0.5208 0.5625 0.7917

(0.2793) (0.3342) (0.3087) (0.2793) (0.5049) (0.5013) (0.4104) senhou 0.9167 0.8750 0.8958 0.9167 0.7708 0.7083 0.7917

(0.2793) (0.3342) (0.3087) (0.2793) (0.4247) (0.4593) (0.4104) samedem 0.2917 0.4375 0.4375 0.5000 0.2500 0.3333 0.1250

(0.4593) (0.5013) (0.5013) (0.5053) (0.4376) (0.4764) (0.3342)

lot 0.0000 0.0000 0.2917 0.3542 0.5833 0.7500 0.7708

(0.0000) (0.0000) (0.4593) (0.4833) (0.4982) (0.4376) (0.4247) *

rm 70.6950 75.2043 81.3203 84.8574 80.2869 80.9512 81.7965

(43.711) (44.092) (47.567) (56.546) (49.979) (50.033) (50.396)

debt pop 0.0508 0.1060 0.4364 0.8346 1.2370 1.6307 1.8009

(0.0514) (0.0908) (0.3364) (0.6018) (0.8186) (1.1290) (1.3395)

debt pi 0.0303 0.0442 0.0500 0.0630 0.0757 0.0787 0.0724

(0.0276) (0.0324) (0.0370) (0.0430) (0.0465) (0.0488) (0.0470)

Numbers reported are averages with standard deviations given in parentheses. Summary statistics for same place values of I on missing data points for Nebraska and Minnesota due to non-partisan elections. 'Thousands

known maximum likelihood methods as presented in Anselin (1988) to estimate (1) and (2) under standard assumptions on vit.

Summary statistics for all variables can be found in Table 2 and a listing of the control variables and their sources are found in Table 3. The majority of the control variables are dummies that describe the political environment and institutions. There are dummies to

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

12 Public Choice (2013) 155:1-18

Table 3 Control variables

Variable Definition Source

nm Net migration US Census Bureau

gov Dummy = if governor is a democrat Book of states

sen Dummy = if senate democrat majority Book of states

hou Dummy = if house democrat majority Book of states

sengov Dummy = if Gov and Senate majority are same party Book of states

hougov Dummy = if Gov and House majority are same party Book of states

senhou Dummy = if Senate and House majority are same party Book of states

samedem Dummy = if Gov, House and Senate are Democratic Book of states

lot Dummy = if state has a lottery Coughlin et al. (2006) rm Road miles Highway Statistics (US DOT) debt Debt outstanding at end of fiscal year Book of states

pop Population US Census Bureau

Pi Personal income US Census Bureau

deb pop debt/pop

deb pi debt/pi

account for the party affiliations of the governor and majorities in the house and senate, and

dummies that describe when the legislative and executive branches of the government are

controlled by the same party. In addition to the political descriptors variables are entered for

net population migration of a state, the total road miles in a state, a dummy for the presence

of a state lottery, and a control for indebtedness. Observations for Alaska and Hawaii were omitted as the dataset predates their statehood. The statistical technique employed requires a balanced panel so missing earmarking data were replaced with a linearly interpolated value.

The inclusion of many of these control variables in the estimation merit further explana tion. Net migration is included as a regressor to control for the potential effects of mobility

as suggested by Tiebout (1956).15 Road miles are used as a control as a large proportion of earmarked taxes come from fuel and highway user taxes and a large proportion of revenues are earmarked to transportation. A larger transportation infrastructure could then be corre lated with a larger percentage of revenues being earmarked. Road miles is a proxy for the

size of the transportation network. Lottery adoption is included as a control variable because

lottery revenues tend to be earmarked. Lottery money is not included in the dependent vari

able of this study because it is not considered to be tax revenue, but lottery adoption may reveal a preference for earmarked styles of financing. There is a simple correlation between

average percentage of revenues earmarked and average state lottery adoption. This simple correlation is evident from observing Table 2. In 1954 and 1963 no state had a lottery. Then

from 1979 to 1997,11% of states in the sample introduced lotteries.16 The growth in lottery

adoption coincides with a decline in the average percentage of revenues earmarked over the sample time period; one explanation could be that states earmarked fewer taxes in response to the increase in the revenues generated by lotteries.

15The empirical models were also run using a lag on net migration to account for possible endogeneity. Using lagged net migration did not change any of the results reported in the following section in any significant way.

l6Alaska and Hawaii are not included in the sample.

â Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 13

The most interesting explanatory variables are those that control for the level of indebt edness. There are two controls included in the regressions separately: debt pop is debt out standing weighted by state population and debt pi is debt outstanding weighted by state personal income. Each of these provides a measure of a state's ability to service its debt (indebtedness). Because earmarking links expenditures to revenues it is commonly viewed as a tool that a state government can use to control debt. The regression results lend support to this view.

5 Results

Table 4 and 5 each present results form the maximum likelihood spatial regressions using two different measures of state indebtedness as a control; the debt to population ratio and the debt to personal income ratio, respectively. Table 4 reports regression results when the missing data for Nebraska and Minnesota were handled by setting same to one while Table 5 show the results from setting same to zero. While all control variables were included in the regressions, in the interest of brevity I have not reported the estimates for the variables that

displayed no statistical significance and were of little interest. The coefficients for V O are positive and statistically significant for all specifications at

the 5% level regardless of the treatment of missing data. The estimates for VO range from as low as 0.0470 to as high as 0.0539. The coefficient estimates for same are all positive but statistical significance depends upon the treatment of missing data. In Table 4 same is statistically significant at the 1% level across all specifications and ranges from a low of 0.1136 to a high of 0.1356, but in Table 5 same is not statistically significant in any specification ranging from a low of 0.0041 to a high of 0.0368. There is more evidence that majorities in the house and senate powerful enough to override a gubernatorial veto increase the use of tax earmarks, yet the evidence on unified governments is too large to completely ignore. Using an estimate of 0.05 for V O and 0.065 for same, the combined average changes in V O and same over the period from 1954 to 1997 explain 18.5% of the 52% decrease in average percentage of revenues earmarked in the same time period.17

The coefficient estimates for senper and houper are not statistically significant in any specification. Coefficients for senper are always negative as expected, yet houper is neg ative when indebtedness is measured by the debt to population ratio but positive when it is measured by the debt to personal income ratio. The effects from V O and same are more powerful than any effect stemming from electoral uncertainty.

The controls for indebtedness yield some interesting results. The coefficients for debt to population (debt pop) are positive and statistically significant at the 1% level across all specifications. The coefficients for debt to personal income (debt pi) are not statistically significant.18 These findings suggest that a government with a high level of indebtedness will tend to earmark more revenue to specific uses, supporting the view that earmarking is used by state's as a means to control debt spending.

Other than the measures of indebtedness there is little significance in the estimated co efficients on control variables. Sengov and samedem enter negatively in Table 4. Sengov is statistically significant for all specifications while there is less significance for samedem.

^Alternatively, the average decreases in VO and same explain five percentage points of the 26.5 percentage point decrease in average percentage of revenues earmarked at the state level.

l8It is worth noting that even though the coefficients for debt pi are never significant they are often nearly so, with t-statistics approaching 1.6.

â Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

14 Public Choice (2013) 155:1-18

Table 4 Spatial regression results: same set to one for missing data

wcrm, Wdebt pop ^debt pi

vo 0.0507** 0.0492** 0.0493** 0.0482** 0.0539** 0.0525**

(2.2291) (2.1391) (2.2349) (2.1335) (2.3901) (2.3146)

same 0.1356*** 0.1351*** 0.1236*** 0.1253*** 0.1174*** 0.1136***

(3.0663) (3.0284) (2.8550) (2.8336) (2.6916) (2.5854)

senper -0.0778 -0.0687 -0.0352 -0.0257 -0.0224 -0.0094

(-0.8812) (-0.7695) (-0.4010) (-0.2868) (-0.2568) (-0.1067)

houper -0.0320 0.0087 -0.0359 0.0318 -0.0485 0.0094

(-0.3173) (0.0868) (-0.3657) (0.3217) (-0.4963) (0.0974)

sengov -0.0783*** -0.0849*** -0.0706** -0.0837*** -0.0626** -0.0687**

(-2.8432) (-3.0650) (-2.5482) (-2.9820) (-2.2615) (-2.4735)

samedem -0.0897* -0.0836 -0.0842 -0.0776 -0.0955* -0.0901*

(-1.6517) (-1.5288) (-1.5476) (-1.3967) (-1.7617) (-1.6488)

debt pop 0.0344*** - 0.0366*** - 0.0311*** -

(2.6616) (-) (3.9795) (-) (3.3430) (-)

debt pi - 0.3846 - 0.3814 - 0.3145

(-) (1.6062) (-) (1.5906) (-) (1.4178)

P 0.2180 0.2390 0.0749 0.1060 0.1050 0.1380**

(0.7822) (0.7300) (1.0336) (1.3446) (1.5740) (2.0890)

X 0.0109 0.0104 -0.4160*** -0.3900*** -0.4010*** -0.4680***

(0.0350) (0.0283) (-2.9183) (-2.6360) (-2.9432) (-3.3658)

Moran-I 0.8682 0.8749 0.0984 0.1980 0.1762 0.0515

(0.3853) (0.3816) (0.9216) (0.8431) (0.8602) (0.9589)

LM 0.0306 0.0461 0.4089 0.2563 0.3195 0.4295

(0.8611) (0.8300) (0.5225) (0.6126) (0.5719) (0.5122)

Nobs 336 - - - - -

Nvars 71 - - - - -

R2 0.799 0.7952 0.8017 0.7933 0.8017 0.7987

R2 0.7478 0.743 0.7512 0.7407 0.7512 0.7474

Significance levels: l%***,5%**, 10%*

Coefficient t-statistics and test statistic marginal probabilities are in parentheses

These coefficients lose their significance with the treatment of same in Table 5. This indi

cates that the percentages of tax revenue earmarked tends to be smaller when the senate and

governorship are controlled by the same party and there is mild evidence that a government unified under democratic control will earmark less. However, the persistent insignificance of the variables gov, sen, and hou indicates that any party-specific effect is quite weak. The theory presented in this paper makes no prediction about which party will want to earmark more than the other. These results show that the desire to earmark is not strictly a party

preference but rather due to the fierce political battle to insulate preferred policies from the

other party. Coefficients on lot and rm are insignificant, indicating that lottery adoption and

the size of a state's transportation network have minimal if any effect on tax earmarking behavior.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 15

Table 5 Spatial regression results: same set to zero for missing data

waml Wdebt pap Wrfebt pi

vo 0.0505** 0.0488** 0.0480** 0.0470** 0.0523** 0.0506**

(2.1874) (2.0900) (2.1586) (2.0557) (2.3125) (2.2328)

same 0.0368 0.0303 0.0314 0.0211 0.0170 0.0041

(0.8100) (0.6610) (0.7106) (0.4638) (0.3793) (0.0920)

senper -0.1062 -0.0969 -0.0573 -0.0513 -0.0378 -0.0230

(-1.1966) (-1.0788) (-0.6500) (-0.5685) (-0.4343) (-0.2627)

houper -0.0142 0.0286 -0.0143 0.0583 -0.0224 0.0390

(-0.1386) (0.2796) (-0.1448) (0.5827) (-0.2293) (0.4095)

sengov -0.0362 -0.0403 -0.0308 -0.0391 -0.0171 -0.0191

(-1.2914) (-1.4259) (-1.0940) (-1.3623) (-0.6175) (-0.6854)

samedem -0.0069 0.0045 -0.0046 0.0122 -0.0147 -0.0025

(-0.1233) (0.0799) (-0.0827) (0.2172) (-0.2650) (-0.0458)

debtpopl 0.0348*** - 0.0370*** - 0.0300*** -

(2.6186) (-) (4.0219) (-) (3.2821) (-)

debtpil - 0.3852 - 0.3737 - 0.3142

(-) (1.5812) (-) (1.5340) (-) (1.4305)

P 0.2180 0.2390 0.0760 0.0950 0.1250* 0.1585**

(0.7211) (0.6420) (1.0423) (1.1682) (1.9129) (2.4688)

X 0.0109 0.0106 -0.4560*** -0.3900*** -0.4870*** -0.5660***

(0.0324) (0.0255) (-3.1561) (-2.6075) (-3.5138) (-4.0190)

Moran-I 0.8336 0.8529 0.1330 0.1777 0.0680 -0.0660

(0.4045) (0.3937) (0.8942) (0.8589) (0.9458) (0.9474) LM 0.0186 0.0359 0.3682 0.2768 0.4438 0.5868

(0.8914) (0.8498) (0.5440) (0.5988) (0.5053) (0.4437)

Nobs 336 - - - - -

Nvars 71 - - - - -

R2 0.7937 0.7899 0.7984 0.7884 0.8005 0.7985

R2 0.7412 0.7363 0.7471 0.7346 0.7496 0.7472

Significance levels: 10%* Coefficient t-statistics and test statistic marginal probabilities are in parentheses

There is also some significance in the estimates of the spatial parameters. It is hard to interpret the spatial error term (À) other than to say that it accounts for unobserved variables

that are spatially correlated. The A. coefficients rise to the level of significance for the weight

ing matrices based on debt to population and debt to personal income ratios. However, there

is no significance on any spatial coefficient with the contiguity weighting matrix. The spatial

lag term (p) is significant only when the weight matrix is based on debt to personal income.

The interpretation of this spatial parameter is clear. Not only does a state with relatively high indebtedness tend to earmark more, there is some evidence that states mimic the ear marking practices of other states with similar debt to personal income ratios. If a state finds

other states with debt to personal income ratios comparable to their own and those states earmark a large percentage of their revenue, then that state is going to earmark more taxes

Ö Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

16 Public Choice (2013) 155:1-18

itself as it replicates the policy of similar states. Interestingly, a state looks more closely at its own debt to population ratio when making earmarking policy decisions yet states mimic the earmarking policies of states who are more similar in terms of the ratio debt to personal income.

Moran-I and Lagrange multiplier tests for residual spatial autocorrelation were con ducted; these tests reject the null hypothesis of spatial autocorrelation for all spatial weight

matrices specified. Therefore the estimates in Tables 4 and 5 are free from the bias and inefficiencies that OLS estimates would suffer from spatial misspecification.

6 Conclusion

A large share of state tax revenue is earmarked for specific uses and few studies have pro vided compelling arguments as to why this occurs. The theoretical studies addressing the issue are sparse and the empirical literature even more so. In this paper I have presented a simple theory of tax earmarking that shows how revenue may be earmarked in an effort by a current government to overcome policy uncertainty it faces over the future government's

policy choices. The ability of the legislature to earmark is somewhat alleviated by guber natorial veto; however, the veto is meaningless if the legislature can get the votes it needs to overturn it. The theory predicts that the earmarking of tax revenue will occur more of ten when the legislature is controlled by a majority large enough to overturn a veto by the

governor or when the government is unified under one party's control. Few theories of earmarking have been tested empirically and the theory presented in

this paper is the first to receive positive empirical support. The theoretical implications are confirmed using spatial econometric techniques on a panel of data from the US states. A state with a legislature that is controlled by a single party with a large enough majority to overturn

a gubernatorial veto will earmark 5% more of its tax revenue than other states, whereas a state with a unified government will earmark 6.5% more. These estimates explain 18.5% of the decrease in average state percentage of tax revenue earmarked over the years 1954—1997.

Acknowledgements 1 would like to offer my thanks to Marcus Berliant, Randall Calvert, and Paul Roth stein for their direction and comments and to Craig Brett, Jeremy Groves, and Tom Skladzien for comments made on an earlier draft. I also thank session participants at the 2008 Public Choice Society Meetings. Com ments from two anonymous referees were especially helpful. All remaining errors are my own.

Appendix

A.l Construction of W,

This appendix details the construction of the weight matrices Wr based on similarity in a characteristic and not on contiguity. Let i and j index generic states and let the set / be the collection of all states. Let C, be a characteristic vector of state where each C,-, refers to the

value of the characteristic for state i in time period t. For notational simplicity I neglect the

time subscript throughout the rest of this section.

Let Ajj = d(Cj, Cj) where d represents the usual standard distance function. Next iden tify all states that are no further than 9 > 0 away from i and call this set N, (6 ). Formally,

Ni(6) = {j el s.t. j ± ». Aij < 9}

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Public Choice (2013) 155:1-18 17

Define 9(n) to be the theta that makes the cardinality of the set N, (fi) equal to n < |/|." That is 6(n) = {f) e K+ s.t. |W,-(0)| = «}. By construction, the set of the n closest neighbors to i is given by Ni(6(n)).

I now begin to define the elements of W. The /th row of W contains weights for all the neighbors of state i. If k £ Nj(9(n)) then k is not a neighbor of i and the A'th element of row i in matrix W will be zero, Wik — 0. Note that i £ Ni(9(n)) so that Wu = 0 for all i regardless of n. For those j e Ni(6(n)) the assigning of weights is more complicated.

Define y, as follows.

Y, = 2 * max A,, jeNi (W(/i))

Assign to each j e Nj(6(n)) a number y(/- > 0 according to the equation Yij — (Yi ~ A,y)2. The weight given to i's neighbor j, Wir which appears in the /th column of row i in matrix W, can now be written as follows

2-,keN(H(n)) Y'k

Constructing the weight matrix W in this manner results in the n closest neighbors to state i in terms of characteristic C getting positive weight in a manner such that the closest

neighbors get the greater weight. This construction also row standardizes the matrix so that the sum of the weights assigned to neighbors of any i sum to unity.

For the analysis presented in the paper I set n = 5. Increasing this number slightly has little to no effect on the results of the analysis but as n gets large computational difficulties build. The spatial econometric techniques I employ work best when the matrix W is sparse. It does not seem reasonable that any state would examine the policies of all the other states when making policy comparisons. It is much more likely that they may look at a few that have similar characteristics. Therefore, setting /? — 5 is reasonable.

References

Anesi, V. (2006). Earmarked taxation and political competition. Journal of Public Economics, 90, 679-701. Anselin, L. (1988). Spatial econometrics: methods and models (1st ed.). Norwell: Kluwer Academic. Anselin, L., Bera, A. K„ Florax, R„ & Yoon, M. J. (1996). Simple diagnostic tests for spatial dependence.

Regional Science and Urban Economics, 26, 77-104. Athanassakos, A. (1990). General fund financing versus earmarked taxes: an alternative model of budgetary

choice in a democracy. Public Choice, 66, 261-278. Besley, T., & Case, A. (1995). Incumbent behavior: vote-seeking, tax-setting, and yardstick competition.

American Economic Review, 85, 25—45.

Bös, D. (2000). Earmarked taxation: welfare versus political support. Journal of Public Economics, 75, 439 462.

Brett, C., & Keen, M. (2000). Political uncertainty and the earmarking of environmental taxes. Journal of Public Economics, 75, 315-340.

Brett, C., Slade, M., & Pinkse, J. (2002). Spatial price competition: a semi-parametric approach. Economet rica, 70, 1111-1153.

Brown, R. P., & Rork, J. C. (2005). Copycat gaming: a spatial analysis of state lottery structure. Regional Science and Urban Economics, 35, 795-807.

Browning, E. K. (1975). Collective choice and general fund financing. Journal of Political Economy, 83, 377-390.

l9The notation \A\ refers to the cardinality of the set A.

"Ö Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

18 Public Choice (2013) 155:1-18

Brueckner, J. K. (2003). Strategic interaction among government: an overview of empirical studies. Interna tional Regional Science Review, 26, 175-188.

Buchanan, J. M. (1963). The economics of earmarked taxes. Journal of Political Economy, 71, 457-469. Calvert, R. L. (1989). Reciprocity among self-interested actors: uncertainty, asymmetry and distribution. In

P. C. Ordeshook (Ed.), Models of strategic choice in politics. Ann Arbor: University of Michigan Press. Case, A. C„ & Rosen, H. S. (1993). Budget spillovers and fiscal policy interdependence. Journal of Public

Economics, 52, 285-307. Coughlin, C., Garret, T., & Hernandez-Murillo, R. (2006). The geography, economics, and politics of lottery

adoption. Federal Reserve Bank of St. Louis Review, 88, 165-180. Deran, E. ( 1965). Earmarking and expenditures: a survey and a new test. National Tax Journal, 18,354-361. Dhillon, A., & Perroni, C. (2001). Tax earmarking and grass-roots accountability. Economics Letters, 72,

99-106.

Dye, R. F., & McGuire, T. J. (1992). The effect of earmarked revenues on the level and composition of expenditures. Public Finance Review, 20, 543-556.

Evans, W., & Zhang, P. (2007). The impact of earmarked lottery revenue on K-12 educational expenditures. Education Finance Review, 2, 40-73.

Figueiredo, R. J. P. D. (2002). Electoral competition, political uncertainty, and policy insulation. American Political Science Review, 96, 321-333.

Figueiredo, R. J. P. D. (2003). Budget institutions and political insulation: why states adopt the item veto. Journal of Public Economics, 87, 2611-2101.

Figueiredo, R. J. P. D., Jacobi, T., & Weingast, B. R. (2000). The new separation-of-powers approach to Amer ican politics. In B. R. Weingast & D. A. Wittman (Eds.), The Oxford handbook of political economy. London: Oxford University Press.

Fiscal Planning Services, Inc. (2000). Dedicated tax revenues: a fifty-state report. Betharda: Budget and Fiscal Research Services and Publications.

Goetz. C. J. (1968). Earmarked taxes and majority rule budgetary processes. American Economic Review, 58, 128-136.

Goetz, C. J„ & McKnew, C. R. (1972). Paradoxical results in a public choice model of alternative government grant forms. In J. M. Buchanan & R. D. Tollison (Eds.), Theory of public choice. Ann Arbor: University of Michigan Press.

Grossman, G. M., & Helpman, E. (2008). Separation of powers and the budget process. Journal of Public Economics, 92, 407-425.

Jackson, J. (2011). A legislative bargaining approach to earmarked public expenditures. Economics Bulletin, 31, 2006-2020.

Landry, C. E„ & Price, M. K. (2007). Earmarking lottery proceeds for public goods: empirical evidence from U.S. lotto expenditures. Economics Letters, 95, 451—455.

Lee, D. R„ & Wagner, R. (1991). The political economy of tax earmarking. In R. Wagner (Ed.), Charging for government: user charges and earmarked taxes in principle and practice (pp. 110-124). London: Routledge.

LeSage, J., & Dominguez, M. (2010). The importance of modeling spatial spillovers in public choice analysis. Public Choice. doi:10.1007/sl 1127-010-9714-6. ISSN:0048-5829.

McCubbins, M. D., Noll, R. G., & Weingast, B. R. (1987). Political procedures as instruments of political control. Journal of Law Economics and Organization, 3, 243-277.

McCubbins, M. D., Noll, R. G., & Weingast, B. R. (1989). Structure and process; politics and policy; admin istrative arrangements and the political control of agencies. Virginia Law Review, 75,431 —482.

Moe, T. M. (1989). The politics of bureaucratic structure. In J. E. Chubb & P. E. Peterson (Eds.), Can the government govern? Washington: Brookings.

Moe, T. M. (1990). The politics of structural choice: towards a theory of public bureaucracy. In O. E. Williamson (Ed.), Organization theory: from Chester Barnard to the present and beyond. Berkeley: University of California Press.

Moe, T. M. (1991). Politics and the theory of organization. Journal of Law Economics and Organization, 7, 106-129.

Novarro, N. (2002). Does earmarking matter? The case of state lottery profits and educational spending. Unpublished manuscript.

Novarro, N. (2004). Do policy-makers earmark to constrain their successors '! The case of environmental earmarking. Unpublished manuscript.

Pérez, A., & Snell, R. K. (1995). Earmarking state taxes (3rd ed.). Washington: National Conference of State Legislatures.

Revelli. F. (2005). On spatial public finance empirics. International Tax and Public Finance, 12, 475-492. Tiebout, C. M. (1956). A pure theory of local expenditures. Journal of Political Economy, 64, 416-424.

Ê Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:18:47 UTC All use subject to https://about.jstor.org/terms

Contents

p. [1]
p. 2
p. 3
p. 4
p. 5
p. 6
p. 7
p. 8
p. 9
p. 10
p. 11
p. 12
p. 13
p. 14
p. 15
p. 16
p. 17
p. 18

Issue Table of Contents

Public Choice, Vol. 155, No. 1/2 (April 2013) pp. 1-188

Front Matter
Tax earmarking, party politics and gubernatorial veto: theory and evidence from US states [pp. 1-18]
The composition and interests of Russia's business lobbies: testing Olson's hypothesis of the "encompassing organization" [pp. 19-41]
Inequality in developing economies: the role of institutional development [pp. 43-60]
Power indices in large voting bodies [pp. 61-79]
Independent central banks, regime type, and fiscal performance: the case of post-communist countries [pp. 81-107]
Satisfaction with democracy and collective action problems: the case of the environment [pp. 109-137]
Policy-seeking candidates who value the valence attributes of the winner [pp. 139-161]
þÿ�þ�ÿ��D��o�� p��r��o��-��m��a��r��k��e��t�� e��c��o��n��o��m��i��c�� r��e��f��o��r��m��s�� d��r��i��v��e�� h��u��m��a��n�� r��i��g��h��t��s�� v��i��o��l��a��t��i��o��n��s��?�� A��n�� e��m��p��i��r��i��c��a��l�� a��s��s��e��s��s��m��e��n��t��,�� 1��9��8��1��2��0��0��6�� [��p��p��.�� 1��6��3��-��1��8��7��]
Back Matter

__MACOSX/Economics Resources/._Earmarking - Jackson.pdf

Economics Resources/Taxation of Income - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Taxation of Income - HOLCOMBE File.pdf

Economics Resources/Authoritarian Government - Haber File.pdf

__MACOSX/Economics Resources/._Authoritarian Government - Haber File.pdf

Economics Resources/Taxes on Business Income and Wealth - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Taxes on Business Income and Wealth - HOLCOMBE File.pdf

Economics Resources/Taxes on Labor Supply - GRUBER File.pdf

__MACOSX/Economics Resources/._Taxes on Labor Supply - GRUBER File.pdf

Economics Resources/Public Goods - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Public Goods - HOLCOMBE File.pdf

Economics Resources/Health Economics and Private Health Insurance - GRUBER File.pdf

__MACOSX/Economics Resources/._Health Economics and Private Health Insurance - GRUBER File.pdf

Economics Resources/Medicare, Medicaid, and Health Care Reform - GRUBER File.pdf

__MACOSX/Economics Resources/._Medicare, Medicaid, and Health Care Reform - GRUBER File.pdf

Economics Resources/Budget Analysis and Deficit Financing - GRUBER File.pdf

__MACOSX/Economics Resources/._Budget Analysis and Deficit Financing - GRUBER File.pdf

Economics Resources/Taxes on Economic Transactions - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Taxes on Economic Transactions - HOLCOMBE File.pdf

Economics Resources/Economic Role of the State - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Economic Role of the State - HOLCOMBE File.pdf

Economics Resources/Taxation and Redistribution - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Taxation and Redistribution - HOLCOMBE File.pdf

Economics Resources/Budget Deficits - Hodler.pdf

Elections and the strategic use of budget deficits Author(s): Roland Hodler Source: Public Choice, Vol. 148, No. 1/2 (July 2011), pp. 149-161 Published by: Springer Stable URL: https://www.jstor.org/stable/41483686 Accessed: 18-11-2018 17:40 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Public Choice

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (2011) 148: 149-161 DOI 10.1007/sl 1 127-010-9650-5

Elections and the strategic use of budget deficits

Roland Hodler

Received: 23 July 2009 / Accepted: 20 April 2010 / Published online: 30 April 2010 © Springer Science+Business Media, LLC 2010

Abstract We present a model in which a conservative incumbent with preferences for low public spending can strategically run a budget deficit to prevent the left-wing opposition candidate from choosing high public spending if elected, and possibly also to ensure his own reelection. We find that the incumbent never manipulates the opposition candidate's public spending if he can ensure his own reelection; and that a conservative incumbent who runs a budget deficit to ensure his reelection may somewhat paradoxically choose high public spending before the election.

Keywords Voting • Budget deficits • Public debt • Distortionary policies

1 Introduction

It is well known that a conservative incumbent who prefers low public spending may strate- gically run a budget deficit to prevent the left-wing opposition candidate from raising public

spending after getting into office (Persson and Svensson 1989). It is however not obvious why an incumbent should target the opposition candidate if he can use the budget deficit to manipulate policy preferences of others. He may alternatively target decisive voters. A con- servative incumbent, for example, may run a budget deficit to ensure that these voters prefer his low future public spending to the opposition candidate's higher public spending. Getting reelected does after all not only allow him to choose low public spending in the future, but also ensures the perks from office for another term.

We present a stylized two-period model to study when and how a conservative incumbent runs a budget deficit to influence the election outcome or the future public spending of a left- wing opposition candidate. In this model, voters derive utility from private consumption

R. Hodler (El) Study Center Gerzensee, 3115 Gerzensee, Switzerland e-mail: [email protected]

R. Hodler

Department of Economics, University of Melbourne, Melbourne VIC 3110, Australia

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

150 Public Choice (201 1) 148: 149-161

and public spending as well as from the charisma or other exogenous characteristics of the

politician in office. The incumbent and the opposition candidate are office- as well as policy- motivated. The incumbent has weaker preferences for public spending than the median voter,

while the opposition candidate has stronger preferences. In the first period the incumbent chooses both the level of public spending and the budget deficit. In the second period the voters (re)elect one of the two politicians. The elected politician then chooses the level of public spending and repays public debt.

We find that the incumbent can be in one of three situations, depending on how inclined

the voters are to reelect him due to his charisma and other exogenous factors. First, if the

voters really want him in office, then he is in the best possible situation and gets reelected

after choosing his preferred policies: a balanced budget and low public spending. Second, if the voters strongly prefer the opposition candidate, the incumbent does not

get reelected no matter what policies he chooses. In this situation, he either accepts that his successor chooses high public spending and runs a budget surplus to smooth private consumption over time; or he runs a budget deficit to induce his successor to keep public spending low. He runs a budget deficit if and only if he is sufficiently averse to high public spending and the opposition candidate not too enthusiastic about it.

In the third scenario, the incumbent is not reelected when choosing a balanced budget. He is however reelected if he runs a sufficiently high budget deficit, because the median

voter then prefers his low future public spending to the opposition candidate's high public

spending. In this situation, the incumbent either runs a budget surplus to smooth private con-

sumption, or a distortionary budget deficit targeted at the median voter. He does not target

the opposition candidate since this would require an even more distortionary budget deficit

and may not ensure the perks from office for another term. Exactly because of these perks,

even a conservative incumbent who is not overly averse to public spending may run a bud-

get deficit to ensure his reelection. This gives rise to the somewhat paradoxical possibility that the incumbent chooses a distortionary budget deficit and high public spending before the election, and is then reelected because the voters anticipate that he will reduce public spending once the public debt that he accumulated must be settled. We will discuss later

how this result may explain, for example, why some Republican presidents increased the budget deficit before their reelection, and why they used the budget deficit not only to cut taxes, but also to raise public spending.

This paper builds on the seminal contribution of Persson and Svensson (1989). They study how a conservative incumbent can use the budget deficit to manipulate the public spending decision of a left-wing government that will be elected into office. However, they abstract from the possibility that the voters' choice could depend on the anticipated public spending decisions of future conservative and left-wing governments and, therefore, on the

level of public debt. We extend their model by allowing the election outcome to be endoge- nous to the budget deficit chosen by the incumbent government.

There are other contributions extending the Persson-Svensson model to allow for en- dogenous election outcomes. Aghion and Bolton (1990) show that a conservative incumbent

may run a budget deficit to improve his reelection prospects if and only if the median voter

is afraid that the opposition candidate would default on the outstanding debt; a possibility from which we abstract. Persson and Tabellini (2000: Chap. 13) present a simplified version of the Persson-Svensson model as well as an extension in which a conservative incumbent

can improve his reelection prospects by running a budget deficit if there are fewer swing vot-

ers among his supporters than among the opposition candidate's supporters. We show that

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 148: 149-161 151

the incumbent can use a budget deficit to improve his reelection prospects in close elections even in the absence of any asymmetry in the voting population.1

Inspired by the Persson-Svensson model, Pettersson-Lidbom (2001) studies how bud- get deficits depend on the reelection prospects of incumbent governments. Using panel data from Swedish municipal elections, he finds that conservative governments run higher budget deficits when facing lower reelection chances (and the opposite effect for left-wing govern- ments). This finding is consistent with our prediction that conservative incumbents choose higher budget deficits when targeting their successors because of dim reelection prospects than when targeting the median voter before close elections. Our paper further relates to contributions suggesting how distortionary policies may im-

prove an incumbent's reelection prospects. Milesi-Ferretti and Spolaore (1994) show that an incumbent who likes to channel funds to his constituency may refrain from tax reforms to ensure that public spending decisions remain unimportant in elections. Besley and Coate (1998) present a citizen-candidate model in which the elected candidate may refrain from ef- ficient public investments if the subsequent change in the income distribution translates into policy revisions that he dislikes. Biais and Perotti (2002) show that a conservative incumbent may underprice shares when privatizing public firms to reduce the appeal of redistributive policies to the middle class. Dellis (2009) presents a model in which the incumbent may not address policy issues on which the voters like his stance before the election. In addition, there is a literature showing that an incumbent may use distortionary policies in an attempt to mislead voters who are imperfectly informed about his competence (Rogoff and Sibert 1988), his preferences (Alesina and Cukierman 1990), or the underlying state of the world (Hodler etal. 2007). The remainder of the paper is structured as follows: Sect. 2 presents the model. Section 3

derives and discusses the equilibrium. Section 4 presents circumstantial evidence from the United States and concludes. All proofs are in the Appendix.

2 The model

We study a two-period model with complete information in which there is a conservative incumbent R , an opposition candidate L, and a measure-one continuum of voters i that differ in their preferences for public spending. The income of the two politicians and the voters is exogenous and, for simplicity, equal to 1 in each period t e {1, 2}. 2

In period one R chooses the budget deficit b e M and the level of public spending gi € {g, g], where g > g > 0. The budget deficit b is financed on international capital mar- kets at a zero interest rate, with b < 0 representing a budget surplus that is invested in these

1 In a model similar to the Persson-Svensson model, Alesina and Tabellini (1990a) show that an incum- bent may run a budget deficit if he and the opposition candidate differ in their preferences for differ- ent types of public goods (rather than different levels of public spending; see also Alesina and Tabellini 1990b). Crain (2001) summarizes the literature on policy durability, and discusses how the Persson-Svensson and the Alesina-Tabellini model relate to this literature. Most other contributions on the political econ- omy of budget deficits and public debt differ from our paper and the literature discussed above by as- suming that politicians have no policy preferences and care only about being in office (Lizzeri 1999), or the rents they can appropriate while in office (Battaglini and Coate 2008; Caballero and Yared 2008; Yared 2010).

2The assumption that all voters earn the same income, but differ in their preferences for public spending is not crucial. The model could easily be transformed into one in which voters differ in their incomes and, therefore, in the effects of public spending on private consumption, but not in the utility they derive from public spending itself. The assumption that income is constant over time can also be relaxed.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

152 Public Choice (201 1) 148: 149-161

markets. The dichotomy of g ¡ simplifies the analysis and helps us to illustrate under what circumstances R may somewhat paradoxically choose high public spending in period one. The policy bundle (gi , b) determines the lump-sum tax X' = g' - b in period one, since the budget deficit must be equal to public spending minus tax revenues. After observing (gj , b ), the voters (re)elect R or L at the beginning of period two. The elected politician then repays public debt b if b > 0, or collects public savings -b if b < 0, and chooses public spending 82 € {g, g}- The lump-sum tax T2 must be high enough to finance public spending and to repay public debt. Hence it is r2 = #2 + b. Following Persson and Svensson (1989: Sect. IV), we assume that private citizens can-

not access international capital markets. Assuming that the budget deficit is financed on international capital markets, while private citizens cannot access these markets, is a sim- ple way of guaranteeing that the budget deficit affects the path of private consumption ct > 0.3 Therefore, private consumption is c' = 1 - X' = 1 - gi + b in period one, and c*2 - 1 - ^2 - 1 - Si - b in period two. Together with the requirement that public debt must be repaid in period two and

82 £ {g> the non-negativity constraint c2 >0 puts an upper bound on the budget deficit. In particular, it must satisfy b < c = 1 - g to ensure 1 - g - b > 0. The same constraints also imply that only g2 = g is feasible in period two if b e (ç, c], where c = 1 - g. For later use notice that c and c are the hypothetical levels of private consumption in case of a balanced budget and public spending g and g , respectively. In each period r, the utility that politicians and voters derive from private consumption

and public spending is given by

gt ) = u(ct) + aih(gt ), (1)

where > 0 measures the intensity of their preferences for public spending, and where u(ct) and h(gt) are continuous with u! > 0, u" < 0 and h! > 0. We assume for technical reasons that u! -> 0 as c, -» oo, and u(g - g) - u( 0) > a¡ [h(g) - h(g)] for all a, .4 For sim- plicity, we abstract from discounting, and we introduce the tie-breaking rules that indifferent

politicians act as if they preferred g to g, and indifferent voters as if they preferred R to L.

The distribution of the voters' preferences for public spending a, is given by F(a¡). The median voter M is characterized by aM, where F(aM) = It is standard in political economy models with partisan politicians that the median voter's preferences are in-between the preferences of the two (major) parties and their candidates. We follow this tradition and assume

aR<aM<aL. (2)

Politicians and voters do not only derive utility from private consumption and public spending. Politicians also like to be in office, and they enjoy a rent ^ > 0 in every period in which they are in office.5 Voters also derive utility from exogenous characteristics of

3 Persson and Svensson (1989) show that the budget deficit has similar effects on the path of private consump- tion in such a simple model as in a richer model in which private citizens can access the international capital market but their labor supply is endogenous.

4These two assumptions, which are reminiscent of standard Inada conditions, guarantee the existence of the thresholds ßß and ßi introduced later (see Proposition 1 and its proof).

5 Results depend only on /?' s office rent and are independent of V s office rent. If R could not stand for reelection himself, e.g., due to a binding term limit, would represent his preferences for a successor from his own party rather than the opposition party (conditional on both politicians choosing the same #2)-

4l$ Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 148: 149-161 153

the politician in office, such as his charisma, integrity, competence and leadership skills.6

Their voting decisions therefore depend on the public spending gJ2 ( b ) chosen by politician

j e{R,L} if elected, and the consequences for private consumption cJ2(b) = 1 - g2(b) - b , as well as the politicians' exogenous characteristics 0j . In particular, voter i votes for R if and only if

cOi(c%(b),g£ (b)) + y&R > »/(Cj (b), g![ (b)) + y6L, (3)

where y > 0 measures the importance of these characteristics in the election. We subse- quently use A = y(0R -0L) and simply refer to it as R's relative charisma. More generally, we can view A as /?'s competitive edge in the election due to exogenous characteristics. Our model nests two special cases: The election outcome depends exclusively on public

finance issues if y = 0, and it is exogenous if y - ► oo. It is this second case that Persson and Svensson (1989) study. We finally impose some restrictions on oír and to ensure that R has strong preferences

for low public spending g , and L strong preferences for high public spending g. We want to ensure that R prefers g and L prefers g in any period with a balanced budget b = 0. We moreover want that R and L have strong enough preferences for g and g, respectively, such that neither of them would be interested in having g in one period and g in the other period

even if private consumption were perfectly smoothed over time, i.e., if ci = c2 = c =

Note that such consumption smoothing would require a budget deficit b = b= ^ > 0 if

gi =g and g2 = g, and b = -b < 0 if g¡ = g and g2 = g. We assume coR(c + g - g, £) > g) or, equivalently,

u(c + g - g) - u(c) 01 R <

h(g) - h(g / ) x

This assumption implies that R's preferences for public spending are so weak that he prefers g in period two for any b>-b. That is, he prefers g even if public savings are so high that private consumption c2 rises to c H- g - g > c, while it would again be = c if he chose g. Similarly, we assume

u(c)-u(c-g + g) aL >

h(g) -h(g) , , v

Hence, L chooses g in period two for any b < b. That is, he prefers g even if public debt b is so high that private consumption c2 drops to c - g + g < c, while it would again be c2 = c if he chose g.

3 Equilibrium

We solve for the Subgame Perfect Nash Equilibrium of our model using backward induc- tion. We therefore start by looking at period two, before deriving the policy choices of the conservative incumbent R in period one.

6Similarly, we could assume that voters care about the politicians' fixed positions in some policy dimension orthogonal to public spending, e.g., abortion.

<£) Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

154 Public Choice (201 1) 148: 149-161

3.1 Election outcome and public spending in period two

We first establish what level of public spending g2 the two politicians choose if (re)elected.

Besides his office rent politician j e {/?, L } gets utility u(c - b) +otjh(g) when choosing high public spending, and u(c - b) + ocjh(g) when choosing low public spending. The benefit of low public spending is therefore

U (b) = u(c - b) - u(c - b) = u( 1 - g - b) - u(' - g - b) > 0, (6)

and the benefit of high public spending is oljH , where

HzEh(g)-h(g)> 0. (7)

Hence politician j chooses g2(b) = g if and only if U(b) >o¿jH. While H is independent of public debt b , U(b) increases in b due to the concavity of u(ct). If b is high, private consumption c2 tends to be low in period two. A consumption difference of g - g > 0, which may make the difference between starving and getting by, therefore results in a large utility

difference U (b). However as b decreases, c2 tends to increase, and a consumption difference

of g - g translates into a smaller utility difference U(b). For this reason, U(b) > ajH is satisfied if and only if public debt b exceeds some threshold ßj. This is illustrated in Fig. 1.

Proposition 1 When in office in period two , R chooses low public spending g$ (b) = g if

public debt b > ßR, and high public spending g*(b) = g otherwise. L chooses g^ib) = g if

b>_ßh > and g2{b) = g otherwise. The unique thresholds ßR and ßi increase in aR and respectively , and satisfy ßR < -b and ßL>b.

Proposition 1 states that politician j e {/?, L] chooses low public spending g already at lower levels of public debt b , the weaker his preferences for public spending are. Figure 1

illustrates that ßj must be increasing in olj , as a rise in a j results in an upward shift of a jH, such that ctjH intersects U(b) at a higher b. This also explains why R already chooses g for lower b than L does, i.e., why ßR < ßL. The substantial difference between ßR and ßL , with ßR < -b and ßL > b , follows from assumptions (4) and (5). Before the election at the beginning of period two, the voters observe public debt b (as

well as g' and ti), and they correctly anticipate the levels of public spending g2 that the two

Fig. 1 Public debt and the players' thresholds in period two

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 148: 149-161 155

politicians choose if (re)elected. They anticipate that g£ (b) = g' (b) if b < ßR or b > ßL. In these cases, the voters base their decisions solely on the politicians' charisma. They all vote for R if A > 0, and for L otherwise.

The voters' decision problem is slightly more involved if the two politicians choose dif- ferent levels of public spending g2 since b e [ßR, ß^. Different voters may support different politicians, and the median voter M becomes decisive. She votes for R who then chooses low public spending g if and only if her benefit from his low public spending and his rela- tive charisma A exceed her benefit from V s high public spending. Hence she votes for R if and only if U(b) > oímH - A. Since she faces a similar decision problem in the election as the elected politician faces afterwards when choosing g2 , we can also illustrate her problem in Fig. 1 as well. There exists a threshold £a/(A) such that M votes for R if and only if b > /W(A). It is easy to see that this threshold must decrease in A, but increase in aM. Let us define Ä = (aM - aR)H > 0 and A == (aM - oil)H < 0. It follows that ßM( A) > ßR if and only if A < Ä, and that /?a/(A) < ßL if and only if A > A.

Proposition 2 R is reelected for any public debt b if A > Ä,/or b < ßR and b > £a/(A) if A 6 [0, à), for b e [/?a/(A), ßL) if A € (A, 0), but never if A < .A- L is elected otherwise. The threshold /3a/ (A) decreases in A but increases in a M>

Proposition 2 shows that the type of the election depends on R's relative charisma A. If A > Ä or A < A , the more charismatic politician has a competitive edge that is sufficiently large to win the election independently of the level of public debt. The election outcome however depends on public debt if the voters consider charisma to be unimportant or the two

politicians to be similarly (un)charismatic such that A e (A> Ä). In any case, the election outcome is independent of public spending gi in period one.

3.2 Budget deficit and public spending in period one

In period one the conservative incumbent R chooses public spending gi and the budget deficit b. He thereby takes into account how the budget deficit affects the election outcome (Proposition 2) and, if he is not reelected, his successor's public spending decision (Propo- sition 1).

We start by specifying R's preferred series of actions:

Lemma 1 R 's utility is maximized if he is reelected , and if g¡ = g2 = g and b = 0.

Hence, R chooses low public spending g and a balanced budget b = 0 in period one if this leads to his reelection. Proposition 2 implies that R is reelected after choosing b - 0 if A > Ä, and also if A € (A> À) and £a/(A) < 0 < ßL- Let us define A by /W(A) = 0, such that /W(A) < 0 if and only if A > Ä. It must hold that A € (A, A) since ßR < 0 < ßL. Consequently, a balanced budget leads to R's reelection if and only if A > A.

Proposition 3 If A > A, R chooses low public spending g' = g and a balanced budget

b = 0 in period one. The threshold A increases in a m and satisfies A e (A, A).

Proposition 3 implies that R can and does get his preferred series of actions (gi = g2 = b = 0 and his reelection) if he is relatively charismatic and also if the median voter has relatively weak preferences for public spending.

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

156 Public Choice (201 1) 148: 149-161

We next look at the case A < A, in which the conservative incumbent R does not get reelected regardless of the budget deficit b . In this case, R has to decide whether to accept that his successor L chooses high public spending g and to run a budget surplus b < 0 to smooth private consumption; or to strategically run a distortionary budget deficit b > ßL that induces L to choose low public spending g. Since R has relatively weak preferences for public spending, we suppose for the moment that he chooses gi = g (which we verify

below), and we focus on his choice of b. The budget deficit b = -b < 0, which corresponds to a budget surplus of ¿¡, smoothes private consumption if g' = g and g2 = g . The least dis- tortionary budget deficit satisfying b > ßL is b = ßL if gi = g2 = g. When choosing g' = g,

R therefore runs either the budget deficit b = to ensure g2 = g, or he accepts g2 = g and chooses b = -b.

Proposition 4 If A < A., R chooses low public spending g' = g and a budget deficit b = ßi

or b = -b in period one. He chooses b = ßL if oír and o¿l are relatively low , and b = -b otherwise.

Proposition 4 highlights that R runs the budget deficit b = ßL to manipulate U s policy choice g2 if V s preferences for public spending are moderate. The reason is that in this case L can be manipulated to choose g2 = g with a lower budget deficit and, therefore, less distortion in the intertemporal path of private consumption. R also chooses b = ßL if his own preferences for public spending are low. However if aR and are relatively high, R accepts that his successor L chooses g2 = g , and he runs a budget surplus of b to smooth private consumption over time. These results mirror Persson and Svensson's (1989) finding that only a stubborn conservative incumbent runs a budget deficit to manipulate his successor's policy, with stubbornness defined as putting more weight on reaching the preferred level of public spending than on the welfare costs of distortionary tax profiles.

Proposition 4 also confirms that R always chooses low public spending g in period one if he cannot be reelected. The main reason is, of course, that his preferences for public spend- ing are weak. He would nevertheless be indifferent between the policy bundles ( g , -¿),

which is followed by g^(b) = g , and (g, b) if followed by g^ib) = g. But V s preferences

for public spending are so strong that a more distortionary budget deficit b> ßL> b would be necessary to ensure g£ ( b ) = g . R therefore prefers (g , -b) in period one to any policy bundle that includes g.

Let us now discuss the intermediate case A e (A, A) in which the conservative incum- bent R cannot ensure his reelection with a balanced budget, but by running a budget deficit b € [ß/u( A), ßL ), where ßM{ A) e (ß/?, )0¿). In this case R also considers various policy bun- dles (gi , b). First, he again considers choosing (g, - b)> which leads to L being elected and choosing g2 = g, but smoothes private consumption over time. Second, he again consid- ers choosing gi = g and the least distortionary budget deficit that ensures g2 = g in period two. This deficit budget is now b = ßM( A), which ensures his reelection and thereby allows him to choose g himself in period two. It is less distortionary than b = ßL, and it moreover guarantees the office rent viz. The budget deficit b = ßL would lead to his reelection and V only if A > 0. Lastly, R may choose high public spending gi = g and the least distortionary budget deficit that ensures his reelection and thereby also g2 = g in period two. This budget

deficit is b = b* = ma x{b, ßM(A)}. We have seen above that R prefers (g, -b) to any policy bundle containing g, = g if he cannot get reelected anyway. However this may no longer hold. R may prefer (g, b*) to (g, - b) because the former guarantees his office rent while the latter does not.

â Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 148: 149-161 157

Proposition 5 If A e (A> A), R chooses one of the following three policy bundles (gi , b) in period one: (g, -b), (g, ßM(A))y or ( g , b*). He chooses (g, -b) ifty is low while aR and a m are high ; (g, ßM(A)) if*'> is high while aR and aM are low ; and (g, b*) ifV and aR are high while a m is low.

Proposition 5 confirms that if the conservative incumbent R runs a budget deficit to affect period two outcomes, he now targets the median voter M rather than the opposition candidate L. He only refrains from targeting M if his office rent ^ is relatively small, if he is

not too averse to high public spending, and if M has relatively strong preferences for public spending such that manipulating her decision would require a highly distortionary budget deficit. But unlike in the situation in which he could not get reelected, R might now run a distortionary budget deficit even if his preferences for public spending aR are relatively high. The reason is that by manipulating A/'s decision, he cannot only ensure low public spending g in period two, but also the office rent

The possibility that a conservative incumbent R who is not overly averse to high public spending may run a budget deficit to ensure his office rent explains why R may now choose high public spending g in period one. Manipulating AT s electoral decision requires a budget deficit b = ßiu( A), which may be substantial and therefore allow for high total (i.e., private and public) expenditures in period one. R may thus prefer high public spending g in period one if he is not overly averse to public spending, as the utility loss from a consumption de- crease of g - g matters little if consumption c' is high anyway. But even if R chooses g¡ = g , M still reelects him because she anticipates that he chooses lower public spending than L in period two, and because she prefers low public spending if excessive debt b = ßM (A) must be repaid.

It directly follows from Propositions 3 to 5:

Corollary 1 R runs a distortionary budget deficit b > 0 if and only if he wants to manipulate the election outcome or the public spending g2 chosen by his successor L. It holds:

1. If R can manipulate the election outcome , he never manipulates L 's public spending. 2. If R manipulates the election outcome , the budget deficit is lower than necessary to ma-

nipulate L 's public spending. 3. If R manipulates the election outcome , he may choose high public spending in period

one , which he never does when manipulating L 's public spending.

Corollary 1 highlights various implications of our model. Statement 1 underlines that the conservative incumbent R prefers to run a strategic budget deficit to manipulate the election outcome rather than to manipulate future policies of his opposition candidate L. One reason is that when being reelected, he can choose his preferred policy and also enjoy the office rent. Another reason, emphasized by statement 2, is that if ensuring his reelection is possible, it requires a lower and less distortionary budget deficit than manipulating U s future policies.

Statements 2 and 3 of Corollary 1 imply what pattern of fiscal policies we should observe depending on how close the upcoming elections are. If the candidate of the conservative incumbent's party loses independently of economic and public finance issues, possibly be- cause the candidate of the opposition party is much more charismatic, then the incumbent chooses low public spending before the election and possibly a distortionary budget deficit to ensure that public spending also remains low in the future. However when the election outcome depends on economic and public finance issues, the conservative incumbent runs a somewhat smaller and less distortionary budget deficit (if he runs a deficit at all). In ad- dition to a budget deficit, he may also choose high public spending before the election. As

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

158 Public Choice (201 1) 148: 149-161

discussed above, this seemingly paradoxical choice follows from the interaction of various forces: The perks from office may even motivate a conservative incumbent who is not overly

averse to high public spending to strategically run a distortionary budget deficit to ensure his reelection; and the necessary budget deficit may make plenty of funds available before the election, such that this incumbent may prefer to use these borrowed funds not only to finance private consumption by cutting taxes, but also to increase public spending.

4 Concluding remarks

In this paper we have shown how and when a conservative incumbent with preferences for low public spending strategically runs a distortionary budget deficit to ensure his reelec- tion or, if this is not possible, to manipulate the public spending decision of the left-wing opposition candidate who has stronger preferences for public spending. Results for the reverse case of a left-wing incumbent are symmetrical. By running a suffi-

ciently distortionary budget surplus , a left-wing incumbent can make sure that the conserv- ative opposition candidate chooses high public spending if elected. The incumbent however prefers to target the median voter instead to ensure his reelection if that is possible. This requires a less distortionary budget surplus and guarantees the perks from office for an- other term. Moreover, a left-wing incumbent may somewhat paradoxically choose low pub- lic spending before the election when running a budget surplus targeted at the median voter. The reason is that the necessary budget surplus would lead to very low private consumption if it were entirely financed by higher taxes. A left-wing incumbent who is not overly keen on a large public sector may therefore decide to finance the budget surplus partly by keeping

public spending low. Nevertheless he gets reelected because the voters anticipate that he will raise public spending after the election when public savings can be consumed. Figure 2 shows the development of the United States' government finances over the

last 35 years. Three presidents were reelected during this period: the Republican presidents Ronald Reagan and George W. Bush, and the Democratic president Bill Clinton. Looking

Fig. 2 US government finances in the recent past (Source: CBO 2009)

^ Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 148: 149-161 159

at the first terms of these presidents, we observe an interesting pattern. The budget deficit increased rapidly during the first terms of Ronald Reagan (1981-1985) and George W. Bush (2001-2005). These increases resulted mainly from lower tax revenues, but also from higher public spending (i.e., outlays). The reverse happened during Bill Clinton's first term (1993- 1997). The budget deficit decreased as a result of higher tax revenues and lower public spending. This pattern is not predicted by the Persson-Svensson model or any other model of the political economy of budget deficits and public debt. Our model however offers a pos- sible explanation for this pattern: The two Republican presidents may have raised the budget deficit to win middle class support in the election at the end of their first terms. These high budget deficits increased the available funds by so much that any conservative incumbent who is not overly averse to public spending would have preferred to use these funds not only to cut taxes, but also to raise public spending. Similarly, Bill Clinton may have used the budget surplus to ensure middle class support in the upcoming election and may have preferred to finance this surplus not exclusively by raising revenues but also by lowering public spending because he was not overly keen on high public spending. While our model offers an explanation for this interesting aspect of budget deficits and

public spending in the recent US history, it is clear that this evidence is circumstantial at best, and that forces not captured by our stylized model also impact upon government fi- nances. More theoretical and empirical work is certainly necessary to fully understand how office- and policy-motivated governments can and do use public spending, budget deficits and public debt to influence election outcomes and future policies. As a first step one could modify our setup to allow for an infinite number of periods, which would lead to a more suitable framework to study how term limits affect the strategic use of budget deficits. Also

looking at the interplay of local, regional and national governments could yield interesting insights.

Acknowledgements I would like to thank an associate editor, an anonymous referee, Rosemary Hum- berstone, Simon Loertscher, Randy Silvers, and seminar and conference participants at Deakin University, La Trobe University, the University of Tasmania, and the Econometric Society Australasian Meeting 2009 for helpful comments.

Appendix

Proof of Proposition 1 It holds that U'b(b) > 0 since u'c > 0, u"c < 0 and c > c; that U (b) - ► 0 as b -» -oo since uc - ► 0 as ct -> oo; and that U(c) > c¿jH since u(g - g) - u( 0) > gljH. Consequently, there exists a unique ßj satisfying U ( ßj ) = olj H, with ßj < c. It follows from U'(b) > 0 that U ( b ) > oljH if and only if b > ßj , and that ßj strictly in-

creases in a j. Hence, ßR < ßL. Assumptions (4) and (5) imply ßR < -b and ßL> b. □

Proof of Proposition 2 We define £a/(A) by U(ßM( A)) = olmH - A. It follows from U'(b) > 0 that U(b) > aMH - A if and only if b > ßM( A), and that /W(A) is strictly in- creasing in a m and strictly decreasing in A. It follows from the definitions of A and A that ßM( A) = ßL and ßM (Ä) = ßR . It remains to prove our results for the four regions of A men- tioned in the proposition. First, if A > Ä, R is reelected for b £ [ßRl ß^) because A > 0, and also for b e [ßR, ßL) because /W(A) < ßR- Second, if A 6 [0, Ä), R is reelected for b i [ßR, ßL ) because A > 0, and for b e (A), ßL ), but L is elected for b e [ßRi £a/(A)), where ßM( A) e (ßR,ßL)- Third, if A e (A, 0), L is elected for b £ [ßR,ßü because A < 0, and for b e [ßR, ßM (A)), but R is reelected for b e [^(A), ßL), where again

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

160 Public Choice (201 1) 148: 149-161

/?m(A) e (ßR, ßL ). Fourth, if A < A, L is elected for b £ [ßR, ßL) because A < 0, and also for b e[ßR,ßL) because ßM( A) > ßL. □

Proof of Lemma 1 If reelected, R gets *I> and is thus better off than otherwise. It follows

from the concavity of u(ct) that for any g' and g2 , /?' s utility is maximized if b ensures

c' = C2. It remains to show that u(c) + aRh(g) > u(c) -f aRh(g) and 2 u(c) + 2 aRh(g) > 2 u(c) + a R[h(g) + h (g)]. It directly follows from the concavity of u(ct) and assumption (4)

that the first of these inequalities holds. To see why the second inequality holds, observe that

assumption (4) implies u{c + g - g) + u(c) + 2a Rh(g) > 2 u(c) + a R[h(g) + h(g)], and that the concavity of u(ct) and 2c = 2c + g - g imply 2 u(c) > u(c + g - g) + u(c) . □

Proof of Proposition 3 It follows from the definition of A that A must increase in aM since ßM( A) increases in aM but decreases in A, and that A e (A> Ä) since £m(Â) < 0 < ßhii A). Lemma 1 implies that R chooses g' = 0 and b = 0 if this leads to his reelection. Proposition 2 and the definition of A imply that R is reelected when b = 0 if and only if A > A. Hence, R chooses g' = 0 and b = 0 if A > A. □

Proof of Proposition 4 It follows from the concavity of u{ct ), the definition of b and -b <

0 < ßi that R strictly prefers b = - b to any other b < ßL if g' = g and g2=-g, and b =

ßL to any other b > ßL if gi = g 2 =£. Similarly, R prefers b = 0 to any other b < ßL if gl = g2 = and ßL to any other b > ßL if gi = g and g2 = g since ßL > b. We next

show that R strictly prefers (g',b) = (g, -b) to both (g,0) and Since 2 u(c) > u(c) + m(c), and since assumption (4) implies u(c) + aRh(g) > u(c) + aRh(g ), it must

hold that 2 u(c) + aR[h(g) + h(g)] > 2 u(c) + 2 aRh(g), i.e., that R strictly prefers (g, -b)

to (¿,0). Since u(ct) is concave and ßL > b , he also strictly prefers (g, -b) to ( g,ßL )•

Hence he chooses either (g, ßL) or (g, - b). He prefers (g, -b) if and only if

2u(c) + aR[h(g) + h(g)] > u(c - ßL) + u(c + + 2 aRh(g).

A higher aR increases the LHS of this inequality by more than the RHS since h(g) > h(g).

It therefore makes (g, -b) relatively more attractive. Proposition 1 implies that a higher aL raises ßL. Together with the concavity of u(ct ), this implies that a higher aL decreases the RHS and makes (g, -b) more attractive. □

Proof of Proposition 5 Similarly as in the proof of Proposition 4, we can show that R chooses b e ßM(A)} if g 1 = g, and b e {0, b*} if g¡ = g, and that R prefers (g, - b) to (g,0). This proves the first statement. For the second statement, observe that R ben- efits from a higher *1/ if and only if he chooses b > ßM( A), i.e., (g,ßM(A)) or (g,b*).

A higher aR leads to a higher increase in his utility when playing (g, - b) or (g,b*) than when playing ( gfßM(A )), as only the latter leads to g, = g2 = g. A higher aM, which raises ßM(A), lowers his intertemporal utility from private consumption, u(c') + «fe), when playing (g, ßM(A)) and possibly also when playing (g,/?*), but not when playing

(£, ~b). ~ □

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 148: 149-161 161

References

Aghion, P., & Bolton, P. (1990). Government domestic debt and the risk of default: a political-economic model of a strategic role of debt. In R. Dornbusch, & M. Draghi (Eds.), Public debt management : theory and history (pp. 315-345). Cambridge: Cambridge University Press.

Alesina, A., & Cukierman, A. (1990). The politics of ambiguity. Quarterly Journal of Economics, 105 , 829- 850.

Alesina, A., & Tabellini, G. (1990a). A positive theory of fiscal deficits and government debt. Review of Economic Studies , 57, 403-414.

Alesina, A., & Tabellini, G. (1990b). Voting on the budget deficit. American Economic Review , 80 , 37-49. Battaglini, M., & Coate, S. (2008). A dynamic theory of public spending, taxation, and debt. American Eco-

nomic Review , 98 , 201-236. Besley, T., & Coate, S. (1998). Sources of inefficiency in a representative democracy: a dynamic analysis.

American Economic Review , 88 , 139-156. Biais, B., & Perotti, E. (2002). Machiavellian privatization. American Economic Review , 92, 240-258. Caballero, R. J., & Yared, P. (2008). Future rent-seeking and current public saving. NBER Working Paper

14417.

CBO (2009). Historical budget data. Washington: Congressional Budget Office. Crain, W. M. (2001). Institutions, durability, and the value of political transactions. In W. E II Shughart, &

L. Razzolini (Eds.), The Elgar companion to public choice (pp. 183-196). Cheltenham: Edward Elgar. Dellis, A. (2009). The salient issue of issue salience. Journal of Public Economic Theory , 77, 203-231. Hodler, R., Loertscher, S., & Rohner, D. (2007). Inefficient policies and incumbency advantage. Cambridge

Working Paper in Economics 0738. Lizzeri, A. (1999). Budget deficits and redistributive politics. Review of Economic Studies , 66, 909-928. Milesi-Ferretti, G. M., & Spolaore, E. (1994). How cynical can an incumbent be? Strategic policy in a model

of government spending. Journal of Public Economics , 55, 121-140. Persson, T., & Svensson, L. (1989). Why a stubborn conservative would run a deficit: policy with time-

inconsistent preferences. Quarterly Journal of Economics, 104, 325-345. Persson, T., & Tabellini, G. (2000). Political economics: explaining economic policies. Cambridge: MIT

Press.

Pettersson-Lidbom, P. (2001). An empirical investigation of the strategic use of debt. Journal of Political Economy, 109 , 570-583.

Rogoff, K., & Sibert, A. (1988). Elections and macroeconomic policy cycles. Review of Economic Studies , 55, 1-16.

Yared, P. (2010). Politicans, taxes, and debt. Review of Economic Studies , 77, 806-840.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 17:40:39 UTC All use subject to https://about.jstor.org/terms

Contents

p. [149]
p. 150
p. 151
p. 152
p. 153
p. 154
p. 155
p. 156
p. 157
p. 158
p. 159
p. 160
p. 161

Issue Table of Contents

Public Choice, Vol. 148, No. 1/2 (July 2011) pp. A1-A4, 1-267

Front Matter
The proximity paradox: the legislative agenda and the electoral success of ideological extremists [pp. 1-19]
Election results and opportunistic policies: A new test of the rational political business cycle model [pp. 21-44]
A theory of entangled political economy, with application to TARP and NRA [pp. 45-66]
Condorcet Polling [pp. 67-86]
Securing the base: electoral competition under variable turnout [pp. 87-104]
Free riders, holdouts, and public use: a tale of two externalities [pp. 105-117]
Timely shirking: time-dependent monitoring and its effects on legislative behavior in the U.S. Senate [pp. 119-148]
Elections and the strategic use of budget deficits [pp. 149-161]
Hold your nose and vote: corruption and public decisions in a representative democracy [pp. 163-196]
One-dimensionality and stability in legislative voting [pp. 197-214]
Foreclosure in contests [pp. 215-232]
Opportunistic and partisan election cycles in Brazil: new evidence at the municipal level [pp. 233-247]
REVIEW ARTICLE

Expressive voting and identity: evidence from a case study of a group of U.S. voters [pp. 249-257]

BOOK REVIEW

Review: untitled [pp. 259-261]
Review: untitled [pp. 263-264]
Review: untitled [pp. 265-267]

Back Matter

__MACOSX/Economics Resources/._Budget Deficits - Hodler.pdf

Economics Resources/Fundamental Tax Reform - GRUBER File.pdf

__MACOSX/Economics Resources/._Fundamental Tax Reform - GRUBER File.pdf

Economics Resources/Principles of Tax Policy - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Principles of Tax Policy - HOLCOMBE File.pdf

Economics Resources/Externalities - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Externalities - HOLCOMBE File.pdf

Economics Resources/US STATISTICS ON THE FIFTY STATES File.pdf

__MACOSX/Economics Resources/._US STATISTICS ON THE FIFTY STATES File.pdf

Economics Resources/Unemployment Insurance - GRUBER.pdf

__MACOSX/Economics Resources/._Unemployment Insurance - GRUBER.pdf

Economics Resources/Corporate Taxation - GRUBER File.pdf

__MACOSX/Economics Resources/._Corporate Taxation - GRUBER File.pdf

Economics Resources/Public Sector Demand - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Public Sector Demand - HOLCOMBE File.pdf

Economics Resources/Government Redistribution Programs - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Government Redistribution Programs - HOLCOMBE File.pdf

Economics Resources/Property Rights and Economic Efficiency - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Property Rights and Economic Efficiency - HOLCOMBE File.pdf

Economics Resources/Federal System of Government - HOLCOMBE File.pdf

__MACOSX/Economics Resources/._Federal System of Government - HOLCOMBE File.pdf

Economics Resources/Local Public Goods - Joanis.pdf

The road to power: partisan loyalty and the centralized provision of local infrastructure Author(s): Marcelin Joanis Source: Public Choice, Vol. 146, No. 1/2 (January 2011), pp. 117-143 Published by: Springer Stable URL: https://www.jstor.org/stable/41483620 Accessed: 18-11-2018 18:59 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Public Choice

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 DOI 10.1007/sl 1127-009-9586-9

The road to power: partisan loyalty and the centralized provision of local infrastructure

Marcelin Joanis

Received: 20 January 2009 / Accepted: 16 December 2009 / Published online: 9 January 2010 © Springer Science+Business Media, LLC 2010

Abstract This paper sets out a simple dynamic probabilistic voting model in which a gov- ernment allocates a fixed budget across electoral districts that differ in their loyalties to the

ruling party. The model predicts that the geographic pattern of spending depends on the way

the government balances long-run 4 machine polities' considerations and the more immedi- ate concern to win over swing voters. Empirical results obtained from a panel of electoral districts in Québec provide robust evidence that districts which display loyalty to the incum-

bent government receive disproportionately more spending, especially close to an election, at odds with the standard 'swing voter' view.

Keywords Partisan loyalty • Swing voters • Political competition • Local public goods • Distributive politics • Long-run relationships

...the new mad turns from pavement into gravel ('Must've elected the wrong guy last time around David says. . .) - Margaret Atwood, Surfacing, 1972, p. 18

1 Introduction

Spectacular events involving aging public infrastructures, such as the Minneapolis bridge collapse in the summer of 2007, inevitably spark debates in the popular press about elec-

M. Joanis (Ë3) Department of Economics and GREDI, Faculté d'administration, Université de Sherbrooke, 2500, boul de l'Université, Sherbrooke, Quebec, Canada, J1K 2R1 e-mail: [email protected]

M. Joanis

CIRANO, Montréal, Quebec, Canada

ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

118 Public Choice (201 1) 146: 1 17-143

toral misallocation of infrastructure spending.1 This is not surprising since public infrastruc-

tures such as roads and bridges are durable and highly visible, two characteristics that are especially desirable from the point of view of politicians interested in securing the endur- ing support of their constituencies. The main goal of this paper is to examine whether the geographic allocation of infrastructure spending by higher tiers of government is indeed distorted by electoral politics. Most public infrastructures are best described as centrally-provided local public goods:

they generate localized benefits - in contrast to pure public goods - but are generally not provided by local governments. The political process is well known to be a fundamental component of the centralized provision of local public goods.2 The existing theoretical liter- ature on distributive politics (or special-interest politics), rooted in the Downsian modelling tradition, has focused largely on the incentive for politicians to target these goods to pivotal

voters, groups or regions.3 As shown by the considerable interest in 'swing states' during U.S. presidential campaigns, pivotal regions clearly attract a disproportionate share of po- litical attention, and the empirical evidence suggests that this is indeed accompanied by a disproportionate share of campaign resources.4 It seems natural to expect that pivotal regions should also attract a disproportionate share of government resources more generally. How- ever, evidence from the empirical literature on the geographic allocation of public spending is somewhat mixed in finding spending patterns that conform to such a 'swing voter' view.5 Despite its intuitive appeal, the swing voter view overlooks one of the most enduring

features of modern democratic societies, namely the fact that political parties engage in long-run relationships with their core supporters. For example, two-thirds of the U.S. popu- lation consider themselves to be either Democrat or Republican, and these partisan loyalties are known to evolve only slowly over time (see Green et al. 2002). Such stable electoral bases are crucial for major political parties to remain credible contenders in upcoming elec- tions. For that reason, parties typically devote ongoing attention to their core supporters, a tendency that has been referred to in the literature as 'machine politics.'6 Political parties thus face a trade-off in the allocation of political favors. Politicians have

an incentive to direct spending towards constituencies in which the marginal dollar spent is most likely to make a difference in terms of immediate electoral outcomes (e.g., in swing districts); however, the existence of long-term relationships between parties and the con- stituencies forming their electoral base provides an incentive for forward-looking incum- bents to favor them as well, so as to secure their support in the future.

thirteen people died on August 1, 2007, when a bridge of the Interstate 35W highway over the Mississippi River collapsed in Minneapolis, Minnesota (USA). On September 30, 2006, five motorists were killed in a similar tragedy in Laval, Québec (Canada), when a bridge over Highway 19 collapsed. Both events were followed by intense debates about the politicization of infrastructure spending. A similar debate followed the collapse of the levees protecting New Orleans when Hurricane Katrina hit in 2005.

2 See Knight (2004) for an excellent discussion.

^Echoing Downs's (1957) median voter theorem, a 'swing voter' view of pork-barrel politics has emerged as a standard prediction in formal models of distributive politics - see Lindbeck and Weibull (1987, 1993) for perhaps the most influential treatment.

4See, for example, Strömberg (2008) on campaign spending in the United States.

5 While Cadot et al. (2006), Milligan and Smart (2005), Dahlberg and Johansson (2002), Schady (2000), and Stein and Bickers (1994) report evidence of swing voter patterns, Francia and Levine (2006), Larcinese et al. (2006a, 2006b), Moser (2008) and Case (2001) do not find such evidence.

6See, for example, Dixit and Londregan (1996). Others, such as Larcinese et al. (2006b), refer to machine politics outcomes as 'partisan supporters' outcomes.

& Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 1 19

To formalize these conflicting incentives, this paper proposes a distributive politics model

with probabilistic voting - an approach pioneered by Lindbeck and Weibull (1987, 1993) - that accounts for the existence of long-run relationships between the incumbent government

and loyal electoral districts. In contrast to the static models typically used in the existing lit- erature, a two-period model captures the time dimension inherent to partisan loyalty. The model's key assumption is that electoral support in favor of the incumbent government exhibits some intertemporal persistence in loyal districts. In equilibrium, the allocation of spending by the government is affected by two conflicting forces: the need to sway the bal- ance in swing districts to win the election in the short-run - a 'political competition effect'- and the need to nurture long-run loyalty relationships to win in the future - a 'loyalty effect.'

Depending on which of these forces dominates, the model predicts that both 'swing district' and 'machine politics' equilibria can arise. The latter 'non-Downsian' equilibria material- ize in the model when future electoral support receives sufficient weight in the incumbent government's decisions. The empirical relevance of both swing district and machine politics equilibria is assessed

by exploiting a rich data set on road expenditure by the provincial government in Québec, the Canadian province with the largest land mass. These data are disaggregated at the electoral district level and cover a ten-year period in the 1980s and 1990s. The empirical analysis contributes to a small but growing empirical literature interested in measuring the effect of local political competition on the geographic allocation of centrally-provided local public goods.7 I follow this literature in using a measure of election closeness to proxy for the intensity of political competition in a district. The empirical strategy also captures the long- run partisan loyalty of some districts in a novel way, by identifying those that repeatedly vote for a given party.8 A non-negligible side effect of controlling for a district's partisan loyalty is the attenuation of a potential omitted variable bias in estimates of the effect of election closeness on expenditures. The empirical strategy involves regressing policy outcomes on electoral outcomes, which

gives rise to well-known endogeneity problems. While previous studies had typically relied on cross-sectional data, the panel structure of the Québec data makes it possible to con- trol for fixed, unchanging geographic determinants of government spending.9 A second op- portunity to control for the potential endogeneity of political variables is provided by the distinctive linguistic pattern associated with partisan loyalty in Québec. A former French, then British colony, Québec is a linguistically divided society. Since the integration of the Province of Québec in the British Empire, linguistic divisions have had profound conse- quences for the political landscape. Local partisan loyalties today are still strongly corre- lated with the linguistic composition of local populations, which is plausibly exogenous to spending decisions. The analysis provides robust evidence that machine politics has played a key role in the

geographic allocation of road spending in Québec in the 1980s and 1990s. The paper's main result is that road spending tended to favor electoral districts that are loyal to the party in power, especially close to elections. There is no consistent evidence that the parties in power

7The recent contributions by Milligan and Smart (2005) and Larcinese et al. (2006a, 2006b) are the closest, in many respects, to the present paper.

8Larcinese et al. (2006a, 2006b) and Case (2001) are also interested in the role played by safe districts in the allocation of spending. However, their measures of 'safeness' do not exploit the dynamic nature of partisan loyalties.

9Milligan and Smart (2005) and Larcinese et al. (2006a) also use panel data, but most existing studies rely on cross-sectional data - e.g., Stein and Bickers (1994), Case (2001), Dahlberg and Johansson (2002).

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

120 Public Choice (201 1) 146: 1 17-143

have favored swing districts. Together, these results thus challenge the swing voter view of distributive politics, and lend support to the theoretical model's loyalty effect and machine politics equilibria.10 That machine politics patterns dominate in the allocation of road spending is consistent

with roads' long-lasting character - arguably a desirable feature from the point of view of politicians who are interested in cementing long-run loyalty relationships with voters. Previ- ous studies have tended to use data on either campaign spending or relatively small transfer programs.1 1 Unlike road spending, it is plausible to think that politicians would not perceive these expenditures to have sufficient long-term significance to be appropriate instruments for building enduring political support.12

The paper is organized as follows: In Sect. 2, 1 discuss the implications of a simple two- district model of distributive politics which nests the swing voter and the machine politics views of distributive politics, and Sect. 3 presents the model's empirical implementation. Section 4 describes the data used in the analysis and provides summary statistics. Baseline regression results are presented in Sect. 5, with instrumental variables (IV) and difference- in-differences results presented in Sect. 6. Section 7 concludes.

2 A dynamic model of distributive politics

In this section, I analyze the role of partisan loyalty in the context of a simple two-district model.13

2.1 A two-district model

Consider a simple model in which an incumbent government can affect its electoral prospects by allocating a fixed budget between two districts. For expositional purposes, one of the districts will be referred to as the 'swing' district (labeled with superscript j = s) and the other, as the 'loyal' district (labeled with superscript j = /).

The model captures two key differences between swing and loyal districts. First, the incumbent benefits from an 'initial electoral advantage' (which will be governed by the pa- rameter y) over potential challengers in the loyal district; however, in the swing district, the incumbent has no advantage and the playing field is level. Second, any electoral ad- vantage favoring the incumbent persists over time in the loyal district but not in the swing

10It must however be acknowledged that within-district swing voter patterns cannot be ruled out here as data on within-district partisan loyalties were not available. Larcinese et al. (2006b) use U.S. survey data to address this issue.

1 'Two examples are Peru's Social Fund in Schady (2000) or Sweden's environmental grants to municipalities in Dahlberg and Johansson (2002). Milligan and Smart (2005) study the allocation of regional development grants by the Canadian federal government. Although a portion of these grants are directed to local infrastruc- ture projects, they serve a variety of other purposes, including transfers to businesses and operating subsidies to local development agencies. Thus, the fact that Milligan and Smart do not find evidence of strong machine politics patterns associated with these grants should not be unduly surprising.

12In a recent closely related contribution, Diaz-Cayeros et al. (2007) argue instead that discretional, private, reversible goods are best suited to build long-run loyalty relationships. The Québec application presented in this paper supplies an instance of a discretional, public, irreversible good emerging as an instrument for machine politics.

13 It is relatively straightforward to extended the analysis to more than two districts - see Joanis (2009) for a generalization of the model to a large finite number of districts.

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 121

district (intertemporal persistence will be governed by the 'persistence factor' 5). These two differences between the districts are captured formally by the following assumptions:14

Assumption 1 y1 = y > 0 and ys = 0.

Assumption 2 8l = 8 e (0, 1] and 8s = 0.

I consider the following timing of events:

1 . At the beginning of period 1 , the government allocates spending between the two districts such that

e'-}-es=ë , with^,^Ä>0. (1)

2. At the end of period 1 , an election is held. 3. In period 2, a second election is held.15

Public spending ( ej ) and initial electoral advantage (yj) affect the incumbent's probabil- ity of being reelected in the period- 1 election (p{) in district j in the following way:

P¡ = ^ + F(yj + ej) for j € {i, /), (2) where yj >0, F' > 0, F" < 0, 0 < F(e) and F( 0) = 0.16 In such a framework, the initial electoral advantage ( yj ) lends itself to an intuitive interpretation in terms of po- litical competition. If yj is high, the incumbent benefits from having a strong advantage over her challengers, which corresponds to a situation involving low political competition. Conversely, if yj is low, the incumbent's advantage is small, which leads to a high degree of political competition.17 Given the concavity of F , the marginal effect of an increase in ej on reelection probability is decreasing in yj .

In the period-2 election, the probability of winning is determined as in (2), with the ex- ception that the electoral advantage derived from yj and ej is subject to some 'depreciation' over time:

PJ2 = ¿ +SJF(yJ +eJ) for j e {s,l}, (3)

14The results derived hereafter do not depend on ys and 8s being set to zero but rather on yl > ys and

8l >8S. However, ys - 8s = 0 is a convenient normalization. The positive correlation between yj and 8Ì implied by Assumptions 1 and 2 captures in a simple way the idea that a safe district today is also a district that is likely to deliver repeated victories in the future.

15 Note that spending takes place only once, i.e., before election 1, and that the entire budget is assumed to be distributed in period 1. However, the spending allocation will have impacts in both periods through the political process. Any subsequent budget to be allocated in the future is abstracted from to simplify the analysis.

16Similar concavity assumptions are adopted by Cox and McCubbins (1986), Lindbeck and Weibull (1993), and Dixit and Londregan (1996).

17To simplify the exposition, the two-district model does not consider districts in which challengers benefit from an electoral advantage, and such districts that are loyal to an opposition party. The reason is that the key trade-off of interest highlighted by the model is a consequence of some districts being loyal to the incumbent. From the point of view of the incumbent, the existence of districts being loyal to the opposition (i.e., sure losers) creates incentives that, if anything, reinforce the incentives associated with a high electoral advantage in favor of challengers.

Ö Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

122 Public Choice (201 1) 146: 1 17-143

where 0 < 8j < l.18

Now, consider an incumbent government whose period-i Bernoulli utility function is linear in the number of seats won:19

M,(n)=n, (4)

where n e {0, 1 , 2} is the number of seats. The government maximizes its total expected utility20 subject to (2), (3), the resource constraint (1) and Assumptions 1 and 2. This yields the following optimization problem for the government, reminiscent of a durable/nondurable

consumption problem or of a consumption/investment trade-off:

max [F(es) + (1 + ß8)F(y + ё- e*)} , (5) es

where ß is a discount factor (0 < ß < 1). Assuming that the problem has an interior solution, spending in the swing district is given by the following first-order condition (spending in the

loyal district is obtained residually):

F V*) = (1 + ßS)F'(y + è - e5*). (6)

The left-hand side of the equation is the marginal benefit of the last unit spent in district s ,

and the right-hand side is the marginal benefit of spending in district / (which has a period- 1 and a period-2 component) or, alternatively, the marginal opportunity cost of spending in district s. In equilibrium, these two quantities must be equal.21

2.2 Predictions

The key issue concerns which of the two districts should be expected to get more funding. The basic mechanism at work involves diminishing returns to spending, which follow from the concavity of F. Because of diminishing returns, public spending is less productive in

18Box-Steffensmeier and Smith (1996) find empirical support for such a 'law of motion' for electoral support.

Their estimates of 8J (in my notation) are in the order of .7-8, which is consistent with the interpretation of 8J as a depreciation factor.

19This government objective assumes away the issue of winning a majority of seats. Cox and McCubbins (1986), Dixit and Londregan (1996) and Lindbeck and Weibull (1993) also assume that political parties are merely vote or seat maximizers. A relevant alternative is the maximization of the probability of winning a majority of seats. Lindbeck and Weibull (1987) and Snyder (1989) contrast these two objectives. See Case (2001) for an excellent discussion.

20In any period, three events can occur: ut( 0) = 0 with probability (1 - p'){ 1 - pst), ut( 1) = 1 with prob-

ability 1 - (1 - p'){ 1 - pst) - p'pst , and M/( 2) = 2 with probability p'pst . This yields expected utility in period/:

Ut = 1 - (1 -p[)( 1 - pst) - p'tpï + 2 p'ps, ,

which reduces to:

Ut =Pt+Pt-

21 Obviously, other factors may affect reelection probabilities: for example, individual characteristics of politi- cians, characteristics of the local population, etc. Such undoubtedly important influences on local politics are abstracted from here in order to keep the exposition as simple as possible, but will be introduced in the empirics. See Sect. 3 for a discussion of the empirical implementation.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 123

terms of period- 1 marginal political support in the loyal district than in the swing district. Thus, the incumbent government has an incentive to direct more spending to the swing district - this captures, in a simple way, the standard 'political competition effect' that has been the main focus of the prior literature, and is consistent with the swing voter view of distributive politics. This incentive is stronger the greater the initial electoral advantage in the loyal district (y). Proposition 1 formalizes this idea.

Proposition 1 (Political competition effect) In a two-district setting , an increase in the initial electoral advantage of the incumbent government in the loyal district (y) unambigu- ously increases equilibrium spending in the swing district (and decreases spending in the loyal district).

Proof See Appendix. □

The fact that political support persists over time in the loyal district leads to a second, opposing incentive for the incumbent government. As long as ß > 0, the incumbent cares about the election to be held in period 2 and therefore values the support of the loyal district in the future. Spending in the loyal district is more valuable to the incumbent the higher the persistence factor in that district (5). Ceteris paribus , this 'loyalty effect' (formalized by Proposition 2) leads to more spending in the loyal district, consistent with the machine politics view of distributive politics:

Proposition 2 (Loyalty effect) In a two-district setting , an increase in the persistence of political support in the loyal district (5) unambiguously reduces equilibrium spending in the swing district (i and increases spending in the loyal district).

Proof See Appendix. □

Thus spending in the swing district is decreasing in the intertemporal link between elec- tions in the loyal district (governed by ß and 5) and increasing in the initial electoral ad- vantage favoring the incumbent in the loyal district (governed by y). Together, these two opposing effects lead to the key insight of the model, which is captured by the following proposition:

Proposition 3 Depending on the values taken by 5, y and ß, the two-district model has three types of equilibria:

(i) Swing district equilibria : es* > | > el*'

(ii) Machine politics equilibria : el* > | > es*' and (iii) An equal distribution equilibrium : es * = el* = | .

Proof See Appendix. □

Spending will be higher in the swing district if the persistence of political support (in the loyal district) is relatively low and the initial electoral advantage (also in the loyal district) is relatively high, leading to the first type of equilibria. However, the standard swing voter view of distributive politics is reversed here if the government cares sufficiently about the future

and if electoral support is sufficiently persistent in the loyal district, leading to the second type of equilibria. Note that the ambiguous result in Proposition 3 is a direct consequence

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

1 24 Public Choice (20 1 1 ) 1 46: 1 1 7- 1 43

of the time component in the government's optimization problem: in the static case, i.e., the case in which ß = 0, only the political competition effect is present and the swing district is always favored.

2.3 Relation to the previous theoretical literature

Relative to existing theories, the main theoretical contribution of the paper is the adoption of

a dynamic perspective of distributive politics to study the role of partisan loyalty. The model

shows that both swing voter and machine politics equilibria can arise in a dynamic context, whereas the static version of the model allows only for the former type of equilibrium.

This paper is not the first attempt to rationalize both machine politics and swing voter equilibria in a probabilistic voting framework.22 Dixit and Londregan (1996) provide a static model in which both types of equilibria are possible. The feature that plays a central role in triggering machine politics equilibria in the Dixit and Londregan model is the lower cost that political parties face when delivering favors to their own support groups. This arises because the government has an informational advantage in loyal constituencies, for example because politicians know their supporters' preferences better than those of citizens who are less loyal. While this assumption is plausible, a different route is followed here: the key effect of partisan loyalty is instead captured by loyal districts delivering enduring benefits to

the incumbent government (versus short-run benefits for swing districts).

Cox and McCubbins (1986) also propose a static probabilistic voting model in which ma- chine politics equilibria can arise, but not swing voter equilibria. Their model predicts that spending in loyal constituencies is a less risky strategy for securing winning coalitions than spending in swing constituencies, and that loyal constituencies should therefore be favored by risk-averse politicians. Studying loyalty building strategies in a dynamic framework per- mits the relaxation of this risk-aversion assumption. More generally, interest in non-Downsian outcomes pre-dates Downs's (1957) seminal

contribution and can be traced back to Smithies (1941), whose work has later been inter- preted as suggesting that threats of abstention may challenge the median voter theorem. Ma- chine politics outcomes can also arise if party leaders maximize not only their own welfare, as is typically assumed in this literature, but also their party members' welfare. Adopting this perspective, Besley and Preston (2007) deal with the implications of a heterogeneous population of loyal and swing voters. In their model, the party in power maximizes the wel- fare of its members, leading to a bias in favor of core supporters. Spending targeted towards swing voters arises as an electorally-driven deviation from this pattern, whereas spending benefiting the loyal voters is not directly driven by an electoral motive. The model devel- oped in this paper differs in that it assumes a purely opportunistic (but forward-looking) government.

The dominance of static models in the political economy literature is reflected in the extensive survey by Persson and Tabellini (1999), which restricts attention to such models. However, at least since Alesina's (1988) account of the crucial role of credibility, there is widespread acceptance of the idea that electoral politics is best thought of in a dynamic

^Probabilistic voting models, in which voters are assumed to react 'smoothly' to government policies, are simple and convenient for studying government behavior under electoral constraints. As a result, their use has become standard in the political economy literature and, more directly relevant to this paper, in models of distributive politics - see Lindbeck and Weibull's (1987, 1993) seminal contributions. For an extensive discussion of probabilistic voting models, see Persson and Tabellini (2000).

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (20 1 1 ) 1 46: 1 1 7- 1 43 1 25

framework.23 This paper is also related to the longstanding literature on ideology - see Hinich and Munger (1994) and Green et al. (2002). In a recent and closely related contri- bution, Diaz-Cay eros et al. (2007) also propose a dynamic model of distributive politics in which a risk-averse and rent-seeking incumbent party must allocate transfers to either swing

or loyal voters. In a game-theoretic framework, they highlight the role that reversible private

goods (e.g., transfers) may play in sustaining partisan loyalties in the future. As in the model presented above, catering to loyal voters becomes more likely when the incumbent party's interest in the future increases. While Diaz-Cayeros et al.'s model is undoubtedly relevant to our purpose, this paper presents a simple and empirically tractable model in which an irreversible public good - roads - is the instrument used by the incumbent party to sustain loyalty. Another noteworthy difference is that Diaz-Cayeros et al. are interested in the prob- lem of targeting swing versus loyal voters, while this paper highlights the trade-off between swing and loyal districts (abstracting from a district's distribution of voters).24 Although the empirical analysis that follows does not directly test for the relevance of one

modelling approach over the others,25 the results presented hereafter support the theoretical perspective adopted in this section, drawing attention to the key role of long-lasting partisan loyalties.

3 Empirical implementation

The empirical strategy is based on a generalization of the theoretical model presented in Sect. 2, to account for more than two districts and a larger set of district characteristics.26 Let us now think of a large finite number of districts differing by their persistence factor (8-i) and their initial electoral advantage (yj). It will be useful to allow the initial electoral advantage to be correlated with partisan loyalty, and to be influenced by other local and economy-wide political conditions:27

yJ = y{8J) + ÇJ, (7)

where y(8j) captures any systematic correlation between yj and 8j , and ^ stands for any other factor affecting local political competition.

23 More recently, influential dynamic political economy models have been developed by Besley and Coate (1998), explicitly extending the standard probabilistic voting model to a dynamic environment, and by Pers- son et al. (2000), setting out a model of politics and public finance, mainly intended to study the role of different political institutions on public finance outcomes. The case for adopting a dynamic perspective in the analysis of the "theory of political failure" has recently been convincingly reasserted by Battaglini and Coate (2007), this time within the framework of a legislative bargaining model.

24The models of Cox and McCubbins (1986), Lindbeck and Weibull (1987, 1993) and Dixit and Londregan (1996) are also cast at the voter level.

25Theory suggests other mechanisms through which the centralized provision of local public goods might lead to inefficiencies in spending decisions. For example, legislative bargaining models such as the one pro- posed by Milligan and Smart (2005) draw attention to the role of politicians' individual characteristics in their ability to attract public projects to their own constituency. Knight (2004) highlights the conflicting in- centives of individual legislators to increase own-district spending and restrain the own-district tax burden, while Cadot et al. (2006) focus on the link between the productivity of public capital and influence activities by corporate lobby groups.

26See Joanis (2009) for the technical details of that model.

27For example, the national political climate undoubtedly influences the incumbent's initial advantage in a given district.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

126 Public Choice (201 1) 146: 1 17-143

The following equilibrium condition forms the basis of the empirical strategy. For esti- mation purposes, this condition is extended to include other observable political and non- political determinants of public spending, that are assumed to enter the equation linearly, yielding:

e>* = G(8j) - (y(SJ) + $J) + ßV + 0Xj + eJ , (8)

where G (8j) is an increasing function of 8J , Zj stands for other political factors that may affect the allocation of spending (e.g., the role of powerful politicians in attracting spending to their own district), and Xj and ej are observable and unobservable district characteristics respectively.

The sign of the relationship between partisan loyalty and expenditure (i.e., the sign of

depends crucially on the sign of the correlation between loyalty and political com-

petition (i.e., the sign of the derivative y'(8j)). For the incumbent government, there is a trade-off if high loyalty districts tend to display large values for both yj and 8j , that is if y'(8j) > 0. In this case (for which this paper provides empirical evidence), the model predicts an ambiguous relationship between district expenditure and the degree of loyalty, depending on whether the political competition or loyalty effect dominates.28 A dominant loyalty effect would be consistent with the machine politics view of distributive politics, whereas a dominant political competition effect would be consistent with the swing voter view.

In Sects. 5 and 6, empirically-relevant versions of (8) will be estimated to test the the- oretical model's political competition effect (governed by yj) and loyalty effect (governed by 8j). Recall that according to the political competition effect (see Proposition 1 above), one would expect lower levels of expenditure where the intensity of political competition is low, e.g., where winning margins are typically high. The loyalty effect concerns the role that local spending plays in securing the support of loyal districts in the future (see Propo- sition 2). According to the loyalty effect, one would expect a positive relationship between expenditure and partisan loyalty.

4 Data and summary statistics

To assess the empirical relevance of the political competition and the loyalty effects de- scribed in the previous sections, I exploit rich data on the Québec government's road expen- ditures in each of the province's electoral districts. The expenditure data cover fiscal years 1986 to 1996, with the exception of 1991, when the data were not compiled by the Depart- ment of Transportation.29 There were 122 (provincial) districts before 1989, and there has been 125 since then.30 The expenditure data set is merged with two other sources of data,

28This case is a natural extension of the two-district model, in which such a positive correlation between yj

and 8J is implicitly assumed (see Assumptions 1 and 2).

29These figures have been produced using administrative data, internal to the Department of Transportation - Béland (various years). Aggregate figures may not match public accounts data. I refer to fiscal years as if they were calendar years, e.g., 1986 refers to the 1986-87 fiscal year. Publication of these data stopped after 1996.

30Over the period covered by this study, some redistricting occurred but most changes to district boundaries have been minor. In these cases, it is straightforward to link old and new districts and no further adjustment to the data has been made. However, in some cases, either districts have been split or new districts have been created from existing districts. Thus, the number of cases varies from year to year. Another source of variation in the number of cases has to do with missing data points in the official publications, which generally relate to urban districts where expenditure is very small.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 127

Table 1 Summary statistics: expenditure data (dependent variable)

Years Obs. Total expenditure Construction Maintenance Mean Min. Max. Mean Min. Max. Mean Min. Max.

1986 119 4840 0 20746 1999 0 14110 2841 0 10718

(4536) (2399) (2780) 1987 119 5129 0 29694 1888 0 23120 3241 0 11744

(5333) (3003) (3218) 1988 119 5480 0 28626 2363 0 21843 3118 0 10645

(5466) (3425) (3010) 1989 124 5328 0 25106 2089 0 17255 3238 0 22436

(5007) (2799) (3541) 1990 124 5775 0 28426 2269 0 21141 3506 0 13551

(5647) (4090) (3629) 1991 Not available

1992 113 5846 0 22170 2508 0 15257 3339 0 10920

(4965) (3067) (3019) 1993 113 5439 1 28609 2389 0 20336 3050 0 11881

(5101) (3337) (2916) 1994 118 5656 0 25855 2613 0 17148 3042 0 11078

(5752) (3429) (3059) 1995 115 5259 0 23071 2187 0 16848 3071 0 11912

(4982) (2698) (2970) 1996 121 5224 19 25995 2700 0 24712 2523 16 11206

(5232) (3435) (2629) All 1185 5396 0 29694 2299 0 24712 3098 0 22436

(5205) (3203) (3096)

Notes : Standard deviations in parentheses. 1992 Canadian dollars ('000$)

used to construct district-level covariates. The first of these sources provides demographic and economic data on each electoral district. The second source of district-level data con-

sists of official election results covering six general elections (1981, 1985, 1989, 1994, 1998 and 2003). Summary statistics on the variables used in the analysis are provided in Tables 1 and 2, which are now discussed in detail.

4.1 Expenditure data (dependent variable)

Table 1 documents the road expenditure data, which are used to construct the dependent variable in all empirical specifications. The average per district road expenditure was $4.84 million in 1986 (in 1992 Canadian dollars) and reached a peak of $5.85 million in 1992.31 In 1996, average expenditure had declined to $5.22 million. The maximum spending re- ceived by a single district varied from $20.75 million (in 1986) to $29.69 million (in 1987). Each year, a fraction of the 'ridings' - Canadian electoral districts - received zero or almost

31 All expenditure and income figures are expressed in 1992 Canadian dollars using provincial CPI (data provided by the Institut de la statistique du Québec).

ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

128 Public Choice (201 1) 146: 1 17-143

Table 2 Summary statistics: district characteristics and political variables

Variable Description Years Obs. Mean Std. Dev. Min. Max.

District characteristics

AREA District area (ln(km2)) 1991 125 5.55 2.81 1.20 12.75 POP District population (count) 1986 125 52242 7753 14530 68820

1991 122 55237 9927 13990 76535

1996 122 57099 11393 13765 82931

URB Urban population (share) 1986 125 .7605 .2655 .1081 1.0

FIRMS Manufacturing firms (count) 1988 124 115.52 75.48 7 426 UE Unemployment rate (%) 1986 125 12.46 4.84 5.3 29.17

1996 122 15.06 7.18 6.6 48.9

INC Mean household income 1985 125 41706 8563 25061 70520

(1992 Canadian dollars, $/year) 1995 122 41066 7971 24813 65892

FRENCH French-speaking pop. (share) 1986 125 .8185 .1990 .1305 .9896 1991 122 .8225 .2023 .1352 .9924

1996 122 .8056 .2087 .1313 .9818

Political variables

MAR Winning margin 1985 120 .2047 .1777 .0029 .8693 1989 125 .1581 .1083 .0024 .4984

1994 122 .2157 .1767 .0009 .7489

GOV Government dummy 1985 120 .8167 .3886 0 1 1989 125 .7360 .4426 0 1

1994 122 .6066 .4905 0 1

Partisan loyalty dummies (elections included)

LI 85,89,94 All 1250 .2912 .4545 0 1

L2 81,85,89,94,98,03 All 1250 .2032 .4025 0 1

L3 All past elections All 1250 .2752 .4468 0 1 LA All future elections All 1250 .3056 .4608 0 1

L5 81,85 All 1250 .2976 .4574 0 1

L6 98,03 All 1250 .3472 .4763 0 1

MIN Cabinet minister All 1250 .2016 .4014 0 1

zero expenditure.32 The expenditure figures include direct expenditure by the Department

of Transportation on the construction and maintenance of roads under its direct jurisdiction

and transfers to municipal governments for road improvement.33 On average, construction

expenditure represents 42% of total expenditure (with a low of 37% in 1987 and a high of

52% in 1995), the remainder being accounted for by maintenance expenditure.

32 A closer look at the data reveals that, each year, roughly one-fourth of the ridings receives essentially no spending. These ridings are typically the smallest urban districts.

33Most roads in Canada are under provincial/municipal jurisdiction. Any direct federal spending on in- frastructure is not included here.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 129

4.2 District characteristics

The following district characteristics are used in the analysis (see Table 2): the area covered by the district ( AREAj ), the size of the population ( POPJt ), the share of the population living

in urban areas ( URBj ), the share of the population that is French-speaking (FRENCH Jt), the number of manufacturing firms (FIRMS y), the unemployment rate ( UE' ), and the average household income (INCj ).34 The AREAj variable is the only one to which a log transfor- mation is applied in order to account for the wide size discrepancy between some large northern districts and the average district. This transformation conveniently linearizes the relationship between expenditure and district geographic size. Perhaps with the exception of population size, the districts vary widely with respect to these characteristics. Whereas the smallest district was 3 km2 (an urban district), the largest was 343,390 km2 (a northern district). The average riding had a population of 52,242 in 1986, 55,237 in 1991 and 57,099 in 1996. The share of the population living in urban areas varies from 10% to 100% and the share of the population whose main language is French (a group which forms more than 80% of the province's population) ranges between 13% to 99%. The unemployment rate varies between 5.3% and 48.9%, while the average household's real income is $24,813 in the 'poorest' riding (in 1995) and $70,520 in the 'richest' (in 1985).

4.3 Election data

Provincial politics in Québec, which is the focus of this paper, operates in a first-past-the- post system and was essentially bipartisan over the period of interest: the 'federalist' Québec Liberal Party and the 'independentisť Parti Québécois (PQ) have alternated in power since 1970.35 For the period most directly related to the expenditure data (1986-1996), the Liber- als were in power from 1985 to 1994, when the PQ took office, only to be replaced in power by the Liberals again in 2003. Table 3 provides some summary statistics on the elections held over the 1981-2003 period.

From the electoral data, several political variables are constructed. The main political variables measure the intensity of political competition - yj in the theoretical model - and the presence or not of long-run partisan loyalty - 8j in the theoretical model. A standard measure of 'closeness' of elections at the riding level (MARj) is used as a proxy for the intensity of political competition. This variable is defined in a straightforward manner for a particular district j and the last election before year t as36

MAR1, = Vj2' , (9) l2k=' Vjkt

34 Data on district characteristics come from the Directeur général des élections du Québec, the body respon- sible for organizing elections in the province - see Directeur général des élections du Québec (various years). Most of these data come from special tabulations from the census and, hence, do not vary every year (see Table lb for available years). Based on data availability, some of these variables are coded as time-invariant (they are AREAJ, URBJ and FIRMSJ).

35Two other parties have been represented in the National Assembly (N.A.) over the 1981-2003 period: the 'English-speaking' Equality Party (four members of the N.A. in 1989) and the 'conservative' Action démocratique du Québec (one elected in 1994). Separate elections are also held at the federal, municipal and school-board levels.

36In election years, the previous election is also used. The same convention is adopted by Milligan and Smart (2005), who use a similar measure of election closeness.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

130 Public Choice (201 1) 146: 1 17-143

Table 3 Summary statistics: Provincial general election results, Québec, 1981-2003

Vote date Number of seats in the National Assembly

QLP PQ EP ÄDQ Ш

General elections

April 13, 1981 42 80 122 Dec. 2, 1985 99 23 122

Sept. 25, 1989 92 29 4 125 Sept. 12, 1994 47 77 0 1 125 Nov. 30, 1998 48 76 0 1 125

April 14, 2003 76 45 0 4 125

Legend: QLP: Québec Liberal Party PQ: Parti Québécois EP: Equality Party (first ran in the 1989 election) ADQ: Action démocratique du Québec (first ran in the 1994 election)

where Vjkt is the number of votes cast for candidate к. К is the total number of candidates, and the candidates are ordered in decreasing order of their number of votes, such that vju stands for the number of votes for the winning candidate in district j, Vj2t stands for the

number of votes for the second most popular candidate, etc. Thus MARJt captures the mar- gin of the winner over total votes cast and will be used in the empirical analysis to capture the effect of political competition. Summary statistics are provided in Table 2. There is wide variation in winning margins across districts. For example, in the 1985 election, winning margins ranged from 0.23% to 86.93%. The average margin was 20.47% in the 1985 elec- tion, 15.81% in the 1989 election, and 21.57% in the 1994 election.

To capture a district's loyalty to the party in power, six closely related measures of par- tisan loyalty are used. They exploit the fact that loyal districts repeatedly vote for a given party, often over long periods. All share the same logic: LOYALJt = 1 if riding j repeatedly voted for the incumbent government in a given series of elections, 0 otherwise. The six loy-

alty variables (labeled LI to L6) capture different combinations of elections (see Table 2 for details). For example, according to LI a district is classified as 'loyal to the party in power' in year t if it voted for the party currently in power in the 1985, 1989 and 1994 elections.37 Depending on the measure being used, on average between 20% and 35% of districts can be classified as 'loyal' to the party in power. This approach to the measurement of partisan loyalty differs from the approaches followed in Case (2001) and Larcinese et al. (2006a). In those studies, vote shares for the incumbent party are used as measures of what Larcinese et al. label 'ideological bias.'38 To capture the dynamic aspect of partisan loyalty, the current application focuses on a measure of loyalty based on the extent of repeated support for the party in power.

Finally, two variables describe the status of individual politicians in the Québec parlia- ment (the National Assembly). The GOVj variable takes values 1 if the district is repre- sented by a member of the National Assembly (MNA) from the government party and 0

37 In an election year, the party forming the incumbent government is deemed the party in power.

38In a related paper, Larcinese et al. (2006b) measure ideological bias using exit polls. Such data are not available in Québec.

Ö Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 131

otherwise. In all three elections directly relevant to the expenditure data (1985, 1989 and

1994), majority governments were elected. Consequently, more than 50% of seats in the National Assembly were held by the government party, and as many as 82% following the

1985 election. Within the parliamentary delegation of the party in power, some MNAs are

also cabinet members. The MIN{ variable equals one if a district's MNA was a cabinet min- ister during the previous calendar year, 0 otherwise. On average, one out of five MNAs were

cabinet ministers in a given year between 1986 and 1996.

5 Main empirical results

In this section, I study the relative roles played by political competition and partisan loyalty

in the geographic allocation of road spending in Québec. The section proceeds as follows: Sect. 5.1 focuses on the effect of political competition. The standard test of the political competition effect in the literature involves regressing expenditure on a measure of elec-

tion closeness, generally winning margin. As a benchmark, results based on this standard approach, i.e., abstracting from partisan loyalty, are presented. Measures of partisan loyalty are introduced in Sect. 5.2. Section 5.3 then explores the composition of road expenditure by presenting separate results for construction and maintenance expenditure.

5.1 Political competition

In this subsection, the basic estimating equation relates spending in district j and year t ( EXPJt ) - the empirical counterpart of ej* in the theoretical model - to winning margin

(MAR{) in the previous election, controlling for a series of district characteristics:

EXP[ = a + rjcMARj * GOVj + rj0MARJt * OPPj + ßZ [ + 0Xj +<pt+ ф] + e{ , (10)

where a is a constant, OPPjt = 1 - GOV/ , <pt is a vector of year effects, and is a vector of district fixed effects. The dependent variable is measured as the level of road spending.39

Z/ includes the political variables GOVJt and MINJt, and X/ includes the following district

characteristics: area covered by the district (. AREAj ), population size ( POPj ), urban popu- lation share ( URBJ ), number of manufacturing firms ( FIRMSj ), unemployment rate ( UE' )

and household income (INCj ). Note that this initial specification excludes partisan loyalty, which will be introduced in

Sect. 5.2, in order to focus first on the correlation between winning margin and expenditure.

Equation (10) allows the effect of winning margin on expenditure to differ between ridings

held by the government (captured by the parameter rjG) and opposition parties ( rjo ).40

39Results are generally insensitive to changes in the definition of the dependent variable. Regressions using as the dependent variable per capita expenditure, budget shares and ratios to the average district yield very similar results, and are available upon request.

40Table 5 will also report benchmark results without this interaction - see Sect. 5.2.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

132 Public Choice (201 1) 146: 1 17-143

5.1.1 Benchmark results

The results for this benchmark regression are presented in the first two columns of Table 4.41

Specification (1) includes the X/ vector but no district fixed effects.42 Most 'economic' controls enter the regression significantly and with the expected signs. The area and urban population variables are strongly significant, with a positive sign for the former and a neg- ative sign for the latter. The unemployment rate is also significant and enters the regression positively (higher unemployment being associated with more spending), perhaps reflecting the role of transportation infrastructure in regional development policies. While the positive signs on the other two economic variables (income and number of firms) suggest a positive relationship between economic activity and spending, only the number of firms coefficient is statistically significant.43

Turning now to the political variables, the main parameters of interest are r]G and rjo (respectively the coefficients on MARJt * GOVJt and MARJt * OPP{ ). The basic empirical test

can be thought of as follows: consistent with the swing voter view of distributive politics, the theoretical model's political competition effect predicts that both rjG and rj0 should be negative. According to this effect, more spending should be directed to ridings with narrow margins regardless of which party currently holds the riding, those ridings being the most likely to be pivotal in the next election.44 However, Specification (1) displays a strong posi- tive effect of winning margin in government-held ridings (í) G > 0). This result thus seems to

sharply contradict the swing voter view of distributive politics and is more in line with the machine politics view. The coefficient on MARJt * OPPJt has the expected negative sign but is not statistically significant. The other two political variables ( GOV{ and MINJt) display insignificant effects.

Specification (2) exploits the panel structure of the data. By including fixed effects, it controls for fixed unchanging district characteristics. The results for Specification (2) show that fjG and fj0 have the same signs as in Specification (1) but neither of them is statistically

significant, with fjG now much smaller. Again, these results provide very little evidence in favor of the swing voter view.

Specification (3) presents the results from a fixed-effects regression on the subsample of districts that were in the first three deciles of the winning margin variable in 1985. The results

from this specification provide useful information with respect to a potentially nonlinear effect of the winning margin on expenditure. Indeed, it is for the largest margins that one would expect the swing district prediction to be the weakest. Hence, limiting the sample to close races introduces a bias against finding machine politics patterns, which are intuitively expected to be more prevalent for larger margins. Both fjG and f¡ 0 now have the negative sign predicted by the political competition effect. While the effect is now significant for opposition-held ridings, it is still insignificant for government-held ridings.45

41 Throughout the paper, standard errors are adjusted for clustering. Groups are defined according to the margin variable, which changes only once per electoral cycle in each district.

42Since some district characteristics are coded as time-invariant, inclusion of fixed effects absorbs them. In

specifications (2) to (5), AREA-ì , URB-i and FI RM S j are dropped and fixed effects are included.

43The number of manufacturing firms is central to the analysis of Cadot et al. (2006), which they interpret as a proxy for lobbying activities. My results corroborate the presence of a significant link between the number of firms and spending.

^There is no a priori reason to expect that the political competition effect should work differently in gov- ernment and opposition districts.

45 It may be argued that Specification (3) controls for the potential endogeneity of political variables, at least to some degree. According to Lee et al. (2004), by following over time a subgroup of districts where winning

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (20 1 1 ) 1 46: 1 1 7- 1 43 133

Table 4 Panel estimation results

(1) (2) (3) (4)

OLS Fixed effects Close races Electoral cycle

MAR*GOV*ELEC 41 20** 3041*

(1720) (1588)

MAR*GOV*PREELEC 1 799

(1334)

MAR *GOV*POSTELEC 1 487

(1012)

MAR*GOV 3507*** 238 -1496 -1881 -411

(972) (928) (3037) (1161) (979) MAR*OPP -1392 -1771 -7733** -1843* -1830

(1521) (1082) (3107) (1062) (1085) GOV -324 435 100 618 494

(466) (417) (784) (416) (418) ELEC -534 47

(326) (411)

PREELEC -341

(292)

POSTELEC -293

(249)

MIN 528 210 -415 111 183

(367) (274) (524) (267) (274)

AREA 867***

(135)

POP .0179 -.0120 -.0394 -.0103 -.0117

(.0213) (.0343) (.0507) (.0308) (.0345) URB -5999***

(1381)

FIRMS 3.71*

(1.92)

UE 115*** -92* -80 -71 -90*

(45) (51) (180) (53) (51)

INC .0336 -.0427 .0630 .0327 -.0401

(.0250) (.0721) (.1409) (.0728) (.0721)

Fixed effects no yes yes yes yes

Year effects yes yes yes no yes R2 .5646 .7555 .8134 .7544 .7568 Observations 1158 1168 345 1168 1168

Notes : Dependent variable: district-level expenditure. Constants included but unreported. Robust standard errors in parentheses, adjusted for clustering.

Levels of statistical significance: 1 % , 5% and 10%

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

134 Public Choice (201 1) 146: 1 17-143

5.7.2 Electoral budget cycle

The first three specifications in Table 4 make the strong assumption that the impact of po- litical variables such as MARJt * GOVj are constant over time. Specifications (4) and (5) allow the impact of MARj * GOVj to vary over the electoral cycle.46 In Specification (4), MARj *GOVj is interacted with three electoral cycle dummies: ELECt (election years: 1989 and 1994), PREELECt (pre-election years: 1988 and 1993), and POSTELECt (post-election years: 1986, 1990 and 1995). The coefficients on all three interaction terms are positive. However, MARj * GOVj is significant only when interacted with the ELECt dummy, re- vealing that a lot of the action is concentrated in election years. Note that the coefficient on MARj * OPPj (which is not interacted with electoral cycle dummies here) has the expected negative sign and is marginally significant. Specification (5) is presented as a robustness test for the positive sign on MARj * GOVj * ELECt in Specification (4). Interactions with PREELECt and POSTELECt are dropped, and year effects are included. The pattern of in- terest (the positive sign on the estimated coefficient for MARj * GOVj * ELECt) appears to be robust.

These results indicate that the dynamics in opposition ridings tend to conform to the standard swing voter view but that, in government-held ridings, there is no supporting ev- idence.47 Furthermore, the effect of winning margin is positive and significant in election years, when electoral competition is expected to be the strongest. On average, government- held ridings with large winning margins in the previous election received greater road spend-

ing in election years. The estimated effect is economically significant, a one percentage- point increase in winning margin being associated with $40,000 worth of spending in elec- tion years. The remainder of this section argues that this pattern is largely explained by the positive correlation between winning margin and partisan loyalty.

5.2 Partisan loyalty

The large positive coefficients on MARj estimated for government-held ridings in the pre- vious subsection are puzzling if one's prior is the swing voter view of distributive politics. Why would rational politicians not target swing districts, especially close to an election? I argue that these estimates might suffer from an omitted variable bias related to the role played by partisan loyalty. High margins tend to be associated with strong partisan loyalty. And the theoretical model of Sect. 2 develops one rationale as to why loyalty might be a determinant of the allocation of spending across districts. In terms of (10), the coefficient on MARj * GOVj will be biased if (i) MARj * GOVj is correlated with partisan loyalty, and (ii) if the error term is also correlated with loyalty.

margins were initially narrow, it is possible to isolate a group of districts that share similar unobservable characteristics. Unfortunately, given that the variable of interest here is the winning margin, this strategy is obviously not fully satisfactory for our purposes since using margin to split the sample effectively treats it as a control variable. Note also that there is a trade-off here in restricting the sample to closer races, which would arguably reduce the endogeneity bias but also reduce the number of observations and hence the precision of the results. Unreported results show that choosing a lower cutoff does not significantly alter the qualitative pattern of the political variables. For a more comprehensive discussion of potential endogeneity issues, see Sect. 6.

There is a large body of literature on political budget cycles, the well-known phenomenon that aggregate government budget fluctuations are influenced by political dynamics. Brender and Drazen (2005) revisit the evidence on the political budget cycle and, in a related paper, Drazen and Eslava (2006) provide a theoretical model of redistributive politics in which swing regions are targeted before the election.

47Milligan and Smart (2005) find a similar dichotomy.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 135

Table 5 Pooled regressions with loyalty

MAR LOYAL GOV MIN FE R2 con(MAR, LOYAL)

Benchmark regression (no control for loyalty) 2089** 453 685* No .5608 (832) (338) (364)

(LI) Loyal for 3 elections (85, 89, 94) 1072 1110*** 118 576 No .5676 .33*** (880) (384) (355) (356) (6.4)

(L2) Loyal for 6 elections 700 1673*** 100 448 No .5739 .36*** (8 1 , 85, 89, 94, 98, 03) (829) (407) (345) (345) (6.9)

(L3) Loyal in the past (81 onwards) 39 1510*** 36 457 No .5710 .50*** (895) (385) (353) (350) (9.3)

(L4) Loyal in the future 1228 1298*** 81 553 No .5717 .28*** (813) (350) (344) (349) (5.4)

(L5) Loyal in the past (81 and 85 only) 103 1608*** 90 458 No .5744.46*** (884) (373) (342) (350) (8.2)

(L6) Loyal in the future (98 and 03 only) 881 1359*** 56 546 No .5729.35*** (811) (331) (345) (344) (6.3)

(L2) Loyal for 6 elections -529 236 729* 268 Yes .7551 (81, 85, 89, 94, 98, 03) (765) (482) (410) (270)

(L2) Loyal for 6 elections -84 907*** 309 Yes .7543 (81,85,89,94,98,03) (751) (306) (268)

(L2) Construction expenditure only -835 778** 273 Yes .4721 (711) (306) (258)

(L2) Maintenance expenditure only 750* 128 36 Yes .8280 (390) (200) (163)

Notes : Dependent variable: district-level expenditure. Constants included but unreported. Robust standard errors in parentheses (robust /-stats in the last column), adjusted for clustering Levels of statistical significance: 1%***, 5%** and 10%* n = 1 158. Full set of district characteristics (X) and year effects included

Regardless of the loyalty measure (LI to L6) being used, there is indeed a strong positive

correlation between MARJt and LOYALj (see the last column of Table 5). The coefficient

of correlation between these two variables varies from 0.28 for L4 (loyalty defined over

all future elections) to 0.50 for L3 (loyalty defined over all past elections) and is always significantly different from zero at the 1 % confidence level. Omitting loyalty from the re-

gressions will therefore be a concern to the extent that partisan loyalty is in itself a factor in

the geographic allocation of spending, as suggested by the theoretical model.

In this subsection, I take this concern seriously and present results based on the following

equation:

EXPJt = a + yMARj + SLOYALj + ßZ [ + <9Х/ + <pt + <¡)j + eJt . (11)

This specification includes the partisan loyalty variable and provides evidence on the rela-

tive influence of political competition and loyalty on the allocation of spending. The main

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

136 Public Choice (201 1) 146: 1 17-143

parameters of interest are now y and 8. In line with the swing voter view, y is expected to be negative. Consistent with the machine politics view, 8 is expected to be positive.48 Table 5 reports results from regressions with the six loyalty variables, with and without

fixed effects. It also reports the results from a benchmark regression excluding LOYALj. Mirroring the results presented above, the coefficient on MARJt (y) is positive and signifi- cant in the benchmark regression. Regardless of which loyalty measure is being used, the inclusion of LOYALJt in the regression considerably decreases the coefficient on MARJt . Al-

though it remains positive in most cases, it is never significant. In contrast, the coefficient on LOYALj (<5) is positive and significant at the 1% confidence level in all specifications but one.

When (1 1) is estimated with fixed effects, the coefficient on LOYAL¡ is still positive but not significant.49 The sudden explanatory power of the GOVJt variable when fixed effects and the loyalty variable are introduced is puzzling, as it is the only specification in which this variable displays a significant effect. Note that with fixed effects, the impact of loyalty - essentially a fixed district characteristic - is identified from changes in the loyalty variable. By construction of the loyalty variables used in this study, such changes occur only when there is a change in government. In the current context, this occurred only in 1994. Given

this limited variation, changes in the loyalty variable are hard to disentangle from changes in the GOVj variable, many of which correspond to the changes in LOYALj - see Sect. 6.2 for a difference-in-differences strategy which actually exploits the 1994 change in government.

To address this concern, I also provide results from a fixed-effect regression without the GOVj variable. These results show an estimate of the effect of loyalty that is strongly significant. Although smaller than in the regressions without fixed effects, the latter effect is economically significant: to illustrate the order of magnitude, a loyal district received 17% more spending than the average district.

5.3 Construction and maintenance expenditure

The data allow for a separate analysis of construction and maintenance expenditure, with the former containing major road improvement projects. One might expect maintenance expen- diture to be less responsive to political considerations and more responsive to local needs than construction expenditure. This is indeed what the results in the last two lines of Table 5 indicate. While partisan loyalty has a positive and strongly significant effect on construction expenditure, the effect is considerably smaller (and not significant) for maintenance expen- diture. This result suggests that major projects, presumably those with the biggest long-term value to voters, are being driven by partisan loyalty. The positive coefficient on MARj in the maintenance expenditure regression (significant at the 10% level) is hard to interpret and once again casts doubt on the presence of a significant political competition effect in the behavior of Québec governments over the 1986-1996 period.

Taken together, the results presented in this section illustrate the difficulty of identifying any evidence of the standard swing voter view in the Québec data. They do, however, pro-

48Since the focus of this subsection is on the partisan loyalty effect, the regressions do not allow the effect of winning margin to differ in government-held and opposition-held districts. However, note that since loyalty to the party in power is taken into account, one should not expect a difference in the effect of winning margin in government versus opposition ridings.

49Table 5 presents results for fixed effects regressions only with loyalty measure L2. As shown by results for the six loyalty measures without fixed effects, the results are only slightly sensitive to the definition of

loyal{.

^ Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 137

vide stronger support for the machine politics view. Section 6 below shows that this overall

picture is robust when accounting for the potential endogeneity of political variables.

6 Robustness and endogeneity

In this section, the robustness of the results presented in Sect. 5 is assessed by means of in- strumental variables (IV) and difference-in-differences strategies to account for the potential

endogeneity of the LOYALjt variable. As suggested by the theory discussion in Sect. 2, par- tisan loyalty is the product of repeated interaction between parties and voters. Hence, while loyalty can be expected to be a causal factor in the allocation of spending, it is also likely that

causality works in the opposite direction if governments actually spend with the intention of nurturing local partisan loyalties. More generally, endogeneity biases will arise if non- observable considerations, e.g., preferences for public goods, are correlated with both elec- toral outcomes (specifically partisan loyalty) and the geographic allocation of road spending. To get a sense of the likelihood that partisan loyalty is picking up some unobserved

heterogeneity across districts, Table 7 compares the 28 districts that were loyal to the Liberal party in all elections between 1981 and 2003 (i.e., according to L2) to the other 97 districts, based on observable characteristics. Suggesting that unobserved heterogeneity might be an issue, 'Liberal strongholds' are statistically different from the other districts along three dimensions: loyal districts tend to be slightly smaller, have a lower unemployment rate, and have a much smaller share of French-speakers. The latter is the main observable difference between liberal strongholds and other districts and will form the basis for the IV strategy that follows.

Based on these observations, the direction of the potential OLS bias affecting the LOYALJt

coefficient is unclear. On the one hand, Liberal strongholds tend to be economically dynamic areas (as suggested by the low unemployment rate) and hence can be expected to have a strong need for new or improved roads. If this is true, one should expect the OLS estimates to be upward-biased. On the other hand, Liberal strongholds tend to be small urban districts, which can be expected to be characterized by a low preference for road spending compared to other public spending. This alternative story suggests that OLS estimates might instead be downward-biased.

6. 1 Instrumental variables

The IV strategy uses the French-speaking population variable (FRENCH /) as an instru- ment for partisan loyalty. The rationale for this instrument comes from a fundamental char- acteristic of the political environment in Québec: partisan loyalties and language spoken are strongly correlated. Roughly 80% of the province's 7-million population are French- speaking, the majority of whom descend from original French settlers and have a Roman Catholic background. The English-speaking population, which forms a majority in Canada as a whole, is the most important linguistic minority in Québec. This British (and usually Protestant) presence in Québec goes as far back as 1760, when New France was integrated into the British Empire. The Parti Québécois, which advocates the province's independence from Canada, draws almost all of its support from the French-speaking community. In con- trast, loyalty to the Liberal Party (in office for most of the period covered by this study) tends to arise in districts where the English-speaking population is concentrated (e.g., Western Montréal). Anecdotal evidence for this is provided by the fact that among the 12 strongest wins for the Liberals in 1985 (the top decile), 1 1 occurred in Western Montréal ridings.

â Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

138 Public Choice (201 1) 146: 1 17-143

Table 6 Summary statistics: liberal strongholds vs. other ridings, 1986

Variable Loyal to the Liberals Others Diff. (r-stat)

AREA 4.7 5.8 -1.8*

POP 52.962 52.034 0.6

URB 81 74 1.1

FIRMS 128 112 1.0

UE 11.1 12.8 -1.7* INC 42.937 41.351 0.9

FRENCH 65 87 -5.6***

Number of ridings 28 97

Notes : $$$ j|( j|c s|e

Levels of statistical significance: 1% ,5% and 10% Loyalty measure: (L2). Two-sided í -tests

The IV regressions are conducted under the assumption that language is in itself not a di- rect determinant of the level of transportation expenditure received by a district. If language

has an influence on spending patterns, it is taken here to be mediated by the political process (through its influence on partisan loyalty). This is what the first stage regression captures: the linguistic composition of a riding is a key determinant of the nature of partisan loyalty in

that riding. In the second stage, partisan loyalty itself (together with the intensity of political

competition) captures the ability of politicians to bias the allocation of spending for electoral purposes.

The bottom panel of Table 6 presents first-stage diagnostics documenting the strong cor- relation between FRENCH j and LOYALJt. The correlation between the two variables is strong, ranging from 0.29 for loyalty variable LI to 0.46 for L5. The usual F-tests and partial R2 measures confirm that, regardless of which definition of the loyalty variable is used, FRENCHj has strong predictive power in the first-stage regression.

IV results, featured in the top panel of Table 6, are qualitatively similar to the previous results. In fact, the effect of partisan loyalty is slightly bigger and still statistically significant

in all specifications (except again for maintenance expenditure). The coefficient on winning margin is negative in most specifications but, as before, is never significantly different from

zero. These results confirm the robustness of the previous section's results, and suggest that causality is working in the expected direction, i.e., from partisan loyalty to spending.

The fact that the IV estimates tend to be bigger than their OLS counterparts is noteworthy

and likely due to the fact that the first-stage regression underscores the effect of politically powerful English-speaking ridings (the core supporters of the Liberal party), hence reinforc- ing the estimated impact of loyalty on expenditure. As suggested by Dixit and Londregan (1996), it may be less expensive for the government to cater to its core supporters, for or- ganizational or informational reasons. If this is the case, then IV results will remain upward biased. Nevertheless, even if they do not allow for a direct test of the theoretical model of Sect. 2 against Dixit and Londregan' s model, these IV results suggest that core supporters within loyal districts are driving the spending allocation in their favor. Indeed, the FRENCH Jt

variable can be interpreted as a rough proxy for the within-district distribution of partisan loyalties in Québec. And results show that a large proportion, in a district, of the language group that is traditionally loyal to the party in power tends to reinforce the correlation be- tween a district's loyalty and road spending.

â Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 139

Table 7 Pooled IV regressions

MAR LOYAL GOV MIN R2

OLS: (LI) Loyal for 3 elections (85, 89, 94) 1072 1110*** 118 576 .5676 (880) (384) (355) (356)

(LI) Loyal for 3 elections (85, 89, 94) -1085 3465** -594 344 .5373 (1604) (1377) (545) (401)

(L2) Loyal for 6 elections (81, 85, 89, 94, 98, 03) -993 3712** -331 160 .5545 (1571) (1510) (476) (432)

(L3) Loyal in the past (81 onwards) -3153 3861** -613 103 .5464 (2367) (1548) (562) (439)

(L4) Loyal in the future 111 2982*** -402 381 .5534 (1206) (1150) (496) (372)

(L5) Loyal in the past (8 1 and 85 only) - 1703 307 1 *** -240 252 .5632 (1753) (1160) (443) (396)

(L6) Loyal in the future (98 and 03 only) -585 3008*** -425 378 .5552 (1367) (1131) (502) (359)

(L2) Construction expenditure only -1224 2552** -180 50 .2104 (1154) (1103) (352) (347)

(L2) Maintenance expenditure only 231 1160 -151 110 .6822 (807) (740) (237) (199)

First-stage diagnostics Correlation F-test Partial R2 (Ll) -.29*** 29.9*** .09 (L2) -.36*** 27.5*** .09 (L3) -.39*** 29.7*** .09 (L4) -.35*** 39.0*** .10 (L5) -.46*** 43.9*** .12 (L6) -.40*** 42.8*** .10

Notes : Dependent variable: district-level expenditure. Constants included but unreported. Robust standard errors in parentheses, adjusted for clustering

Levels of statistical significance: 1%***, 5%** and 10%* n = 1 158. Full set of district characteristics (X) and year effects included. No district fixed effects. LOYAL instrumented with FRENCH. First-stage diagnostics for the excluded instrument (FRENCH): robust test sta- tistics, adjusted for clustering

6.2 Difference-in-differences

An additional caveat of the above IV strategy follows from the fact that FRENCH ' is es- sentially a time-invariant district characteristic. Therefore, in this particular application, it is

not a suitable instrument in the fixed effects regressions (fixed effects are accordingly ex- cluded from the IV regression). But the fact that there was a change of government in 1994 allows for a different identification strategy which exploits variation over time in the loyalty variable.

The rationale is simple: the extra spending directed to ridings that are loyal to the Liberals

while this party is in power should go away when the PQ takes office in 1994. This suggests a difference-in-differences strategy that compares spending in ridings that are loyal to the Liberals (/) to spending in the other ridings (o), before and after the 1994 election. Here, the

Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

1 40 Public Choice (20 1 1 ) 1 46: 1 1 7- 1 43

Table 8 Difference-in-differences estimates

All expenditure Construction Maintenance

Loyal Liberal ridings - Liberals in power 5999 2882 3117 Loyal Liberal ridings - PQ in power 4634 1981 2653

Difference (1) 1365 901 464 (1249) (598) (760)

Other ridings - Liberals in power 5270 2079 3191 Other ridings - PQ in power 5417 2586 2830

Difference (2) -147 -507 360

(579) (328) (335)

Difference-in-difference (1H2) 1511 1407** 104 (1377) (683) (831)

D-in-D with full set of controls 990 1 160** - 1 70 (734) (535) (406)

Notes : Dependent variable: district-level expenditure. Robust standard errors in parentheses, adjusted for clustering. Loyalty measure: (L2) ** Significant at the 5% confidence level Full set of controls includes district characteristics (X), political variables (Z) and year effects. No district fixed effects

effect of partisan loyalty is identified as follows:

5 = (ШРт_94 - mP95^) - (ЁЩ 6-94 - £^95-96). (12)

where the upper bars denote averages. In terms of controlling for the potential endogeneity

of partisan loyalty, the main advantage of this approach is that it differences out any fixed

systematic difference between ridings that are loyal to the Liberal party and the rest of the

province. Table 8 presents the results pertaining to this difference-in-differences exercise. Results

are presented for all expenditure and for construction and maintenance expenditure sepa-

rately. I also present results from a regression with the full set of district characteristics.

The first panel of Table 8 shows that ridings that were loyal to the Liberals experienced on

average a $1.4-million drop in total road expenditure per district after the PQ took office

in 1994, two-thirds of this drop being attributable to construction expenditure. Meanwhile,

the other districts experienced a modest $147,000 increase in total expenditure, which hides

a $0.5-million increase in construction expenditure coupled with a $360,000 decrease in maintenance expenditure (see the second panel of Table 8). The difference-in-differences estimate is positive and significant for construction expenditure, but again not for mainte-

nance expenditure. This result is robust to the inclusion of the full set of controls. Although

the estimated loyalty effect is still positive and of the same magnitude as in other identifica-

tion strategies presented above, it is not estimated with sufficient precision to be statistically

significant for all expenditure. Nevertheless, these results provide additional evidence that

loyal ridings have received more road construction expenditure over the 1986-1994 period.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 141

7 Conclusion

This paper has examined an important dimension of government behavior with respect to the centralized provision of local public goods, namely the geographic patterns of pork- barrel politics. Two opposing predictions dominate the theoretical literature on this issue: the swing voter view, following Lindbeck and Weibull (1987, 1993) among others, and the machine politics view, formalized by Cox and McCubbins (1986). According to the former, public spending is expected to favor voters likely to be pivotal in the next election; according to the latter, spending is instead expected to favor voters that form the traditional electoral base of the incumbent government, namely loyal voters.

The dynamic political economy model laid out in this paper, in which electoral districts are heterogeneous with respect to their partisan loyalty, combines the two views of pork- barrel politics in a transparent way, making clear how they follow from incentives pertaining to different time horizons. The model demonstrates that a political competition effect and a loyalty effect can operate at the same time, working against each other to produce an ambiguous short-run relationship between political competition and public spending at the district level.

To shed light on the relative importance of these two forces empirically, I exploited a rich data set which documents the allocation of public expenditure on roads amongst elec- toral districts in Québec. Specifically, I explored the empirical relationship between partisan loyalty, political competition and the geographic distribution of public spending, providing robust evidence that districts which display loyalty to the incumbent government receive disproportionately more spending. The evidence also indicates that the standard swing dis- trict prediction is not the main factor driving the interaction between politics and expendi- ture allocation in Québec 's recent experience, although there is some evidence of additional spending being directed towards districts held by opposition parties where election out- comes were close. Furthermore, road spending exhibits an electoral cycle, with machine politics patterns especially discernible close to elections. Overall, these results show that, in the case of road spending, long-run political relationships are a key determinant of the allocation of centrally-provided public goods.

In a more general setting than the one developed in the paper, one might envisage the government being able to pull a variety of pork-barrel levers, ranging from those well-suited to yielding short-term political advantages just prior to election time (in the limit, pure cash) to much longer-term investments that may help secure enduring political support. In provid- ing a panel data analysis of an important example of the latter (road spending), this paper complements other work in the literature that has focused on more short-term discretionary projects. The results suggest that a minimal requirement for observing machine politics pat- terns is that the spending instrument in question has the necessary long-term significance for voters. In future work, it will be useful to revisit these issues using comprehensive data on different types of public expenditure displaying different degrees of durability. A promis- ing first step in that direction is provided by Diaz-Cayeros et al.'s (2007) model of political portfolio diversification.

An important caveat of the analysis is that it does not directly tackle the key issue of within-district distributive politics. As the relevant data becomes available, future research should assess whether the extra money flowing to loyal districts benefits loyal voters or swing voters.

Acknowledgements I thank Robert McMillan, Michael Smart, Michael Baker, Adonis Yatchew, François Vaillancourt, André Blais, Timothy Besley, Brian Knight, and an anonymous referee for their suggestions. I also thank seminar participants at U. de Sherbrooke, U. of Toronto, Industry Canada, U. of Oklahoma, CRA

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

142 Public Choice (201 1) 146: 1 17-143

International, U. of Louisville, U. of Kentucky, UC-Merced, PPIC, SUNY-Albany, CEA (Montréal 2006), SCSE (Montréal 2006) and the First World Meeting of the Public Choice Society (Amsterdam 2007). I am most grateful to Benoît Méthot for his input in the early stages of this research. All remaining errors are mine.

Appendix

Proof of Proposition 1 Totally differentiating (6) with respect to es* and y yields

de5* ^ (l+ß8)F"(y + e-es*) dy F"{es*) + {'+fi8)F"(y + ë-es*)~ '

which is also signed in a straightforward way by means of the properties of F. □

Proof of Proposition 2 Totally differentiating (6) with respect to es* and 8 yields

des* F'(v + ë - e5*) } - des* = d8 F"(es*) + (1 + ß8)F"(y + e - e5*)

which is signed in a straightforward way by means of the properties of F. □

Proof of Proposition 3 First, consider the case where es * = e1* = | (i.e., the two districts

receive an equal share of the budget). Condition (6) must be satisfied, soF/(|) = (l + ß8)F'(y + f ). (a) For a given value of y, denoted y , the latter condition defines the required

F' (^ ) value of 8 as a function of y and ë : 8(y , e) = - 1). Note that to have 8 < 1 it

must be the case that y is not too high. Now consider an increase in es* of € above | and,

accordingly, a reduction of e in el*. This yields: 8(ý , ë , e) = i( л ~~ !)• Since F" < P F (y+|- €) л

0, we have: 8(y, ê, e) < 8 (y, ë, 0). Similarly, we have: 8 (y, e, -e) > 8 (y, e, 0). Hence, for

a given value of y, es* > | > el* iff 8 is relatively low, and e1* > | > es* iff 8 is relatively high, (b) Now, for a given value of 8 , denoted 8, this condition defines the required value

- - « f'(4) ■ - of y as a function of 8 and e : y(8 , e) = F'~ « ) - f , which must satisfy y{8 , ë) > 0.

Consider again an increase in es* of e above | and a reduction of e in el*. This yields:

y(8 , ê, €) = II ß^) - § +€. Again since F " < 0, we have: y (8, ë , e) > y (8, ë , 0) and y (8, ë, -e) < y (5, ë , 0). Thus, for a given value of 5, es* > | iff y is relatively high, and e1* > | iff y is relatively low. □

References

Alesina, A. (1988). Credibility and policy convergence in a two-party system with rational voters. American Economic Review , 78(4), 796-805.

Battaglini, M., & Coate, S. (2007). Inefficiency in legislative policy-making: A dynamic analysis. American Economic Review, 97(1), 118-149.

Béland, M. (various years). Compilation des dépenses ministérielles par circonscription électorale provin- ciale sur le réseau routier (title varies from year to year). Ministère des transports du Québec.

Besley, T., & Coate, S. (1998). Sources of inefficiency in a representative democracy: A dynamic analysis. American Economic Review , 55(1), 139-156.

Ö Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Public Choice (201 1) 146: 1 17-143 143

Besley, T., & Preston, I. (2007). Electoral bias and policy choices: Theory and evidence. Quarterly Journal of Economics, /22(4).

Box-Steffensmeier, J. M., & Smith, R. M. (1996). The dynamics of aggregate partisanship. American Political Science Review , 90(3), 567-580.

Brender, A., & Drazen, A. (2005). Political budget cycle in new versus established democracies. Journal of Monetary Economics, 52 , 1271-1295.

Cadot, O., Roller, L.-H., & Stephan, A. (2006). Contribution to productivity or pork barrel? The two faces of infrastructure investment. Journal of Public Economics , 90, 1 133-1 153.

Case, A. (2001). Election goals and income redistribution: Recent evidence from Albania. European Eco- nomic Review , 45 , 405-423.

Cox, G. W., & McCubbins, M. D. (1986). Electoral politics as a distributive game. Journal of Politics, 48(2), 370-389.

Dahlberg, M., & Johansson, E. (2002). On the vote-purchasing behavior of incumbent governments. Ameri- can Political Science Review, 96(1), 27-40.

Diaz-Cayeros, A., Estévez, F., & Magaloni, В. (2007). Strategies of vote buying: Poverty ; democracy and social transfers in Mexico. Book manuscript.

Directeur général des élections du Québec (various years). Dossiers socio-économiques des circonscriptions électorales.

Dixit, A., & Londregan, J. (1996). The determinants of success of special interests in distributive politics. Journal of Politics, 55(4), 1 132-1 155.

Downs, A. (1957). An economic theory of political action in a democracy. Journal of Political Economy, 65(2), 135-150.

Drazen, A., & Eslava, M. (2006). Pork barrel cycles (NBER working paper, no. 12190). Francia, P. L., & Levine, R. (2006). Feast of famine at the federal luau? Understanding federal net spending

under Bush. The Forum, 4, article 6. Green, D., Palmquist, В., & Schickler, E. (2002). Partisan hearts and minds: Political parties and the social

identities of voters. New Haven: Yale University Press. Hinich, M. J., & Munger, M. С. (1994). Ideology and the theory of political choice. Ann Arbor: University

of Michigan Press. Joanis, M. (2009). The road to power: Partisan loyalty and the centralized provision of local infrastructure

(CIRANO working paper, no. 2009s-46). Knight, B. (2004). Parochial interests and the centralized provision of local public goods: Evidence from

congressional voting on transportation projects. Journal of Public Economics, 88, 845-866. Larcinese, V., Rizzo, L., & Testa, C. (2006a). Allocating the U.S. federal budget to the states: The impact of

the President. Journal of Politics, 68(2), 447-456. Larcinese, V., Snyder, J. M., & Testa, C. (2006b). Testing models of distributive politics using exit polls to

measure voter preferences and partisanship, Political economy and public policy series, no. 19. Suntory Center, London School of Economics and Political Science.

Lee, D. S., Moretti, E., & Butler, M. J. (2004). Do voters affect or elect policies? Evidence from the U.S. House. Quarterly Journal of Economics, 807-859.

Lindbeck, A., & Weibull, J. W. (1987). Balanced-budget redistribution as the outcome of political competi- tion. Public Choice, 52(3), 273-297.

Lindbeck, A., & Weibull, J. W. (1993). A model of political equilibrium in a representative democracy. Journal of Public Economics, 51, 195-209.

Milligan, K., & Smart, M. (2005). Regional grants as pork barrel politics (CES-ifo working paper, no. 1453). Moser, С. (2008). Poverty reduction, patronage or vote-buying? The allocation of public goods and the 2001

election in Madagascar. Economic Development and Cultural Change, 57(1), 137-162. Persson, T., & Tabellini, G. (1999). Political economics and public finance. In A. Auerbach & M. Feldstein

(Eds.), Handbook of public economics. Amsterdam: North-Holland. Persson, T., & Tabellini, G. (2000). Political economics: Explaining economic policy. Cambridge: MIT Press. Persson, T., Gérard, R., & Tabellini, G. (2000). Comparative politics and public finance. Journal of Political

Economy, 108(6), 1121-1161. Schady, N. R. (2000). The political economy of expenditures by the Peruvian Social Fund (FONCODES),

1991-95. American Political Science Review, 94(2), 289-304. Smithies, A. (1941). Optimum location in spatial competition. Journal of Political Economy, 49(3), 423-439. Snyder, J. M. (1989). Election goals and the allocation of campaign resources. Econometrica , 57(3), 637-

660.

Stem, R. M., & Bickers, K. N. (1994). Congressional elections and the pork barrel. Journal of Politics, 56(2), 377-399.

Strömberg, D . (2008). How the electoral college influences campaigns and policy: The probability of being Rorida. American Economic Review, 98(3), 769-807.

Ô Springer

This content downloaded from 140.211.65.60 on Sun, 18 Nov 2018 18:59:31 UTC All use subject to https://about.jstor.org/terms

Contents

p. [117]
p. 118
p. 119
p. 120
p. 121
p. 122
p. 123
p. 124
p. 125
p. 126
p. 127
p. 128
p. 129
p. 130
p. 131
p. 132
p. 133
p. 134
p. 135
p. 136
p. 137
p. 138
p. 139
p. 140
p. 141
p. 142
p. 143

Issue Table of Contents

Public Choice, Vol. 146, No. 1/2 (January 2011) pp. A1-A6, 1-267

Front Matter
COMMISSIONED EDITORIAL COMMENTARY

In memoriam: Melvin J. Hinich, 1939-2010 [pp. 1-8]

Social or political cleavages? A spatial analysis of the party system in post-authoritarian Chile [pp. 9-21]
Economic growth with endogenous corruption: an empirical study [pp. 23-41]
Public employment and income redistribution: causal evidence for Brazilian municipalities [pp. 43-73]
Does tenure in office affect regional growth? The role of public capital productivity [pp. 75-92]
Political regime change, economic liberalization and growth accelerations [pp. 93-115]
The road to power: partisan loyalty and the centralized provision of local infrastructure [pp. 117-143]
Loyalty and competence in public agencies [pp. 145-162]
Public sector efficiency: leveling the playing field between OECD countries [pp. 163-183]
How much income redistribution? An explanation based on vote-buying and corruption [pp. 185-203]
LITERATURE SURVEY

þÿ�þ�ÿ��P��o��s��i��t��i��v��e�� c��o��n��s��t��i��t��u��t��i��o��n��a��l�� e��c��o��n��o��m��i��c��s�� I��I��a�� s��u��r��v��e��y�� o��f�� r��e��c��e��n��t�� d��e��v��e��l��o��p��m��e��n��t��s�� [��p��p��.�� 2��0��5��-��2��5��6��]

BOOK REVIEWS

Review: untitled [pp. 257-259]
Review: untitled [pp. 261-263]
Review: untitled [pp. 265-267]

Back Matter