Provision of State and local goods

Farcus by David Waisglass Gordon Coulthart

CHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS—USER CHARGES

171

PRICING OF GOVERNMENT GOODS USER CHARGES

The economic case for the expansion and rationalization of pricing in the urban public sector rests essentially on the contribution it can make to allocative efficiency. Prices will provide correct signals to indicate the quantity and quality of things citizens desire. . . —SELMA J. MUSHKIN AND RICHARD M. BIRD

"GOLF AND SWIMMING ARE ACTIVITIES THAT ARE GOOD CANDIDATES FOR USER FEES. BOTH

REQUIRE PHYSICAL STRUCTURES THAT ARE EXPENSIVE TO MAINTAIN, BUT NOT ALL RESIDENTS OF

A COMMUNITY WILL. USE THESE FACILITIES. THE LOCAL GOVERNMENT CAN OFFSET SOME OF THE

MAINTENANCE COSTS BY ASSESSING 'A FEE FOR USE OF THE FACILITIES. OFTEN RESIDENTS OF A

COMMUNITY DO NOT WANT TO PAY FOR SERVICES THAT THEY DO NOT USE. ALTHOUGH GOLF AND

SWIMMING POOL FACILITIES ARGUABLY ENHANCE THE QUALITY OF LIFE IN A COMMUNITY, A

LOCAL GOVERNMENT THAT CHARGES USER FEES MORE FAIRLY DISTRIBUTES THE RESPONSIBILITY

FOR PAYMENT." (MOULDER, 2002, P. 4)

A SURVEY OF LOCAL GOVERNMENTS IN 2001 BY THE INTERNATIONAL CITY/COUNTY

MANAGEMENT ASSOCIATION SHOWED THAT NEARLY 34 PERCENT OF LOCALITIES PROVIDE GOLF

FACILITIES AND NEARLY 67 PERCENT REPORTED OWNING A SWIMMING POOL OR LEASING A

PRIVATE POOL FOR PUBLIC USE. OF THOSE LOCALITIES PROVIDING GOLF COURSES, 99 PERCENT

CHARGE USER FEES AND 78 PERCENT CHARGE RENT FOR PRIVATE USE OF THE FACILITIES. AMONG

THOSE LOCALITIES WITH SWIMMING POOLS, 93 PERCENT CHARGE USER FEES AND 79 PERCENT

COLLECT RENTAL FEES FOR PRIVATE USE. 2

E 0

ui

I "You know, of course, there's a toll."

FARCUS ® is a registered trademark of LaughingStock Licensing Inc. 0 LaughingStock Licensing Inc. All Rights Reserved.

User charges, which are prices governments charge for specific services or privileges to pay for all or part of the cost of providing those services, have always been important (although perhaps not as much as the Farcus cartoon suggests) but have become increasingly so in the past decade. User charge financing is different from financing through general taxes, because with tax financing no direct relationship exists between tax payment and service received. Common examples of user-charge financing in the state—local govern- ment arena include water charges, tuition at public colleges and universities, public hospital charges, parking fees, highway tolls, subway or bus charges, and park entrance fees.

8 CHAPTER

HEADLINES

TYPES AND USE OF CHARGES

1 1 "Public Prices: An Overview." In Public Price--; for Public Products, edited by Selma Mushkin, 11. Washington, D.C.:

The Urban Institute, 1972.

'Moulder, Evelina. Fmanerng Parks and Recreation: User Fees and Fund-Raising as Revenue Sources. Washington, D.C.: International City/Cotmty Management Association 2002.

The types of financing methods considered as user charges include direct charges for using a public facility or consuming a good or service, license taxes or fees paid for undertaking some activity (such as fishing license and driver license fees), and special assessments, a type of property tax levied for a specific service and based on some physical characteristic of the property, such as front

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

172

Table 8.i

Amounts oh Charges and Fee& State and Local Governments, 2002 4

173

State-local current

charges (nominal

and real dollars)

cHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS-USER CHARGES

-III- Charges, nominal -a- Charges, real 2002 dollars

200000

0 0

150000 -

100000

0 I i i 1 I i 1 1 1 t 1 1 I i 1 :11.

cot‘ cia° Ab cbb oP 41 ' c? 4° cjib c,,d5

Year

Figure 8. tb

Charges, nominal III Charges, real 2002 dollars

State-local current

charges, selected

years

1977 1982 1987 1992 1997 I 2002

Year

"8 150 -

0 - 1962 1967 1972

300

250

r. 200 f0

0 -o

C 0

CT3 100

50

Percent of Type Amount (billions of dollars) General Revenue Total User Charges $2531 9 15.0% 77.0% License Taxes and Fees 36.72 2.2 11.2 Special Assessments 4.78 0.3 1.5 Other Unallocable Taxes 34.08 2.0 10.4

Total 328.77 19.5 100.0

'This measure of charges excludes revenue from public utilities (electric, gas, water) and liquor stores. In most cases, these services are sold directly, so that user charges represent the bulk (more than 80 percent) of financing. See Netzer, 1992.

SOURCE: U.S. Department of Commerce, Governmental Finances: 2001-02 (2005).

Table 8.2

State-Local User Charges and Expenditures, by Category, 2002

Category User Charge as Percentage of All User Charges

Category User Charge as Percentage of Direct

Expenditures in Category 28.6% 12.2% 25.9 75.2 15.1 76.1 4.9 76.1 3.3 7.2 2.8 23.3

14.9

'For comparison, Netzer (1992) reports similar ratios in 1989 for water supply (81.8), electric power supply (93.7), and gas supply (103.1).

SOURCE: U.S. Department of Commerce, Governmental Finances: 2001-02 (2005).

footage (for example, assessments for sidewalk construction). State-local gov- ernments collected nearly $329 billion in these types of charges in 2002, with traditional user charges accounting for about 77 percent of the total, as shown in Table 8.1.

The magnitude of state-local traditional user charges, either in nominal or real dollars, clearly has risen substantially since the early 1960s and particularly fast since the mid-1980s (Figures 8.1a and 8.1b). Current charges increased faster than the general level of prices over this entire period, except for 1978-1980, and much faster than the general inflation rate since then. All charges and fees together rep- resented 19.5 percent of the general revenue of state-local governments in 2002 with traditional user charges alone representing 15 percent of revenue.

Most state-local user charges arise in the budget categories for education and hospitals. 3 As shown in Table 8.2, on average nearly 55 percent of all subnational government direct user charges are attributable to those categories. Of all other

3 When state or local government owns and operates public utilities (electric or natural gas) and liquor stores, the prices for those services also represent user charges. See Netzer, 1992.

individual categories, sewers and sanitation account for 15 percent of charges, the only one that is more than 10 percent. For that reason, extreme caution must be used when comparing user-charge use among different states (or localities). Without large public higher-education and hospital systems, user charges may appear as a small fraction of revenue simply because those services

Category

Education Hospitals Sewers and Sanitation Air Transportation Highways Parks and Recreation Other

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES CHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS-USER CHARGES

r 75 1 7 4

are not provided. Interjurisdictional comparisons should be made by budget category. 4

More than half of state-local expenditures on airports, hospitals, and sewer and sanitation systems are financed by user charges, whereas only about 12 percent of education expenditures are financed that way. Although tuition and other charges by public colleges and universities are a large fraction of total user charges, they represent a small fraction of total state-local education expenditures when the mostly tax-financed primary and secondary school expenditures are included. The opposite is true for airports, sewer and sanitation systems, and parks and recre- ation services, for which user charges are a small fraction of all charges but repre- sent a large fraction of spending in those categories.

By 2002, user charges (broadly defined) accounted for more than 14 cents for every dollar of state revenue and 18 cents per dollar for local governments (includ- ing nearly 23 cents per dollar for cities). Changes in the pattern of user-charge reliance since 1962 for different types of subnational governments are depicted in Table 8.3. Reliance on user charges, license fees, and special assessments together has increased for both state and local governments since the 1970s, compared to decreases in reliance on those charges in the 1962-1977 period. Among local gov- ernments, county governments have continually increased reliance on user charges although municipalities and townships followed the general pattern of decreasing reliance until the late 1970s and then subsequently increasing. Charges provide a small (and decreasing) fraction of revenue for school districts and a rel- atively large fraction (more than 40 percent) for special districts. Focusing on direct user charges alone, reliance by states and localities changed little from 1962 to the late 1970s and increased substantially since. Clearly, total user charges of state-local governments, whether broadly or narrowly defined, have increased faster than other revenues since the late 1970s.

THEORY OF USER CHARGES

In theory, user charges should operate as benefit taxes (discussed in Chapter 2), with an individual's charge depending both on the benefit (use) and cost of the provision. The principal rule for economic efficiency requires that marginal bene- fit equal marginal cost. For services that primarily benefit the direct consumer, the price charged should equal marginal cost.

The principal reason this makes economic sense is simple. If consumers believe that public services and facilities are "free"-that is, that more can be produced at no cost to the consumer (when in fact additional amounts do entail a production

'In fact, user charges as a percentage of general revenue are largest for states with relatively lower per-capita revenue and smallest for high-revenue states. The five states with the largest fraction of revenue from user charges are Alabama (25.2 percent), South Carolina (21.56 percent), Mississippi (20.1 percent), Utah (20.0 percent), and Idaho (19.7 percent), whereas the five with the smallest user-charge ratios are Connecticut (7.7 percent), Rhode Island (8.4 percent), Alaska (9.6 percent), Massachusetts (9.8 percent), and Maine (9.9 percent). Similarly, Netzer (1992) reports that increases in state per-capita incomes are associated with decreased reliance on user charges.

User Charges and User-Associated Taxes as a Percent ob General

Revenue, by Level oil Government, Various years

Year 1962 Charges

States

7.1%

All

10.6%

Counties

Local Governments

Municipalities Townships School Districts Special Districts

Alla 16.3 13.7 11.9 17.8 7.8 6.6 482

1972 Charges 7.9 10.5

All 14.0 12.7 13.4 15.3 7.7 5.7 44.3

1982 Charges 6.4 11.4 All 11.7 14.7 16.6 17.3 9.3 4.9 38.1

1987 Charges 7.6 13.2 All 11.9 15.6 17.2 19.2 10.9 4.5 38.4

1992 Charges 8.7 14.6 All 12.6 16.8 18.5 20.9 12.1 4.6 42.8

1997 Charges 8.9 15.9 All 12.7 18.1 21.0 23.1 12.5 4.4 42.9

2002 Charges 9.4 15.4 All 14.3 17.7 20.7 22.8 13.2 4.3 41.6

'Charges plus license taxes and fees plus special assessments plus other unallocable taxes.

SOURCE: Census of Governments, Compendium of Government Finances, various years.

cost)-consumers will be induced to demand more than the efficient amount of those services or facilities. One function of user charges is to make consumers face the true costs of their consumption decisions and create an incentive for efficient choice. 5

The basic idea of that choice is illustrated by Figure 8.2, which depicts the mar- ginal benefit schedules for both direct users (MBa) of a service or facility and all of society (MBs), who also benefit generally. Those marginal benefits are added together to determine the aggregate marginal benefit to the entire society or com- munity from an additional unit of the service (E MB). Given a cost of producing one more unit equal to MC, the efficient amount of the service or facility is Q*. The private marginal benefits to users and general marginal benefits to all at that quan- tity determine how the production costs should be divided among users (a user charge) and all of society (general taxes). In this case, user charges should account for MB u*/ MC of the facility's cost. Because direct users face a marginal cost of MB u*, they demand quantity Q*, which is the efficient quantity.

'You may want to review the section on public goods and benefit taxation in Chapter 2.

IMB ; = MBu + MBs

Marginal cost MC

MB

MB

Quantity

Figure 8.2

Allocation of costs

to direct users and

society in general

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

Price

($)

In contrast, if users perceive the marginal cost to be zero, they demand amount Q2. This is not efficient because the marginal benefits to everyone—the sum of the marginal benefits to direct users and to society generally—are less than the cost of producing all the units of output between Q* and Q2. That difference between mar- ginal cost and aggregate marginal benefit represents the potential efficiency cost of not charging appropriate prices for this service.

Several general principles of efficient user charges follow from this analysis:

t. The greater the share of marginal benefits that accrues to direct users, the more attractive user-charge financing becomes.

2. User-charge financing requires that direct users can be easily identified and excluded (at reasonable cost) from consuming the service unless the charge is paid, assuming that most of the benefits of a service or facility go to direct users.

3. The efficiency case for user-charge financing is stronger, the more price elastic is demand. In the special case of a perfectly inelastic (vertical) demand, price does not matter. No inefficiency results if consumers underestimate cost. Obviously, the more price elastic demand is, the greater the potential for inefficiency if consumers do not face true costs.

4. Marginal benefits, not total benefits, matter for determining user charges. For instance, in Figure 8.2, quantities of the facility beyond Q i provide benefits only to direct users. Thus, despite the fact that all of society benefits some from this facility, direct users should finance entirely the production of amounts greater than Qi.

This last principle deserves additional explanation. The general rule is that costs should be allocated in proportion to benefits, but the question is which

,TER EIGHT ■ PRICING OF GOVERNMENT GOODS—USER CHARGES

costs. 7 Here it is helpful to distinguish between the capital costs for the amount of

a service or facility to provide—the long-run production decision—and the oper- a- ting costs associated with the use of a given facility—a short-run decision. For it.,starice, a local community faces a decision about the appropriate number and size of parks to provide, whereas a state government selects the number and size of public colleges. After a given amount of those facilities is provided, how-

ever, each government also faces a choice about how much and by whom those facilities are to be used. Should park use be free or should there be an entrance

charge? Should the charge be different for residents and nonresidents? Should the charge be different at different times? Similar questions apply to college tuition. User charges can have a role to play in the decisions both about amount

and use. 6

Allocating Access (Capital) Costs

The costs for constructing or acquiring a public facility should be paid by those groups in society who will benefit from the existence of the facility, which may be different from those who benefit from using the facility directly. For instance, an individual who may never drive a car still benefits from roads as a result of trans- portation of goods and other people. Individuals may benefit from a facility, even if they do not use it directly, in three main ways. First, the facility's existence pro- vides individuals the option of using it in the future, should their demands change. An individual may not use a particular bridge currently, but may move in the future to an area that would require using the bridge to commute. Such option value is particularly significant when it would be very costly, or even impossible, to pro- vide the public facility in the future. Such might be the case with public parks that preserve land in a relatively undeveloped state (as it might be impossible to reverse development after it has occurred).

Second, individuals who are not direct users also might benefit if the facility generates spillovers in the form of additional economic activity. Such monetary benefits to nonusers associated with public facilities or services might include spending on private services that complement public services (a private bait shop near a public park), attracting funds from other jurisdictions (tourism), or improv- ing the environment and attracting workers (which increases the supply of labor and holds labor costs down).

Third, nonusers might benefit from pure altruism, receiving personal benefits from providing service to others. If all residents of a jurisdiction benefit from the existence of or access to a public service or facility, then all residents should con- tribute toward acquiring that facility based on those general benefits, which are independent of use.

qt may not be practical to separate long-run and short-run pricing decisions, however, because that might require prices to change substantially over time. If price is set equal to short-run marginal cost, higher prices are called for as demand rises. Because those prices will be greater than long-run average cost, funds will be provided to finance the desired capital expansion. But after the facility is enlarged, marginal costs, and thus prices, will fall again. Some type of average cost pricing would maintain more price stability but would be inefficient. Price sta- bility in itself might be desirable, however.

176 177

Demand y Demandz

Z Qx

Quantity

CHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS—USER CHARGES

1 79

Efficient user

charges with

and without

congestion

SRMC

Price

($)

(a) (b)

price ($)

SRMC

Demand y •

Quantity

00

Q w

178

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

If all residents as well as users should pay all or part of the long-run produc- tion costs of public facilities, these charges should be independent of the amount of actual use of the facility. These charges therefore might be flat per-capita or per-household charges, or perhaps charges based on property size if long-run capital costs vary by size. Examples include a fixed-service charge common in public water systems to cover the capital costs (pipes, pumps, storage) and special assessments for sidewalks, street lights, and neighborhood parks. Of course, these access costs might also be covered from general tax revenues if everyone benefits equally from the existence of the facility.

Allocating Use (Operating) Costs After a public facility—whether a park, road, water system, or college—has been provided, attention must turn to covering the variable or operating costs. How this is accomplished determines how much and by whom the facility is to be used. The general principle of efficiency, again, is that marginal benefit should equal mar- ginal cost, but now the relevant marginal cost is short-run marginal cost, the cost of accommodating an additional consumer or providing another unit given the capital input selected. At issue here is the appropriate charge for each gallon of water consumed or for each admittance to the park.

Operating costs should be allocated based on marginal benefit from use. In many and perhaps most cases, the benefit from additional use (as opposed to existence) goes only to users. If so, then it may be appropriate to charge fees to users to cover all the marginal operating costs. In some cases, however, external (nonuser) bene- fits may be associated with the additional use of a facility or service, such as the benefit to society from having an additional person educated. In those cases, only that portion corresponding to their share of marginal benefits from use should be charged to direct users.

Assuming that users are to pay all the operating costs, some possibilities are illustrated in Figure 8.3. In Figure 8.3a, the short-run marginal cost is positive and constant; each additional unit of service imposes a constant additional variable cost of ao . Demand for this service is represented by Demandy . The appropriate use charge (if users can be identified and excluded) is ao dollars per unit consumed— for example, $.O1 per gallon of water or $2 per car for admittance to a park. In Fig- ure 8.3b, marginal cost is zero up to quantity Qc. If demand is Demandy, then the correct user charge is zero; there are no operating costs at the margin to cover.

Allocating Congestion Coats For some services, an additional consumer may impose extra costs on other users, called congestion costs. As roads and bridges become more crowded, traffic slows and the (time) costs to all users increase; as parks become more crowded, there is less space for those in the park to enjoy activities; and when all the parking spaces and tennis courts are occupied, other potential users incur a waiting cost (or must forego the activity). Because the government incurs no additional cost by provid- ing a service to an additional consumer (if one additional car parks in a space or

an additional couple uses a tennis court), the government does not need to collect more revenue for operating expenses. Yet, governments should and sometimes do charge user fees for all these services. The purpose of use fees in those situations is to allocate a scarce resource among competing demands.

This economic notion of congestion is shown in Figure 8.3b. For quantities of use or service less than Qc, additional consumers can be accommodated without impos- ing any costs on other users. In essence, the facility is not yet "crowded." Because marginal cost (operating and congestion) is zero, the efficient price is also zero; no use fee is required. If demand for the service is Demandy, no use fee should be charged, with capital costs covered either out of general taxes or by some fixed charge as dis- cussed earlier. For quantities of use or service above Qc, the facility starts to become crowded; additional consumers impose congestion costs on other users (at an increasing rate in Figure 8.3b). Therefore, if demand for this service is Demandz, the

appropriate use fee is co, with a resulting amount of use equal to Qz . If no use fee is

charged, the amount of use is Q x, and the facility is overused, that is, "too crowded." 7 Correcting for congestion costs may require charging different fees at different

times. For the service represented in Figure 8.3, demand might sometimes be Demandy, requiring an efficient use charge of zero, and sometimes Demandz, when

the efficient use charge is c o. For instance, parks may be crowded on weekends and not during the week, demand for bridge crossings may be great at the com- muting hours and low at other times, or public-transit facilities may be used exten- sively at rush hour and little at other times. In other words, demand may be dif- ferent at peak times (when higher use fees are appropriate) than at off-peak times

(when lower or even zero use fees may be appropriate). 8

'At Qv the marginal benefit to the last user is c o, equal to the marginal cost that user imposes on all other users.

'Examples of congestion pricing for roads and highways are discussed in Chapter 20.

Figure 8.4

Two-part pricing

by a natural

monopoly

Price

($) A

Q3 QI

Average cost

— Marginal cost

Demand

Quantity

P1 =AC

P2 = MC

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES "TER EIGHT • PRICING OF GOVERNMENT GOODS—USER CHARGES

1 8o

ger and larger output, the decreasing average fixed cost (combined with rela- 'vely small marginal costs) causes average total cost to decrease as well. Average

to tal cost always decreases as output rises and marginal cost is always less than

average cost. This leads naturally to monopoly because any given output can be

produced at lower average cost by one large firm than by several smaller firms. This cost situation is believed to characterize most utilities, including electric-

ity natural gas, and perhaps water and sewer and public-transit services, but also may apply to many other facilities provided by state and local governments, including parks and beaches, roads, bridges, airports, and others. In all these cases,

fixed (capital) costs are large relative to variable (operating) costs, and the facility has some unique aspect that generates monopoly power. Therefore, the usual

approaches are either for the government to grant a private firm monopoly rights to a given market and then regulate the prices the firm may charge or for the government to become the producer directly.

Setting an efficient price or use charge when there are increasing returns to scale faces an inherent conflict. If price is set equal to marginal cost (for instance, at P2), then price is less than average cost, and the firm or operating authority cannot cover all its costs. If price is set equal to average cost, at P i, so that the firm or pub- lic authority can cover both operating and fixed costs, then price is greater than marginal cost, causing the capital facility to be used less than efficiently. It is some- times suggested that the government should set or enforce a price equal to mar- ginal cost, with the government using general tax revenues to cover the resulting financial losses. That could be an appropriate user charge strategy if everyone were to contribute toward the capital costs, with users covering only operating costs. But that, too, leads to inefficiency because of the inherent inefficiencies cre- ated by the taxes necessary to offset the entity's operating losses. In practice, utility regulation often settles on solutions that effectively set charges equal to average cost—for example, at P i—allowing the regulated utility just to cover costs and earn average profits.

One possible and often practical solution to this difficulty is to set a two-part price, charging different prices for different quantities of the service. For example, in Figure 8.4, one could set a price of P3 for quantities up to Q3 and a price of P2 for amounts greater than Q3. Because the charge for marginal units of output is equal to marginal cost, total consumption equals the efficient amount of Q2. But the producer may be able to avoid operating losses because the price discrimination generates larger revenue than if a single price is charged. With the two-part price, revenue is P3Q3 + P2(Q2 – Q3) or (P3 — P2)(23 + P2Q2, which is greater than the revenue from a single price, equal to P2Q2.

Two-part pricing takes advantage of the fact that some consumers are willing to pay prices higher than marginal cost for so-called inframarginal units (units other than the last one purchased). The two-part price captures some of that con- sumers' surplus for the producer, allowing the producer to charge a marginal cost price for marginal units and still cover all costs. With a single price of P2, con-

sumers enjoy a surplus represented by the area of triangle ACE. With the two-part price involving P2 and P3, consumers' surplus is smaller, represented by the areas

Obviously, efficiently applying use charges could generate revenue, which is not necessary to cover extra operating expenses. In Figure 8.4b, a congestion charge at price co during peak demand time generates net revenue because the marginal operating cost is zero. This is one of the advantages of user charges, that they measure the real demand for new facilities and provide the resources to cre- ate those new facilities. If current users are paying the appropriate costs of their consumption (including congestion costs) and excess demand still exists (evi- dence of serious congestion), then the facility is too small and consumers would pay, and are paying, to expand the facility. The revenue above operating costs, which was paid by the peak-time users, can be used to expand the facility or cre- ate another one.

Potentially, therefore, user charges can be composed of three separate parts: (1) an access charge to cover all or part of capital costs, (2) a use fee to cover all or part of the operating costs to the government associated with use, and (3) a con- gestion charge to cover the costs imposed by an additional user on other users. An alternative but equivalent way to think of determining user charges is to con- sider how a single producer should set its price based on its costs, including both its fixed and variable costs. That is the approach that follows for a natural monopoly.

User Charges with Natural Monopoly A natural monopoly exists if the production of a good or service exhibits increas- ing returns to scale, so that the long-run average cost continually decreases as output increases, as depicted in Figure 8.4. This situation often arises when capital or fixed costs are large relative to variable costs. As the fixed cost is spread over a

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES CHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS—USER CHARGES

1 83

182

of triangles ABF plus BCD. Rectangle BDEF represents the added revenue to the producer.

Of course, there is no reason why the inframarginal price needs to be set at P3; that price must be selected to generate enough extra revenue to cover the producer's operating losses, if possible. In fact, the inframarginal price could apply only to the first unit consumed, effectively serving as a type of cover or access charge. In the case of Figure 8.4, that could entail charging a price of $A for the first unit and a price of P2 for all subsequent units. This is equivalent to charging a set access fee to cover capital costs and then a use fee to cover operating costs.

Two-part prices of this type already are common, but this technique probably could be expanded with other user charges. Many public water systems charge a fixed monthly access charge as well as a per-gallon use fee. The access charge is a second price effectively imposed on the first gallon of water consumed and covers the fixed costs. Some public-transportation systems sell passes that allow riders to pay a lower fee for each ride than paid by consumers without the pass. Those who purchase the pass effectively pay a high price for the first ride in each period (the inframarginal ride) and a low price (usually zero) for all subsequent rides (the marginal ones). It is not hard to think of other potential user charge applications of this type. A public refuse system might levy a fixed monthly access charge in addition to a small fee per unit or bag collected, or a public parking facil- ity could offer lower hourly parking charges to individuals who have purchased a monthly pass.

Two-part prices also can make sense even if the marginal cost of an additional user is zero. In that case, the first price covers the fixed costs, and the second price (the marginal use fee) is zero. Some private amusement parks (including the Disney parks) price this way, charging a single admission fee and no extra charge for each ride. Similarly, the Michigan Department of Natural Resources allows users to purchase an annual vehicle pass for Michigan state parks for $24, which entitles that vehicle to unlimited admittances without further charge to all state parks for that year. Those who purchase the pass therefore pay $24 for the first admittance in a year and a zero price for all others. Without such a pass, each vehicle admittance costs $6. 9

It is relatively easy to add a congestion charge to the two-part price when appropriate. For instance, for those who purchase a monthly pass for $20, a public- transit system might charge $.50 per ride during off-peak times (compared to $1 for others) and $1 during peak periods (compared to $1.50). Or a park system might offer an annual pass for $20 that permits free use of the park on week- days, although weekend use entails an additional $4 fee. In both cases, the addi- tional marginal price during high-demand periods represents the marginal congestion cost.

Application 8.i

At one university, the school's policy was to

not charge students and faculty members any

fee for using the tennis courts, the argument

being that using the university's facilities

should be "free" to those who already paid

tuition or worked for the school. Because this

university is in a northern city and the tennis

courts are outside, this policy posed no prob-

lem for half of the academic year. In the fall

and especially in the spring, however, there

was substantial excess demand for the tennis

courts; waits of 30-60 minutes for a court

were common. The courts were therefore not

"free" but were allocated by having people

wait. Presumably, those who had the lowest-

valued time ended up using the tennis courts

more. This university had no summer session

and so made their tennis courts available

to the general public during the summer

PRICING AT CONGESTED TENNIS COURTS

months. The difference was that a use fee

was charged in the summer to everyone—

students who remained in the town, faculty,

and the public. Not surprisingly, there were

many vacant courts during the summer.

This situation is represented in Figure 8.5.

The supply of tennis courts is fixed at Qc, so

marginal cost is zero for quantities less than

Qc (vacant courts mean there is no cost to

accommodate another player); however, mar-

ginal cost becomes very high after all the

courts are in use (the cost of accommodating

another user is the cost of building another

court). When demand is Demandy, as during

the summer in the preceding story, there is

excess capacity, and no fee should be

charged. When demand is at the peak level of

Demandz, a use fee equal to Cz would gener-

ate efficient use—only those who are willing

Supply

Efficient prki

when supply is fixed

Number of tennis courts or parking spaces

9Even with the pass, there is a daily charge for camping in the park, rather than just a visit.

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

184

CHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS-USER CHARGES

1 85

to pay Cz, that is, those who get Cz dollars

worth of benefit from using the tennis court

would play. With the fee, no excess demand

exists for the facility. By charging a fee during

the summer when demand was low (when

the university was not in session) in an

attempt to generate revenue from the public,

the university's facilities were wasted from

society's viewpoint. By not charging a fee dur-

ing the spring when demand was high, the

university made an implicit decision to allo-

cate the scarce tennis courts by having peo-

ple wait—what is often called "first come, first

served." Although allocation either by fees or

waiting gives some consumers an advantage

over others, under the first come-first served

system, potential tennis players do not know

the charge (the required amount of time to

wait) until they arrive at the courts.

One might believe that the proper policy

for the university in this case is to build more

tennis courts because the excess demand

during the spring and fall suggests that more

are "needed." That analysis is faulty, however,

because the excess demand occurs only

because tennis court use appears to be "free."

Although the extra courts would be used dur-

ing the peak times, they would enlarge the

excess capacity that exists during the off-

peak time. As the discussion about the effi-

cient amount of public facilities showed,

more tennis courts should be built only if

those who demand the courts are willing to

pay the full cost of constructing them

(assuming that extra tennis courts benefit

only direct users). Interestingly, use fees pro-

vide a test of that hypothesis. The efficient

use fee Cz in Figure 8.5 equals the marginal

benefit of a tennis court to users. If the fee

that equates supply and demand turns out to

be large enough to finance another court—

that is, if marginal benefit is greater than mar-

ginal cost—then another court should (and

can) be built.

Application 8.1 — Pricing at Congested Tennis Courts

PARKING FEES AND PARKING METERS

Here is a do-it-yourself application that you

can use to see how well you understand user

charges and congestion. The analysis and dia-

gram of Application 8.1 can be applied to the

question of parking meters. Consider the fol-

lowing argument:

The streets have been paid for and belong

to the people. Therefore, parking meters

should be abolished.

Analyze that position and prescription in

light of the preceding discussion. If there

were no parking meters or fees, how do you

think the available parking spaces would be

allocated? Would that allocation system be

better? For whom? Without parking fees, do

you think people would perceive that there is

more or less of a "parking problem"? If parking

fees are to be used, should they vary by loca-

tion? Time of day? Time of year?

Application 8.2

Other User-Charge Issues The theoretical discussion suggests that user charges are most appropriate when most of the benefits of a government service go to identifiable direct consumers whose demand shows some price elasticity. Two other poteltial advantages of user charges should be noted as well. Many public services provide benefits to individ-

uals who are not residents of the providing jurisdiction, and user charges are one way for those nonresidents to pay for the benefits they enjoy. Having users directly finance (at least partly) the services and facilities from which they bene- fit may portray a type of fairness in public policy that results in more public

acceptance of state and local government provision of certain services. Several potential problems with user charges should be addressed as well.

Sometimes objection is raised to user fees on the grounds that they are a disad- vantage for consumers with lower incomes. That notion is often coupled with the statement that general taxes, in contrast, are based on "ability to pay." The pre- sumption of such an argument is that it is not fair to base consumption of the gov- ernment service in question on income or "willingness to pay," as is done when fol- lowing the "benefit principle" of public finance.

It is certainly true that allocating any good or service by money prices gives an advantage to those consumers with more money. But because that point is general, the relevant issue is why a particular government service that mostly benefits direct users should be treated differently than privately provided goods and ser- vices such as medical care or Mercedes automobiles. One possible explanation is that the service is a means of redistributing income, which is one of the funda- mental economic roles of government. This is undoubtedly part of the reason pri- mary and secondary education is financed almost entirely from taxes. Education provides external benefits to all of society; one of these benefits is a means of im- proving the economic conditions of the poor. One needs to be careful not to carry this argument too far, however. It is not clear that free use of public golf courses, for instance, is a very effective way of assisting the poor.

Avoiding user charges also may be an inefficient way of helping the poor. Some state—local government services are consumed much more by higher-income con- sumers than lower-income ones. Avoiding user-charge financing in those cases to assist lower-income consumers may actually benefit higher-income consumers to a greater degree. It could be more efficient for the government to charge everyone the user charge and give direct assistance of some type or a specific subsidy to the lower-income consumers affected by the charge.

A second potential problem with user charges is that the administration costs (to the government) and compliance costs (to the consumers) of collecting the charge are, in some cases, large enough to offset any expected efficiency gains from user-charge, as opposed to tax, financing. Typically, administration costs include the costs of measuring use, billing users, and collecting the fee, whereas compliance costs include delay at road or bridge toll booths and the time and postage costs of making the required payments. Besides the other necessary con- ditions, therefore, user-charge financing is attractive only if the charge can be collected at a reasonable cost. (For instance, the advantages and disadvantages

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES CHAPTER EIGHT ■ PRICING OF GOVERNMENT GOODS—USER CHARGES

186 187

of alternative ways of administering highway user charges are discussed in Chapter 20.)

APPLICATION OF USER CHARGES

The application of user charges to five specific state–local government services— public higher education, K-12 education, water and sewer service, refuse collection, and parks—is discussed in this section. Discussion of transportation- related user charges is presented in Chapter 20, and charges to offset the infrastructure costs associated with economic growth and development are covered in Chapter 22.

Financing Public Higher Education"

For most readers of this book, and especially for those attending public colleges and universities, tuition is the best known of all subnational government user charges and the one with the most immediate personal as well as academic impli- cations. In the United States, tuition historically covered between 30 and 50 per- cent of the expenditures of public colleges and universities, with the remainder financed mostly from state (and for community colleges, sometimes local) taxes. In recent years, the share financed by tuition has grown and is now commonly more than 50 percent. There is substantial variation among states in the reliance on tuition, however; for many years, in fact, some states provided "free" college edu- cation to qualified residents. This naturally leads to a question of whether public college students should pay a larger (or smaller) fraction of the cost of their college education and how those charges should be structured.

Those who argue that tuition (or other user charges) should be more important in financing public higher education usually suggest that most of the benefits of that education are captured directly by the students in the form of higher incomes, jobs with more prestige, and information that assists those individuals in all aspects of their lives. Moreover, those beneficiaries are directly identified, the charge can be collected at low cost (indeed at zero extra cost after any tuition is levied), and students easily can be prevented from consuming the service unless they pay the charge. With that viewpoint, higher education seems to meet all the tests for substantial user-charge financing. However, at least four issues suggest that this view is incomplete.

First, public institutions of higher education usually produce research and pub- lic service in addition to education of students, although those three outputs are clearly not independent. Even if one believes that all the benefits of the education component of output are captured by students, the research and public service components of output benefit all of society and are thus appropriately financed by the government. Pure scientific research is usually identified as a classic public good; discoveries, once made, can be used by anyone at zero marginal cost to

ETor discussion and evidence concerning these issues for specific states, see Lee Hansen and Burton Weisbrod (1969) and John Goddeeris (1982).

society. Research and public service should therefore be financed by the general society and not directly by students. 11

It remains to be determined what fraction of public higher-education output is research and public service compared to education, a fraction that undoubtedly differs by type of institution. In major state universities, research and public ser- vice usually represent at least half of a faculty member's job and similarly at least half of the university's output. The output of community colleges, in contrast, is usually almost entirely education. It follows therefore that the appropriate degree of user-charge (tuition) reliance might be greater for institutions primarily produc- ing education as opposed to those producing education and research. Indeed, pub- lic subsidies to four-year colleges generally are greater than to two-year colleges.

Second, students already bear a larger fraction of the social cost of public higher education than it appears from comparing tuition and state appropriations. A hypothetical but illustrative computation of both the social and private cost of public higher education for one student is shown in Example 8.1.

[ Per-Student Economic Costa off Higher Education

Category Social Cost ($) Student Cost ($) Percentage

Instruction $ 16,000 $ 8,000 50% Books, supplies, transportation 1,500 1,500 100

Foregone income 15,000 15,000 100

Total 32,500 24,500 75.4

The cost of instruction, which is essentially the college or university expendi- tures per student, is assumed to be $16,000, of which one-half is covered by student tuition. The cost of books, supplies, and transportation represents expenditures on these items that are greater than if the student did not attend college. These costs are therefore true opportunity costs of choosing to attend college. Similarly, fore- gone income represents the difference between the income the student could have earned if not attending college and actual income earned. In the example, $15,000 is the approximate annual earnings for a full-time employee paid $7 per hour, slightly above the minimum wage. This foregone income is a true social cost, in addition to a cost to the student, because society gives up the goods and services that this individual's work would have produced, the value of which can be estimated by the factor payment. 12 In the example, then, students bear more than

'Tor that reason, it is often argued that research should be largely financed by the federal government. And the fed-

eral government, through such entities as the National Science Foundation, the National Institute for Education, and the National Institutes for Health, does substantially support university research.

'This computation is different than the out-of-pocket budget students usually consider; for instance, costs of room and board are not included. Because some room and board costs are incurred regardless of whether the individ- ual attends college, those costs would be included in the economic cost computation only to the extent that they

are larger because of college attendance.

Example 8. 1

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

188 CHAPTER EIGHT ■

PRICING OF GOVERNMENT GOODS—USER CHARGES

189

75 percent of the social cost of their public college education, not the 50 percent that appears from comparing tuition to college operating expenses. 13

Third, even if, after considering these two issues, a state desires greater reliance on tuition for financing public higher education, it may be difficult for that indi- vidual state to act unilaterally. Potential college students can change their states of residence toward those states that rely on tuition less and away from those states that rely on tuition more. Moreover, with an increasingly mobile society, the social benefits from higher education are not likely to be confined to any given state.

Fourth, the cost to the university of adding another student or having a student take more classes—the marginal cost—might be close to zero, at least for some lim- ited number of additional students. If the university is not crowded—that is, if another student can be accommodated without reducing the education provided to other students—then it is inefficient to charge a positive price at the margin. Of course, the solution to this problem might be a two-part price, charging a fixed tuition per year or semester and a lower or zero charge for each class or credit taken. The fixed tuition could cover the fixed costs of the university without creat- ing a disincentive for students to take additional academic work.

The argument usually raised against increased reliance on tuition—that it would prevent many lower-income students from attending college—also may be faulty. The evidence shows that college students, including public college students, tend to be mostly from higher-income (above the median) families; at least, the fraction of students attending college increases with family income. To maintain low reliance on tuition for all students, then, provides substantial benefits to many stu- dents who clearly are not poor.

If equity is the concern, an alternative to low reliance on tuition, and an alterna- tive to low reliance on user charges generally, is targeted assistance to lower- income consumers. Of course, this is already done in higher education. The state or university can set tuition at a level that seems efficient given the perceived social benefits and costs, and lower-income students can then be assisted with some type of income-based financial aid. This method has the potential to be a more efficient way of improving equity because society decides which consumers require and deserve assistance and provides aid based on those criteria. 14

Financing K - 12 Education's Among all state and local governments, local school districts rely on user charges least. As shown previously in Table 8.3, user charges account for only about 4 per- cent of school district general revenue currently, a share that has actually declined in the past 40 years at the same time that user charge reliance has grown for all

The results in the example are very similar to the results reported by Hansen and Weisbrod (1969) for the California state college and university system.

"It is also sometimes argued that lower-income students face a problem in financing higher education because the

capital markets do not work properly; if these students will indeed earn higher incomes due to education, then

financial institutions should be willing to make loans against those future earnings. If financial institutions will not, then an appropriate solution is government-sponsored education loans.

15See Robert Wassmer and Ronald Fisher, 2002.

other governments. Nearly two-thirds of school user charges arise from prices or

fees for school lunches. Activity fees, which represent about 20 percent of the total,

are the other major category of school charges. This relatively minor reliance on user fees by school districts raises the question

of whether K-12 schools might be able to increase use of charges either to supple-

ment revenue or to permit tax reduction. Work by Robert Wassmer and the author

suggests that the strongest case for increased use of user fees by public schools is for providing auxiliary services—such things as meals, transportation, after-school

care, medical care, adult education, and perhaps certain clubs or special activities— that are not necessarily part of the standard curriculum. These types of services tend to provide substantial private benefits, may be consumed by only a fraction of stu- dents in a school, have close private substitutes, and have relatively low-cost col- lection mechanisms available. By one measure, schools spent more than $30 billion on these types of services in 1992, amounting to about 13 percent of expenditure.

So, why don't public schools use charges and fees more? Or equivalently, what fac- tors have affected the choice of fees by schools in the past? One factor is obvious, as school user fees are prohibited or limited by state law in some states (although that fact just raises the parallel question of why some states have adopted these limits). User fees tend to be used more by schools in large districts where there may be greater variation in the types of services that are used by different students. User fees also are used to a greater degree by districts that are constrained by property tax lim- its, as those limits effectively force districts to find alternative revenue sources. User fees also seem to be relied upon less in states that exhibit a liberal or left-leaning polit- ical bias. This last point may arise because individuals with those political beliefs may be more concerned about the equity implications of charging all students, regardless of family income, prices for some school-related services.

Financing Water and Sewer Servicea l6 Water use fees are common, whether water service is provided by a local govern- ment or by a privately operated water utility company. These fees actually com- prise, either explicitly or implicitly, three separate charges: a connection charge, a capital and distribution charge, and a water-supply charge. The water-supply charge is intended to cover the marginal cost of additional gallons of water and therefore ideally should be based on the amount of water used. Use is sometimes approximated by the number of water outlets per structure or by the number of persons per structure, but it is far more common for the actual number of gallons of water consumed to be measured by a water meter. Assuming that marginal cost per gallon of water is constant, which appears reasonable for all but some special industrial users, a use fee can be computed from the measured usage and the appropriate constant per gallon charge.

The capital and distribution charge is usually a fixed charge, which may depend on the location or size (front footage) of the structure served. A charge based on front footage is intended to represent the extra cost of the water-supply pipe, as

"For a more comprehensive discussion of these issues, see Paul Downing and Thomas DiLorenzo (1981).

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

190

distribution costs depend on user density. It is sometimes argued that these distribu- tion charges also should depend on distance from the supply source, although apply- ing that concept is problematic. The location of the supply source, the water-treatment plant, is not fixed but is selected by the government. In fact, that location may be changed after many consumers have selected their locations. In addition, although a new and isolated development far from the supply source entails large extra costs of service for running new supply lines, a new development next to an existing one requires only extension of the water line (unless an entirely new and larger supply line was required). In practice, this charge is most often a fixed, front-footage charge.

Water users also are often charged for the direct costs of hook-up to the water system. This one-time connection charge may depend on the length of pipe required or it may be a flat charge reflecting the large fixed costs to the utility.

Analysis of potential user charges for sewer services is essentially similar to that for water (sewer disposal is a result of indoor water consumption); costs depend on the amount and type of sewage disposed and on the size and location of the struc- ture. However, actual metering of sewer discharge is not common, except for certain industrial users. Apparently, sewage flow meters are relatively expensive compared to water meters. The usual approach, particularly for residential users, is to assume that sewage flow is some percentage of water consumption and to compute a sewer- use fee from that number of gallons and a per-gallon charge. Of course, there is no reason for the sewer per-gallon charge to be the same as the water per-gallon charge. This method does not allow for variation among users concerning the purpose for water consumption, but it may still be the best option given the measuring costs.

These charges usually are collected from consumers through monthly or quar- terly billings, much the same as electricity, natural gas, or telephone bills. The water and sewer charges are usually on one bill, and the capital/distribution charges may be combined into a single amount per front foot, paid through a monthly service charge, or included in the gallonage charge.

Randolf Martin and Ronald Wilder (1992) estimated the potential effects on res- idential water use from user charge pricing. Using monthly household data from Columbia, South Carolina in the 1980s, they estimated the price elasticity of water demand to be between –.3 and –.6 when price was measured by the per gallon charge. If price was measured by the average total water bill per unit of water, the estimated elasticity was between –.5 and –.7. Therefore, demand for water is price inelastic, so that residential consumers do reduce use when user charges are applied. Noting the relative magnitudes of the estimated elasticities, the authors conclude the results are "consistent with the notion that households tend to respond to the total water and sewer service bill, rather than to the marginal price alone" (Martin and Wilder, 1992, p. 100).

Financing Rebuse - Collection Services

The costs of collecting refuse arise from both collection and disposal. Disposal costs depend on the amount and type of refuse and should include the cost of any environmental damage resulting from the disposal. A user charge to cover these disposal costs should therefore be a unit charge that varies by type of unit (the

CH APTER EIGHT ■ PRICING OF GOVERNMENT GOODS-USER CHARGES

191

disposal cost of a pound of household garbage is different from that of a pound of

used nuclear fuel). One difficulty in applying such a use fee is in measuring the amount of refuse. Possible measures include the number of specific-size cans col- lected or the weight of refuse collected. The first measure suffers because different

amounts of garbage may be packed into a fixed-size container, and both entail sub- stantial administrative costs to make and record the measurement. Even if those problems could be overcome, individuals would have an incentive to deposit their refuse at a neighbor's location, which gives rise to all sorts of silly notions about enforcement and neighborhood wars.

One innovative solution to this measurement problem, used in many localities, is to require that all refuse be deposited in specific bags sold only by the local gov- ernment. Typically, the bags are delineated by unusual colors and insignia. The fee per bag charged by the government includes not only the cost of producing the bag (what would be charged in a store) but also the disposal cost per bag. This method avoids both the administrative costs of measuring use and the incentive for indi- viduals to shift their costs to neighbors. There is a compliance cost to users, how- ever, because they must arrange to purchase the special bags. To facilitate this and reduce those compliance costs, the local government may arrange to have the bags sold by private retailers rather than just at the government offices, although counterfeiting is a potential problem. An alternative is to permit residents to use any disposal bag, but afix a sticker sold by the city. This method could also be extended to provide different charges for different types of refuse (bottles and cans versus paper, for example) by having different color bags or stickers sold for dif- ferent fees. As noted previously, it also might be appropriate to charge a fixed disposal fee per month or year to be eligible to use the bag system.

Any use fee based on the actual amount of refuse generates an incentive for con- sumers to avoid the charge by littering and creates a corresponding cost to the gov- ernment for enforcement. For instance, illegal dumping might occur on vacant land, in business dumpsters, or into surface-water sewer systems. The costs from those externalities might outweigh any gains from requiring a refuse-collection fee. On the other hand, a use fee based on quantity also generates an incentive for consumers to avoid refuse through recycling, using returnable containers, and substituting reusable for disposable materials (such as cloth towels rather than paper towels). One alternative is to impose a fee on manufacturers or sellers to induce them to change the packaging or nature of products, such as a disposable- diaper tax considered in Arizona. Another option is a recycling fee that is returned to the consumer if the product is recycled, such as bottle and can deposits.

Refuse collection costs depend on the type of refuse and the density and location of the users. Obviously, collections requiring a special vehicle or extra trip (for example, collection of household durables, such as refrigerators) should ideally entail a specific charge. In practice, however, it is not clear that the absence of such a charge generates much inefficiency—replacing those durables is probably insen- sitive to disposal costs. The argument for these special collection charges, then, must be fairness. Routine collection costs, on the other hand, depend mostly on time and the density of consumers. It takes longer to collect from widely spaced single-family residences, for example, than from multifamily residences with all

CH APTER EIGHT ■ PRICING OF GOVERNMENT GOODS—USER CHARGES

1 93

PART III • PROVISION OF STATE AND LOCAL GOODS AND SERVICES

192

refuse in one location, perhaps deposited in specially designed large containers. It may be appropriate therefore, as some localities do, to charge a lower fee per unit of refuse for apartments and commercial establishments than for residences.

In practice, both local governments and private firms provide refuse-collection services. In the first case, financing out of general taxes still is most common, although fixed charges per structure per month are sometimes used. Among private firms, fixed monthly charges are most common, although the charge often applies to a fixed, maximum amount of service; extra service brings extra charges. In many rural areas, refuse disposal is still the responsibility of individual con- sumers who make the weekly trip to a disposal site or recycling facility, which may be operated by the government or a private firm and is financed either from taxes or dumping charges.

Financing Public Parks and Recreation Areas

Paying for admission to private recreational facilities (such as beach clubs, pools, tennis courts, and camping facilities) is expected. Similar use fees are used for some types of public parks, beaches, and recreation facilities, such as the Narra- gansett (R.I.) Town Beach, but three issues seem to be important in limiting the broader application of state—local user fees in this area.

First, taxpayers often question the fairness of charging for the use of public facil- ities that have been acquired with general tax revenues, arguing that such facilities already have been paid for and thus should be "free" to all taxpayers. Partly this viewpoint reflects a misunderstanding about the difference between fixed capital or access costs and variable operating costs. Both types of costs must be paid some- how, and it might make sense to charge everyone for the first cost and only users for the latter cost. Indeed, one can ask whether it is "fair" not to charge users for operating costs, if those services primarily benefit those users.

This distinction and resulting policy was explained as follows in an unusually candid letter to the editor by the director of a state natural resources department:

Proposal D [allowing the state to borrow funds by selling bonds to be repaid from future taxes] provided $60 million for badly needed repair of existing facilities .. . includling] updating sewage systems, replacing electrical services, repairing bath- houses, picnic shelters and rustic cabins, and repaving roads. The bond money will not pay operational costs.

Operation is primarily financed by the users. Approximately 80 percent is from the motor vehicle entrance fee and fees collected for camping and other services. The remaining 20 percent is from general tax revenues. Fees pay wages, provide mainte- nance, and pay operational expenses such as electricity. . . . (Hales, 1989)

It seems important, as is done in the preceding, to explain to taxpayers the types of costs to be paid by all taxpayers and those specifically by users.

Second, in some cases, there is just not a sufficient level of use of these facilities to warrant user fees. Two forces often work together here. At low use levels, there may be no marginal operating costs, which calls for a zero price. Even if an effi- ciency reason exists for use fees, collection costs may be prohibitive when use is

low. It usually doesn't make sense to pay a toll collector $5 an hour, if toll collec-

tions aren't greater than that (and perhaps substantially so). 17 Third, combining the first two issues, even if sufficient use or crowding calls for

charges, congestion seems to be the least understood and often most opposed rea-

son for fees. A probable reason is that the government or public authority incurs no identifiable direct costs to justify the charge. Rather, the reason for the conges- tion charge is to ration use of a public facility, with some preferring a different rationing mechanism (first-come, first-served; a lottery; or whatever) and others denying that rationing is called for. Of course, one possibility is to dedicate the con- gestion charge to a fund for expansion or improvement of the facility, thereby cre- ating a direct reason for the extra use fee. Even though congestion charges are

exceedingly common in the private sector (higher prices at recreational resorts or parks during peak demand periods such as holidays and weekends), their use by state and local governments remains problematical.

INTERNATIONAL COMPARISON

Whether to utilize user charges to finance publicly provided goods with private- good characteristics is a classic issue in many nations, not just the United States. For instance, there has been increased attention paid to how water service should be financed in Australia. In 1992, the federal Industry Commission recommended that the state and local water authorities move toward a complete user charge sys- tem, under which consumers would be billed for water by the liter. The Commis- sion based its position on a concern about efficient use of resources, arguing that if consumers see water as "free," they are more likely to waste it. The Commission noted that user charge financing ". . will reduce water consumption and thereby waste water discharges and bring financial savings by deferring investments to expand water and sewage networks" (Tideman, 1992, p. 9).

In the past, water service had been financed mostly through a property tax. Most consumers paid a separate water property tax (a water rate), allowing consumption of water with no additional charge up to an annual limit. Water charges apply over the limit. Because the limits often are high, however, in most cases, a household's water payment depended on the value of the house or property rather than the amount of water actually used. Not surprisingly, therefore, many consumers living in high-valued houses support a move to water charges, because they believe that their water property tax is more than what direct water charges based on use would be.

Some changes in financing have occurred. The Melbourne Water Corporation increased use of charges, which now account for 31 percent of its revenues. Officials in Melbourne believe an opportunity exists for consumers to conserve, especially because they estimate that 40 percent of water consumption goes for gardens and 20 percent for toilets. One local government official in Adelaide noted that the discussion of charges "has made people aware of the value of the

17/in alternative here is a voluntary "honor system" for collection of fees. But even such a system is likely to require

enforcement (and costs) some of the time to encourage participation.

c opTER EIGHT ■ PRICING OF GOVERNMENT GOODS-USER CHARGES 1 95

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERVICES

resource. ... If people waste water they have to pay for it." But the great bulk of water consumption in Australia nationally is for agricultural uses. Some agricul- tural industries have opposed water use charges over concern both that consumer prices for some agricultural commodities might increase and that the higher production costs that result might hurt the ability of agricultural producers to export their goods, which is an important part of the Australian economy. For instance, reacting to the Industry Commission recommendation, an official of the sugar cane industry argued that many sugar cane growers would go out of business if the Queensland state government adopted the recommendations. (Daily Mercury, 1992, p. 9). The absence of water fees, then, may not only encour- age and subsidize excessive consumption by households, but also represent a subsidy to specific industries.

government) and compliance costs (to the consumers) of collecting the charge

00y offset any expected efficiency gains. Even if there is no additional cost to the government of providing a service to

an additional consumer, that consumer may impose congestion costs on other

users. The purpose of use fees in those situations is to allocate a scarce resource

among competing demands and provide a measure of the demand for new capital

investment. potentially, user charges can be composed of three separate parts: (1) an access

charge to cover all or part of capital costs, (2) a use fee to cover all or part of the

operating costs to the government associated with use, and (3) a congestion charge to cover the costs imposed by an additional user on other users. Two-part prices

can be one way to accomplish this; the first price covers the fixed costs and the sec- ond price (the marginal use fee) covers marginal operating and congestion costs.

DISCUSSION QUESTIONS User charges, prices charged by governments for specific services or privileges and used to pay for all or part of the cost of providing those services, have always been important but have become increasingly so in the past decade. They are to be dis- tinguished from financing services through general taxes, with no direct relation- ship between tax payment and service received. User charges create an incentive for efficient choice because consumers face the true costs of their consumption decisions.

Financing methods, which can be considered as user charges, include direct charges for use of a public facility or consumption of a service, license taxes or fees paid for the privilege of undertaking some activity (such as fishing license and dri- ver license fees), and special property tax assessments levied for a specific service.

Charges and fees represented about 19.5 percent of the general revenue of state-local governments in 2002, with traditional user charges alone representing 15 percent of revenue. Education and hospitals are the two budget categories from which most state-local user charges arise; on average about 55 percent of all subnational government direct user charges are attributable to those categories. Total user charges of state-local governments, whether broadly or narrowly defined, have increased faster than other revenues and faster than the general price level since 1980.

User-charge financing is more attractive, the greater is the share of marginal ben- efits that accrues to direct users, the greater the percentage of benefits of a service or facility that go to direct users, the more easily users can be identified and excluded (at reasonable cost) from consuming the service unless the charge is paid, and the more price elastic is demand. Two other potential advantages of user charges are that they are one way to have nonresidents pay for the benefits they enjoy, and the perception of fairness from users paying may result in more public acceptance of state and local government provision of certain services.

Objection is raised to user fees on the grounds that they are a disadvan- tage for consumers with lower incomes and that the administration costs (to the

1. In many large cities, the government operates a museum, library, and zoo that are visited by substantial numbers of people who are not residents of the city. They may come from the metropolitan area or from around the state. What economic reasons would justify the city financing these services through user charges? What problems would user-charge finance present in these cases? Consider how the charges might be structured for each service.

2. Suppose that your state provides a number of parks with majestic mountains, beautiful beaches, and unspoiled wilderness areas. These parks were acquired and operated in the past using the state's general tax revenue. Now the state proposes to charge a daily entrance fee of $5 per vehicle, with the revenue earmarked for the "state park fund" (to be used for operating expenses, capital improvements, and acquisition of new parks). At a public hearing on the proposal, one citizen complained "It is unfair to require taxpayers who have paid for these parks with their tax dollars to now also pay a fee to use them." As director of the state parks department, how would you respond to this citizen?

3. Suppose that partly as a result of this type of complaint, the state park user- fee proposal is revised so that no fee will be charged for park use Monday through Friday, a $5 fee will be charged on weekends, and a $10 fee will be charged on holidays and holiday weekends (Memorial Day, Fourth of July, Labor Day, and so on). Is there any economic rationale for such a structure? Do you think it is fairer than charging the same fee at all times? More efficient?

4. Suppose that the apartment building you live in at college has only one water meter for the entire building. The landlord receives a water bill from the city each quarter based on the gallons of water used, but each apartment or tenant is not charged separately—the cost of water is

196

PART III ■ PROVISION OF STATE AND LOCAL GOODS AND SERV!

effectively included in the rent. Now the water department decides t o install separate meters for each apartment and to bill each separately

rather than the landlord (so the rent is reduced by $X per person for all

tenants). The city justifies the cost of the extra meters and billings on th e grounds that the city's scarce water resources will be used more efficiently

,

What is the price to a tenant or apartment per gallon of water before a nd after the new meters are installed? Do you think the new procedure will

reduce water use? If so, how might the student tenants of these apartments

act to conserve water? Will there be a gain in economic efficiency?

SELECTED READING

"Costing and Pricing Local Government Services." Governmental Finance, 11 (March 1982): 3-27.

Gramlich, Edward, "Let's Hear It for User Fees." Governing (January 1993): 54-5. Muskin, Selma, ed. Public Prices for Public Products. Washington, D.C.: The Urban Institute,

1972.

Netzer, Dick, "Differences in Reliance on User Charges by American State and Local Governments." Public Finance Quarterly, 20 (October 1992): 499-511.

CHAPTER 9

RGOVERNMENTAL GRANTS

The basic economic justification for federal functional grants-in-aid is provided by the widespread, and ever- increasing, spillover of benefits from some of the most

important state and local expenditure programs.1 —GEORGE F. BREAK

HEADLINES

GOVERNMENT SPENT $1.9 TRILLION IN THE STATES, THE DISTRICT OF COLUMBIA,

AND OUTLYING AREAS DURING 2002, ACCORDING TO TWO REPORTS RELEASED

COMMERCE DEPARTMENT'S CENSUS BUREAU. THIS WAS AN 8 PERCENT INCREASE

IANS BENEFITTED THE MOST, RECEIVING $206 BILLION, FOLLOWED BY THE

EW YORK ($129 BILLION), TEXAS ($123 BILLION), FLORIDA ($105 BILLION),

VANIA ($86 BILLION). ONE - THIRD OF ALL FEDERAL. EXPENDITURES WENT TO

IN THESE FIVE STATES, WHICH ACCOUNT FOR 36 PERCENT OF THE TOTAL U.S.

ER, SOCIAL SECURITY, MEDICARE AND MEDICAID ACCOUNTED FOR $890 BIL -

NT) OF THE U.S. GOVERNMENT'S 2002 DOMESTIC SPENDING.

WARDS CLIMBED TO $412 BILLION, AN 11.6 PERCENT INCREASE OVER 2001, WITH

E LARGEST, AMOUNTING TO $148 BILLION, UP 11.1 PERCENT. 2 "

ri tss ca: June Relations 42xi )t .tite United States. Washington, D.C.: The tirookings Institution, 1967, 105. of Commerce. Press release entitled "Federal Domestic Spending Up 8 Percent in 2002, Census

197