week 10

HEADLINES

organize the discussion of each specific revenue source. Thereafter, each

revenue source is discussed in turn beginning with the "big three taxes"—

property, income, and sales—and finishing with business taxes and gov-

ernment enterprises, such as utilities, lotteries and gambling, and sale of alcoholic beverages.

CHAPTER

PRINCIPLES OF TAX ANALYSIS

I2

. . . No local, state, or federal government conducts its finances in an economy closed to the outside. . . . So there is little reason to believe either that all taxes are borne by res-

idents of taxing regions or that the ultimate interregional distribution of these tax loads is very simple.'

—CHARLES E. MCLURE, JR.

THE FACES OF STATE AND LOCAL TAXES

Consumption Wealth Income E Business

PERSONAL INCOME TAX

CORPORATE INCOME TAX

VALUE-ADDED TAX

LICENSE TAXES

SALES TAX

USE TAX

MOTOR FUEL TAXES

ALCOHOLIC BEVERAGE TAXES

TOBACCO PRODUCTS TAXES

HOTEL/MOTEL TAX

RESTAURANT MEALS TAX

TELEPHONE CALL TAXES

GAMBLING TAXES

PROPERTY TAX

ESTATE TAX

INHERITANCE TAX

TRANSFER TAXES

'Commodity Tax icidence in Open Economies.' N atiertal Tax Journal, 17 On 1%3) 187.

295 2 94

opT ER TWELVE ■ PRINCIPLES OF TAX ANALYSIS PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS

297

The basic economic issues and tools of tax analysis are introduced in this chapter. You should know and understand the methods and results discussed in this chap- ter because they will be directly applied to specific taxes in the following chapters. If you have never studied economic analysis of taxation, this should be sufficient introduction to allow a general analysis of the effects of different subnational gov- ernment taxes. For others, this chapter may be a review.

THE ECONOMIC ISSUES: INCIDENCE AND EFFICIENCY

Tax Incidence Tax incidence is the analysis of which individuals bear the ultimate burden of taxes, that is, the burden after the economy has adjusted to any changes caused by the taxes. Incidence is usually defined as the change in private real incomes and wealth because of an adoption or change of a tax. Because individuals and firms may react to taxes by changing behavior, the taxpayers who bear the ulti- mate burden of a tax—that is, the economic incidence—may be different from the taxpayers from whom the tax is initially collected or levied upon, the statutory incidence of a tax. Incidence analysis usually considers the distribution of the amount of revenue generated by a tax, the revenue burden. That burden must be compared to something, however, so incidence is usually a relative concept. One possibility is to compare the incidence of one tax to the incidence of another tax that generates an equal amount of revenue, the differential incidence. A second possibility is to compare the incidence of the revenue of a tax to the incidence of the benefits of the goods and services financed by the tax, the balanced-budget incidence.'

The first step in doing incidence analysis is to determine which prices change and by how much they change as a result of the tax (or the tax and spending package). Of course, the prices of consumer goods and services as well as factors of production can change, so a tax may affect individuals both from their uses of income (consumer purchases) and their sources of income (factor prices such as wages, rents, and interest). Suppose such an analysis shows that the price of consumer good X rises and the price of factor of production Y falls because of a tax change. Thus, consumers of good X and suppliers of factor Y bear the burden of this tax. With this information, how can one determine the burden on a specific individual or a group of individuals, say those with incomes between $20,000 and $30,000? You must know the amounts of good X consumed and factor Y supplied by this individual or group of individuals. Those quantities, multiplied by the change in prices caused by the tax, show the magnitude of the tax burden imposed on each class. For instance, a person may bear none of the tax burden (if that person neither consumes X nor supplies Y),

some of the tax burden (if that person consumes X but does not supply Y, or vice versa), or the full effect of the tax burden (if the person both consumes X and supplies Y).3

After the burden of a tax change—because of the changes in prices of goods and services—is determined, that burden is usually characterized by its effect on

income distribution. The terms progressive, proportional, and regressive are used to describe the effect of a tax on private-income distribution. Unfortunately, these

terms can have more than one definition and meaning in tax analysis. In this book,

we adopt the most common usage of these terms, describing tax burden as a per-

centage of income (unless a different specific definition is given). Those definitions

are as follows:

■ Progressive tax: Tax burden/income rises as income rises.

■ Proportional tax: Tax burden/income constant as income rises.

■ Regressive tax: Tax burden/income falls as income rises.

A progressive tax change therefore imposes a burden that is a greater fraction of income for higher-income persons than lower-income individuals. In contrast, a regressive tax change imposes a greater percentage burden on lower incomes as opposed to higher incomes. Continuing the previous example, suppose that the amount spent on good X as a fraction of income is greater for higher-income than lower-income taxpayers, although suppliers of input Y are distributed equally throughout the income distribution. A tax change that increases the price of X and decreases the price of Y would be progressive. This definition is still somewhat uncertain, however, because income could be annual income or some longer-term measure such as lifetime income. The importance of these different measures of income will be considered in Chapter 14.

Efficiency As discussed in Chapter 2, an economy is efficient if marginal social cost equals marginal social benefit for all goods. The efficiency cost of a tax change refers to the changes in production and consumption of goods caused by the tax change that causes marginal social cost and marginal social benefit to no longer be equal. The tax revenue generated by the tax change does not represent an efficiency cost because that money is simply transferred from one part of the economy to anoth- er; the tax revenue is used to provide government goods and services that have corresponding benefits. The efficiency cost of a tax arises, rather, because indi- viduals and businesses change their behavior due to the tax. By consuming dif- ferent goods, which are less desirable than those that would be consumed in the absence of the tax, and by supplying different amounts of factors of production, the economy is moved to a less efficient or lower welfare position by the tax change.

2A third possibility is absolute incidence, which is the incidence of a tax change when neither other taxes nor government spending are changed. The tax change would alter the government surplus or deficit and have macroeconomic effects.

31f input Y is also used to produce other products, then the analysis is still more complicated. Consumers of those other products could be affected.

G 1 Go

Quantity of gasoline

CHAPTER TWELVE ■ PRINCIPLES OF TAX ANALYSIS PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS

2 99

The efficiency cost of a tax change refers to the lost private welfare beyond that caused by the transfer of private income to tax revenue for the government. This is called the excess burden of taxation, that is, the burden over and above the rev- enue generated. The implicit assumption in this definition is that it may be possi- ble to use some tax structure to collect a given amount of revenue at zero efficiency cost. Any other potential tax structure that can be used to generate the same revenue can be evaluated against this standard in terms of the efficiency cost, which is the welfare burden in addition to the revenue (the same for both tax structures).

A General Rule for Tax Analysis If there is one general rule for economic analysis of taxes, it is this: The only way to avoid a tax (legally) is to change your behavior. For instance, if a tax is imposed on the consumption of cigarettes, consumers can reduce their tax burden only by reduc- ing the amount of cigarettes consumed or by purchasing cigarettes in a different (lower-tax) location. Similarly, a tax on the sale of gasoline can be avoided or shifted by firms only if producers sell less gasoline or sell it in a lower-tax juris- diction. The rule also applies to broader-based taxes, in addition to specific excise taxes. An individual can reduce income tax liability only by earning less income or earning income in a lower-tax jurisdiction. 4

This rule makes clear that tax incidence and tax efficiency are inherently con- nected. If individuals and businesses do not change their behavior in response to a tax change, then no efficiency cost is created and determining tax incidence is simple—the tax change is a burden only for those directly taxed. If, on the other hand, individuals and businesses do change their behavior because of the tax- induced price changes, then the tax change will have an efficiency cost. Determin- ing tax incidence will be more complicated as individuals and businesses act to shift the tax burden to others.

There is an important corollary to this general rule. If the only way indi- viduals and businesses can avoid tax burdens is by changing their behavior, it stands to reason that the more an economic agent is willing to change behavior, the more the tax burden can be avoided. For instance, an individual who drives a car but stops driving entirely because of the imposition of a gasoline tax and switches to a bicycle obviously pays less of the gasoline tax than someone who continues to drive exactly the same amount as before the tax. But an effi- ciency cost may have been created if the bicycle transportation that this individ- ual substitutes for driving is less preferred by that person (such as when it is cold or wet). A person who does not drive a car at all still may bear part of the gasoline tax if that tax is included in the prices of other goods this person consumes.

'It is sometimes argued that businesses also can avoid a tax by raising prices. But higher prices are expected to reduce the amount demanded by consumers, requiring lower output. Thus the businesses would change production.

SINGLE-MARKET TAX ANALYSIS

It is useful to consider how to apply the general principles outlined previously to specific tax situations. The easiest way to illustrate those principles is to consider the effect of a tax on only one market, the market in which the tax is directly levied, called partial-equilibrium analysis.

A Unit Excise Tax Suppose that a tax of $t per gallon is to be imposed on the consumption of gaso- line. Suppose also that gasoline is a commodity provided in a competitive market, as represented in Figure 12.1. Before the tax is imposed, the market is in equilibri- um at price Po and quantity of gasoline Go.

The imposition of a specific tax on a commodity can be analyzed either by shift- ing the demand curve down by the amount of the tax or by shifting the supply curve up by the amount of the tax—the methods are equivalent. In this case, because the tax is imposed on the consumers, we analyze the tax by shifting the demand curve down to Demand s . If consumers are to consume the same amount of gasoline after the tax is imposed, the price the seller charges must fall by the amount of the tax so that consumers would still pay Po. That is exactly what demand curve Demand s represents. In essence, Demand s shows the amount of gaso- line demanded for different prices received by the seller after tax, whereas demand curve D shows the amount demanded for different prices paid by the buyer. The two prices differ by the amount of the tax, so Demand and Demands also differ by t. 5

FiSL/Ye 12.1

Incidence of a

unit excise tax by

shifting demand

P1 + Tax

Po P t

tIt is said that demand is shifted down (rather than to the left) because the change is of $t and dollars are measured vertically on this graph.

Price

($)

PART IV ■ REVENUE FOR STATE—LOCAL GOVERNMENTS C HAPTER TWELVE ■ PRINCIPLES OF TAX ANALYSIS

301

Incidence of a unit

excise tax by

shifting supply

Price (s)

P2

PO P2—Tax

Demand

G 1 Go

Quantity of gasoline

After the tax is imposed, the new market equilibrium is shown by the intersec- tion of supply and Demands, which is the demand defined by the seller's price. As a result of the tax, the amount of gasoline sold falls to G s , and the price charged by sellers falls to P i . Remember, consumers must pay the seller's price plus pay the tax in this case, so the full price to a consumer is P i + t, which is shown on the graph as the price from demand curve D at quantity G 1 . In sum, the tax causes con- sumers to pay a higher price for gasoline and thus to buy less, while sellers also receive a lower price for gasoline than they did before the tax.

As noted, the same results are obtained if the tax is analyzed by shifting supply. Suppose, instead of the preceding example, that a tax of $t per gallon of gasoline is levied on the sale of gasoline and collected from sellers. Such a tax increases the marginal cost of selling gasoline by exactly $t and thus can be represented as shift- ing the supply curve up to Supply i , as shown in Figure 12.2. Because the sale of gasoline is more costly to sellers than it was previously, less is offered for sale or supplied at every price. The original supply curve shows the quantity supplied for different prices received by the seller excluding the tax, whereas the new supply curve shows the amount of gasoline supplied for different prices includ- ing the tax.

The new market equilibrium is shown by the intersection of Demand and Sup- ply i . As before, the amount of gasoline sold falls to G 1 , the price paid by consumers rises to P2 (equivalent to P 1 + Tax in Figure 12.1), and the price received by the seller after paying the tax falls to P2 — Tax (equivalent to P 1 in Figure 12.1). These are the same results that were obtained by analyzing this tax with a demand shift. This illustrates a more important point, however, than just the equivalence of these two analytical methods. Concerning incidence and economic effects in a competi- tive market, it does not matter whether a given unit tax nominally is levied on or collected from sellers or buyers—a unit tax levied on consumers produces exactly the same market effects as the same tax collected from sellers.

incidence Who bears the revenue burden of this tax? In this case, consumers and sellers of gasoline bear the revenue burden. Referring to Figure 12.1, due to the tax, the price consumers pay has risen from Po to Pi + t, which is less than the amount of the tax. The price sellers receive has fallen from Po to P1. The total tax revenue collected is tGi, with the consumers' share being (P1 + t – Po)G1 and the sellers' share equal to (Po — In this particular case, consumers bear a larger portion of the burden than sellers. You should understand, however, that the burden on sellers is a bur- den on people, not some business entity. The sellers' burden may result in lower profits to the owners, lower wages to employees, or lower prices for other factors of production. How the sellers' burden is divided among factors cannot be deter- mined in single-market analysis.

What determines the division of revenue burden between consumers and sell- ers? Following the general rule noted earlier, the agents (consumers or sellers) who are less willing to change their behavior bear the larger share of the burden. Will- ingness to change behavior as a tax alters prices is characterized by the price elas- ticity. If consumers are more willing to change behavior than sellers, then demand will be relatively more price elastic than supply, and sellers will bear the greater burden of any tax. In contrast, if sellers are more willing to change behavior, then supply will be more price elastic than demand, and consumers will bear the greater share of the revenue burden.

Following this rule, two special cases are presented in Figure 12.3. In Fig- ure 12.3a, supply is perfectly inelastic, reflecting the fact that the same quantity is supplied regardless of price; in essence, there is a fixed amount of this product. The imposition of a tax is shown by shifting the demand curve down by the amount of the tax, so price falls from Po to Po minus tax. Because sellers will not change their behavior—that is, alter production—as the tax changes the price, sellers bear the full burden of this tax. In Figure 12.3b, just the opposite situation is depicted. If demand is perfectly inelastic, then consumers will not change their behavior as a tax alters price, so the consumers' price rises by the full amount of the tax, and thus consumers bear the full burden of the tax.

Efficiency This unit excise tax on gasoline also creates an efficiency cost. When the tax is imposed and the consumers' price rises, consumers move up along their demand curve and purchase less gasoline. Presumably, consumers are instead purchasing substitute fuels such as gasohol, substituting more fuel-efficient vehicles, traveling less, or purchasing gasoline in a different market, perhaps from stations in a neigh- boring locality. In any case, consumers have been induced to switch to less desir- able alternatives, creating an efficiency cost. Similarly, as the tax causes the sellers' price to fall, sellers move down along their supply curve and produce less gaso- line. Instead, those resources previously used for gasoline production are switched to the production of something else. If those resources cannot be used as effi- ciently in the production of those other commodities, an additional efficiency cost is created.

Supply i

Supply

Incidence with

perfectly inelastic

supply and

demand

(a)

PART IV ■ REVENUE FOR STATE—LOCAL GOVERNMENTS

Figure 12.3

it TWELVE ■ PRINCIPLES OF TAX ANALYSIS

303

Efficiency cost of

a unit tax Price

($)

Supply = Marginal cost

Demand

Demand

Q

Qo

Quantity

Pc, +Tax

(b)

Measurement of the efficiency cost due to changes in consumers' behavior is depicted, for a simplified case, in Figure 12.4. In this case, the market-supply curve, which you may recall arises from the firm's marginal cost of production, is per- fectly elastic (horizontal). The assumption of perfectly elastic supply means that any amount of the product can be supplied at the market price, but that none will be supplied if the price falls below that market equilibrium. One example of such a situation is a product that is sold in many locations but whose price is set in a national or world market according to costs. For instance, after the world price of oil is determined, sellers need not sell oil in any market where the price is below that world price, because they can sell in other markets at the world price. This exact situation is common in the world of state—local government finance, with

individual states or localities being small enough that they are price takers for goods sold in national (or world) markets. 6

The imposition of a tax is analyzed by shifting the demand curve down by the amount of the tax. As a result, the quantity falls from Qo to Qi, the sellers' price remains constant at P o, and the consumers' price rises by the full amount of the tax

to (Po + tax). 7 The efficiency cost is the difference between the benefit to consumers and the

opportunity cost to the society of each unit of the product not produced—that is, the difference between marginal social benefit and marginal social cost. Assuming that one can approximate marginal social benefit by demand, the efficiency cost of the tax is the difference between Demand and marginal cost (MC) for those units no longer produced due to the tax (Qo – Qi)• Thus, the efficiency cost is represented graphically by triangle ABC in Figure 12.4. The tax generates revenue of tQl. The resources no longer needed to produce as much of this product, equal to (Qo – Q1)MC, are shifted to the production of other products at no efficiency loss because marginal cost is constant.

The efficiency cost, which is represented by triangle ABC, can be computed with a simple formula. The area of any triangle is equal to % X the length X the height. Applying that formula to triangle ABC, the area is %(AB)(BC), which

'Even for commodities sold in national markets, it is possible that prices may differ by location because of transporta- tion costs, for instance. But for some pricing strategies, firms will bear transportation-cost differences and charge equal prices in all locations, such as the single "destination charges" used by automobile manufacturers. For a

discussion of the theoretical issues, see Martin Beckmann (1968).

'Because supply is perfectly elastic, all the tax burden is borne by the consumers. If the sellers' price fell below Po,

none of the product would be offered for sale, as the price would be less than marginal cost.

PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS ?copTER TWELVE ■ PRINCIPLES OF TAX ANALYSIS

Figure 12.5

3 0 5

Price ($1

Q Qo Quantity

Tax = rp

Supply

Demand

Demand s

C

equals 144Q0 – Q1 ). One half the tax times the change in quantity can be rewritten to produce the result that

Efficiency Cost = 1/2t2EQ/P

where E = price elasticity of (compensated) demand. 8 In other words, the efficien- cy cost depends on the price elasticity of demand, the amount purchased, and the tax rate squared. This last factor is very important because as a tax rate is increased, the efficiency cost rises at a faster, quadratic rate.

One should note three important warnings about using this formula to approx- imate efficiency costs. First, the formula applies exactly only if the demand curve is linear so that the efficiency cost is represented exactly by a triangle. Second, the formula applies only if the supply function is perfectly elastic (horizontal). If sup- ply is somewhat elastic (the function is upward sloping), then the formula is more complicated and includes the price elasticity of supply. Third, the formula suggests that if the price elasticity of demand is zero (the demand curve is perfectly inelas- tic or vertical), then the efficiency cost is also zero. This is generally not correct. The problem arises because this is single-market analysis and ignores the behavior of consumers in other markets. Because the price of the taxed product has changed, consumers may alter their behavior in other markets (by purchasing less of some other product or by working less, for example), which would create an efficiency cost. This possibility is examined in the appendix to this chapter. 9

One other case exists in which a tax may not have an excess burden or efficien- cy cost. You learned in Chapter 2 that efficiency requires that marginal social cost equal marginal social benefit. The production or consumption of some commodi- ties produces external costs, however, so the social cost of the activity is greater than the private cost and private benefit. Air pollution resulting from gasoline combustion and use of automobiles is one example. In such a case, a tax actually can improve economic efficiency by forcing consumers or producers to perceive the full costs of their activity. If automobile use imposes costs on everyone that are not taken into account by drivers, there will be too much auto use and too much air pollution from society's viewpoint. A tax on auto emissions or a tax on gasoline use would make drivers see the full cost of their activity and lead to less use and less pollution. An excise tax used to offset such an externality is often called a Pigouvian tax, named after the British economist A.C. Pigou.

How I a Percentage Tax Ditterent? The preceding analysis is for a unit tax—that is, a tax of so many dollars per unit of product, such as $.18 per gallon of gasoline. The analysis of the more common percentage or ad valorem tax, such as a sales tax of 5 percent of the price, is only

slightly different. As before, the tax can be analyzed by shifting the demand curve down by the amount of the tax per unit or by shifting the supply curve up. The dif- ference is that the tax per unit depends on the price. If the tax rate is r percent, then

Tax Revenue = r(Price)(Quantity) The tax per unit is then

Tax Revenue/Quantity = r(Price)

Obviously, the higher the price, the larger the tax per unit in dollars, and the more the demand or supply curve must be shifted to reflect the tax. In Figure 12.5, the demand curve is shifted down by the amount of the tax with the distance being larger, the higher the price is.

Although the analysis of the efficiency cost of this tax is exactly the same as for the unit tax, the formula to compute the approximate efficiency cost is different, as shown here:

Efficiency Cost = ir2EPQ

As before, the efficiency cost depends on the price elasticity of demand, the amount of the commodity purchased (now measured in dollars), and on the tax rate squared. 1°

Incidence and Etticiency of a Subsidy Single-market analysis also can be applied to examine the incidence and efficiency of a subsidy offered in that market. A subsidy is a payment from the government

Analysis of a

percentage tax

8The price elasticity of demand, E, is the percentage change in quantity/the percentage change in price. That is, E = [(Q° - (21)/13)1/[t / N. Solving for (Qa - Qi ) and substituting into the equation for area gives the result.

9For a good discussion, see Harvey Rosen (1999, p. 285-294). mAgain, the area of the efficiency cost triangle is 16(AB)(BC), which equals 1/2rP(Qo - (21). Because Qo - (21 equals

rEQ, the efficiency cost area equals lir2 EPQ.

Subsidy

C Supply

-I -V,

Demand s

Demand

Qo Qi Quantity

D B

(b)

PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS

Figure 12.6

Incidence and

efficiency effects

of a subsidy

Price (5)

Price (5)

Subsidy \A C

Po Po Supply

Po - S Po - S D9 Bt‘ Supply i

Demand

Qo Qs Quantity

(a)

that lowers the price or cost of some economic activity to individuals or business- es. Examples among state—local government programs include food stamps, which reduce the price of purchasing food for eligible households; mortgage revenue bonds sold by states to make low-interest housing loans to eligible families, which reduce the price of homeownership; state or local support of higher education, which allows the private price of higher education to fall short of total cost; and perhaps Medicaid, which pays all or part of the cost of health care for certain low- income persons (although some argue that health care is not sold in a competitive market).

Suppose the government offers a subsidy of $S per unit consumed or sold for some commodity sold in a competitive market. For clarity, assume that the supply of this product is perfectly elastic, reflecting the idea that the price is determined in some broader market. Just as with taxes, the subsidy can be analyzed by either adjusting supply or demand. In Figure 12.6a, the subsidy on units sold reduces the marginal cost of producers and is analyzed by shifting supply down by the amount of the subsidy, S. As a result of the subsidy, the price to consumers falls by the amount of the subsidy (from Po to P o — S), and quantity bought and sold rises from Qo to Q l . Because we have assumed that supply is perfectly elastic, the price received by sellers, including the subsidy, remains at Po. Thinking about the inci- dence of the subsidy, then, all the benefits go to consumers, who see the price fall by the full $S.

If the subsidy were offered directly to consumers, an equivalent analysis res- ults from shifting the demand curve up by the amount of the subsidy, as in Figure 12.6b. Again, for the same market conditions, the price to consumers falls to Po — S and consumption rises to (21.n

"Subsidies can be thought of as negative taxes, and thus the analysis is exactly reversed. A tax increases producers' costs and sifts supply up vertically, although a subsidy reduces costs and shifts supply down. A tax reduces demand (shifting it down), whereas a subsidy increases demand (shifting it up).

CHAPTER TWELVE ■ PRINCIPLES OF TAX ANALYSIS

Continuing the parallel to tax analysis, the benefits of a subsidy are divided

between consumers and sellers based on the relative elasticities of demand and supply. If supply is perfectly elastic, as in Figure 12.6, then price is set at a given

level and consumers benefit from the full subsidy. Sellers can lower the price to Po — S and still receive the world price of P o when the subsidy is added. If the quantity supplied is fixed (perfectly inelastic supply), then suppliers get all the benefits from the subsidy. Consumers continue to pay the original price for the fixed amount of the product and sellers pocket the subsidy.

In addition to costing the government a direct amount, the subsidy also has an

efficiency cost. In Figure 12.6a, the subsidy lowers the price and provides benefits to all those who would have purchased this good without the subsidy (Q o), and the subsidy provides benefits to those consumers who are induced to consume more of the good (from Q o to Qi ) due to its lower price. The magnitude of the benefit to original purchasers is shown graphically by the area S Qo or PoADP0 — S, and the benefit to new purchasers is represented by the area ADB, which is the difference between the marginal benefit (demand) to those consumers for each unit from Q o

to Qi and the marginal cost (supply). The sum of these two benefits, represented by area PQABPQ — S, is less than the amount of the subsidy paid, represented by area PoCBP0 — S. The difference between the two, triangle ABC, represents the effi- ciency cost or excess burden of the subsidy.

This efficiency cost also can be explained or understood from a different per- spective. For each unit from Q o to Qi, the marginal social cost (Supply) is greater than the marginal social benefit (Demand). The subsidy makes it seem that this commodity is cheaper then it really is and thus induces society to allocate too many resources to its production or consumption. For instance, if a state subsidizes the consumption of housing, then consumers ignore the source of the subsidy, believe that housing is now less expensive, and increase consumption of housing. That increase in housing consumption may require that consumers change their behavior elsewhere as well, however, perhaps consuming less of something else (clothing) or working more. Because consumers made this change due to the sub- sidy, they really prefer more clothing or leisure. The subsidy has induced con- sumers to make an inefficient choice.

Just as with taxes, however, this efficiency analysis has one important qualifica- tion. As noted in Chapter 2, if a product provides benefits to other than the direct consumers, which occurs if there are externalities, then a subsidy is called for to offset the external benefits. For instance, if education provides benefits to everyone in society in general, then each person would underestimate the benefits and choose too little education from society's viewpoint. In that case, a subsidy for edu- cation corrects an inefficiency rather than creates one.

Limitations ofr Single -Market Analysis Although single-market analysis is helpful in illustrating the general principles of tax analysis, it is often not very precise for two reasons. First, the effects in other markets, whether for other goods or for the same good in a different location, are not considered. Second, the manner in which any sellers' burden gets distributed

307

Price

(S)

Supply6

(a) (b)

309

Figure 12.8

Demand G

Demancl

Gp Quantity of gasoline

Mo M 1 Quantity of motor scooters

Supplym

DemandM

Demand M

Multimarket analy-

sis of a unit excise

tax with increasing

costs

p 1 +Tax

P1

Price (S)

Figure 12.7

Multimarket analy-

sis of a unit excise

tax with constant

costs

Price

(S)

Po + Tax

Price

(5)

SupplyG

Demand G

Demand

G 1 Go Quantity of gasoline

(a)

Supplym

DemandM

Demand M

Mo Mi Quantity of motor scooters

(b)

PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS

308

among the various factors of production is not analyzed explicitly. Although this may not be much of a problem in some cases where intermarket effects are small, often intermarket effects can be substantial, particularly in the world of state-local governments with relatively easy mobility among jurisdictions. Therefore, we turn now to multimarket analysis, effectively applying the same type of supply- and-demand analysis not only for the market in which the tax is directly imposed but also for other, closely connected markets.

MULTIMARKET ANALYSIS

Etteets in Parallel Markets

Here we consider the effects of a tax, including the effects of the tax in the mar- kets for complementary or substitute goods. As an example, we can expand con- sideration of the effects of a unit tax of $t on gasoline to include those in the market for motor scooters, assuming that cars and motor scooters are substi- tutes. That situation is shown in Figure 12.7, with the simplifying assumptions of perfectly elastic supply of both gasoline and motor scooters. Given the national price for Best Unleaded Gasoline and Your Favorite Motor Scooter, sellers will require that price in all markets in the long run. As before, the imposition of the unit tax on gasoline is represented by a downward shift in the demand for gaso- line (to DemandG 1 ). As a result, the quantity of gasoline consumed decreases and the consumers' price rises, in this case by the full amount of the tax because of the perfectly elastic supply. Consumers now purchase G1 units of gasoline at a price of Po + t.

COAP TER TWELVE ■ PRINCIPLES OF TAX ANALYSIS

Because the price of gasoline has increased, consumers will act to reduce con- sumption, perhaps by substituting 80-mile per-gallon motor scooters for 20-mile per-gallon cars. Thus, the demand for motor scooters is expected to increase, shown by the rightward shift of demand from DemandM to Demandmi in the motor scooter market. Assuming perfectly elastic supply, the amount of motor scooters purchased and produced rises, but the price remains the same in the long run. By consuming less gasoline, consumers have reduced the amount of gasoline tax they pay. (Consumers pay tGi rather than tGo.)

The situation is only slightly more complex if constant costs do not prevail so that the supply curves in both markets are not perfectly elastic but are positively sloped, as depicted in Figure 12.8. In this instance, as previously discussed, the unit tax on gasoline causes both an increase in the consumers' price (but by less than the amount of the tax) and a decrease in the sellers' price. The increase in the consumers' price of gasoline causes an increase in the demand for motor scoot- ers, which now causes an increase in the price of motor scooters due to the upward sloping supply. Because additional numbers of scooters cost more to pro- duce than the previous ones, the price must rise to make that extra production worthwhile.

Because of this price increase, the original motor scooter consumers (those who purchased quantity M0) also are hurt by the gasoline tax; the higher motor scooter price is charged to all consumers, not just those who switch from cars due to the gasoline price increase. Motor scooter consumers pay an increased amount equal to - Po)Mo; however, this amount is not transferred revenue to the government, nor is it an efficiency cost lost to the economy. This extra amount consumers pay is transferred to the sellers through the higher price of motor scooters. To complete this multimarket analysis, then, it is also necessary to expand the analysis to the factor markets behind these consumer-goods markets.

Price

(5)

Effects of an excise

tax on gasoline in

factor markets

CO Cl

Demand°

Demands

Tt To Quantity of tanker trucks

(a) (b)

PART IV ■ REVENUE FOR STATE—LOCAL GOVERNMENTS

310

Figure 12.9

Long-run supply

Po

SRS

Price

($)

Supply

P2

Demand s

Demand °

Quantity of plastic

ER TWELVE ■ PRINCIPLES OF TAX ANALYSIS

3''

r

Meets in Factor Markets Changes in consumption away from gasoline and cars and toward motor scooters as a result of an excise tax on gasoline also may have implications for the factors of production used in producing those goods. Some of those potential implications are shown in Figure 12.9. The decrease in the consumption of gasoline could lead to a decrease in the demand for the services of tanker trucks to carry gasoline to wholesale distributors and retail outlets. The immediate effect, given the number of trucks To, is a decrease in their value to P 1 . If the long-run supply of tanker trucks is perfectly elastic, as depicted earlier in Figure 12.7a, then the effect will be a reduction in the number of tanker trucks over time, so that the value of the trucks, or the rental rate for tanker-truck services, returns to the previous level. Of course, the reduction in the number of tanker trucks or in the amount of tanker- truck service used may have implications for the drivers or producers of trucks.

Similarly, the increase in demand for motor scooters due to the tax on gasoline may increase the demand for plastic, assuming that motor scooters primarily are constructed from plastic (and little plastic is used in producing cars). In this instance, we assume that the long-run supply of plastic is positively sloped, thus requiring a price increase to induce more production. The tax on gasoline therefore has the effect of increasing the revenue to producers of plastic, who benefit in effect from the gasoline tax. Recall that in Figure 12.8 you saw that motor scooter consumers pay an increased amount to motor scooter producers as a result of the gasoline tax. In the example, at least part of that gain to motor scooter producers becomes a gain to plas- tic producers. The excise tax on gasoline imposed a burden on gasoline consumers, but also caused a transfer of resources from consumers to plastic producers.

Obviously, this story can continue, for instance, by asking whether the gain to plastic producers ultimately benefits workers in the industry or suppliers of

icals used in plastic production. One important aspect of multimarket tax

ysis is determining into how many different markets or how many different ges of production to carry the analysis. The appropriate answer depends on the

including both the economic conditions in a market, which determine how ge a price change is expected, and the importance of thabmarket for the equity

efficiency result.

Application to State and Local Government Issues ultimarket analysis is essential when dealing with state-local government

taxes because the focus is often on the effect of a tax levied by one state or local- ity when there is mobility among states or local jurisdictions. Examples abound. A consumer may go over a boundary to a store in a different location or order through a catalog or Web site to avoid sales tax. An individual may move his residence and work location to avoid an income tax, or an individual may change residence (but not job) location to reduce the residential property tax. Finally, a business may change its operating location to avoid a state busi- ness tax or local property tax. In all these cases, there might very well be eco- nomic effects in more than one market or location, both the one that imposes the tax and the one to which the economic activity moves. Multimarket tax analysis is required.

A simple relabeling of Figure 12.7 shows how the models in this chapter can be applied to these types of issues. Rather than thinking of one market for gasoline and one for motor scooters, it is just as correct to let Figure 12.7a represent the mar- ket for gasoline in jurisdiction G and Figure 12.7b represent the market for gasoline in jurisdiction M. Before taxes, gasoline sells for the same price in both locations. Now G imposes a $t unit tax on gasoline, so that consumers in jurisdiction G pay a price equal to Po + t, which is greater than the price in M. Consumers in G now not only have the choice of switching to motor scooters from cars but also of pur- chasing gasoline at a station in jurisdiction M. Obviously, some consumers from jurisdiction G decide to buy their gasoline from a station in M where the price is lower because there is no tax. 12

Why don't all consumers switch their gasoline purchases to a station in M? They will unless switching is costly or unless they are not aware of the price difference. It might be costly to buy gasoline at a gas station in M rather than a station in G if an individual had to drive, say 10 miles, from his house to the nearest gasoline station in M. In that case, the cost (both in money and time) of the drive could outweigh the tax savings on gasoline. In contrast, someone who works in jurisdic- tion M but lives in jurisdiction G could switch gasoline purchases to M at little extra cost.

What is the gain to jurisdiction M from more gasoline sales? Possibly, there are now more retailers in M and fewer in G or at least more employment in M and less in G. The increased retail sales activity in jurisdiction M could also mean that

zlt is just as correct to think that both G and M tax gasoline, but the tax in G is higher by $1.

g TWELVE ■ PRINCIPLES OF TAX ANALYSIS

3 1 3

Application 12.1

STATE DIESEL FUEL TAXES: MULTIMARKET ANALYSIS IN PRACTICE

All states levy taxes on diesel fuel, which is

used mostly by trucks, including long-haul

trucks making interstate trips. Because diesel

fuel is transported easily and the market is

worldwide, the supply to any one state is

expected to be very elastic (perhaps essen-

tially perfectly elastic). As a result, prices will

differ by state due to the state tax differences.

Because interstate truckers have some leeway

in deciding what route to take or where to

fuel their trucks, demand is expected to be

somewhat elastic. Thus, fuel price differences

due to tax differences can affect fuel sales in

multiple states.

The experience of Colorado in 1986-1987 is

an example. In July 1986, Colorado increased

its diesel fuel tax by 7.5 cents per gallon to a

then national high of 20.5 cents per gallon,

a 57-percent increase in the unit tax. In the

following 12 months, diesel fuel sales in

Colorado fell by 11 percent from about

204 million galloris to about 182 million.

Despite the decrease in the amount of fuel

sold, state revenue from the tax increased

from about $26.5 million to about $37 million,

a revenue increase of about 40 percent.Tax rev-

enue did not increase as much as the increase

in tax rate because of the fall in gallons sold.

The Colorado case is illustrated in Fig-

ure 12.10. The tax increase is analyzed by shift-

ing the supply up by the amount of the tax

change. Because a perfectly elastic supply is

assumed, the price of diesel fuel rises by the

full amount of the tax increase, 7.5 cents.

Because of that tax and price increase, quan-

tity sold falls from 204.5 to 182 million gallons.

If the initial price of diesel fuel was about $1.00

per gallon, a 7.5 percent increase in price

resulted in about an 11 percent decrease in

quantity, implying that the price elasticity of

demand for diesel fuel in Colorado was about

1.46 (11 percent/7.5 percent).

Figure I 2. le

Colorado cl fuel tax

Price

($)

Supply 1 (tax = 20.5C)

Supply() (tax =13C)

Demand

182 204 Diesel fuel (millions of gallons)

PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS

property values in M increase. These changes would benefit workers in M (regard- less of where they live) and property owners in M. The increased retail activity could (although it is not guaranteed) also increase the tax revenue to jurisdiction M from property taxes or from a local sales or income tax, if one exists.

If the price of gasoline is not determined in a national market (which is shown by the perfectly elastic supply) but rather determined in each local market, then the supply curves in each jurisdiction are positively sloped, as in Figure 12.8. In that case, as consumers switch their gasoline purchases from jurisdiction G to M, the price paid by consumers for gasoline in jurisdiction G falls, and the price in M rises. The market now creates a natural constraint on the movement of purchases from G to M; in the absence of costs of changing purchase location, consumers will reallocate their purchases until the consumers' prices in G and M are again equal.

This analysis of the interjurisdictional effects of taxes using a standard multi- market model is not limited to taxes on consumer goods but can be applied just as easily to taxes on factors of production, such as labor, land, and capital. Of course, firms' payments for these factors become the wages, rent, and profits received by individuals, so these factor taxes are sometimes referred to as taxes on the sources (as opposed to uses) of income. One common application of this type is for subna- tional government taxes on capital. The rate of return on capital investment is determined in a national (or world) market, so any one jurisdiction is a price taker; that is, the supply of capital to that jurisdiction is perfectly elastic (Figure 12.7). The suppliers of capital are individual investors, however, whereas the demanders are business firms. If one jurisdiction imposes a tax on capital, then the effect (just as with the gasoline tax in Figure 12.7) is expected to be a decrease in the amount of capital in the taxing jurisdiction and an increase in the other jurisdic- tion. These changes in the amount of capital are expected to have implications (considered in detail in Chapter 14) for consumers, workers, and landowners in both jurisdictions.

According to The Wall Street Journal (Carlson, 1988), there were two primary rea- sons for this elasticity to diesel-fuel demand. Some drivers on transcontinental trips were taking routes that avoided Colorado, traveling across Wyoming or New Mexico instead. In addition, some drivers who traveled in Colorado arranged for fuel stops to occur in neighboring states. For instance, at the time, the diesel fuel tax in Wyoming was 8 cents per gallon.

In addition to reducing diesel fuel sales in Colorado, it is easy to trace at least two other effects in other markets due to the relatively high tax rate in Colorado. First, as a result of the changes in behavior of some drivers, demand for and sales of diesel fuel in neighboring states, particularly Wyoming and New Mexico, are expected to increase. Thus, both of these states receive increased tax revenue with- out increasing tax rates (indeed, their increased revenue came from Colorado's higher tax rate). Second, if fewer truck drivers were stopping in Colorado, then decreases in sales of other goods in Colorado, such as food and incidentals, are also expected. Indeed, the Journal reported that one truck stop owner in Colorado believed the typical driver spent an average of $28 in purchases other than fuel for

312

DISCUSSION QUESTIONS

Suppose that the local legislative body in Your College Town (YCT) decides to levy a tax of $.50 for each 12 ounces of beer sold in the city (both by-the- drink and packages). The city sees the tax as a way to have students pay more for the city services they receive. Suppose that the beer market in YCT is competitive, the long-run industry supply in YCT is perfeCtly elastic, and the demand for beer in YCT is very price-elastic.

a. What will the effects of the tax be on the price of beer in YCT, the amount of beer sold, and the number of liquor stores and bars in YCT?

b. Why might the demand for beer in YCT be so price elastic, given that it is known that overall demand for beer is rather inelastic? In view of that, what do you expect the effect of the tax will be on beer sales and the number of stores and bars in surrounding cities?

2. "If supply of a good is perfectly inelastic, then the sellers of that good are expected to bear the full revenue burden of an excise tax on the sale or consumption of that good." Evaluate this statement. Can you think of any examples of goods whose supply is (at least almost) perfectly inelastic?

3. If a unit tax is increased from $1 per unit sold to $2, the efficiency cost of the tax more than doubles. Explain.

4. Under what conditions would it be possible for an excise tax to have no efficiency cost and, in fact, increase economic efficiency? Give an example or two.

SELECTED READINGS

Oates, Wallace E. Fiscal Federalism. New York: Harcourt Brace Jovanovich, 1972. See Chapter 4.

Rosen, Harvey S. Public Finance, 5th ed. Boston: Irwin/McGraw-Hill, 1999. See Chapters 13 and 14.

1.

Inditterence-Curve Analyste• of Tax etticiency

An exposition of the consumer-demand model using indifference curves and budget lines was presented in the appendix to Chapter $, Those tools can be used to more carefully describe the consumption changes and resulting effi ciency cost from taxation than is possible with basic supply-and-demand analy- sis* Therefore, the consumer-theory model is used in this, appendix to compare excise taxes on specific commodities with a general lump-sum tax.

APPENDIX

3 1 4

CHAP TER TWELVE ■ PRINCIPLES OF TAX ANALYSIS PART IV ■ REVENUE FOR STATE-LOCAL GOVERNMENTS

3 1 5

each stop. Thus, the decrease in sales for the fuel tax could spill over to markets for other commodities as well.

If supply is very elastic, as hypothesized and as seems likely, then the revenue burden of the tax increase falls on consumers of diesel fuel, which are the truckers and the consumers of trucking services. It is important to emphasize that even though the tax increase caused sales of fuel in Colorado to fall, it brought about an increase in revenue. The 57 percent increase in the tax rate generated about a 40 percent increase in revenue due to the fall in sales. The tax increase also created a larger excess burden or efficiency cost of the tax, represented by area ABC in Figure 12.10. In practical terms, this increased efficiency cost arose from the changes in trucker behavior—taking less efficient routes or longer driving times without stopping—that were used to avoid the tax increase.

Interestingly, in the period since this experience in Colorado, both the level and variation in state excise taxes on diesel fuel have increased. Average diesel fuel tax rates have risen from about $.15 per gallon at that time to about $.21 today; how- ever, Colorado's tax rate has not changed so the relative tax rate between Colorado and other states has decreased.

Tax incidence is the analysis of which individuals bear the ultimate burden of taxes, that is, the burden after the economy has adjusted to any changes caused by the taxes. Incidence is defined as the change in private real incomes and wealth because of an adoption or change of a tax. This is different than statutory inci- dence, which refers to the actual payments made by taxpayers from whom the tax is collected.

The general rule of tax analysis is the only way to avoid a tax (legally) is to change your behavior. Consumers or sellers who are less willing to change their behavior will bear the larger share of the burden.

The efficiency cost of a tax change arises because consumers or producers change their production or consumption so that marginal social cost no longer equals marginal social benefit.

Tax incidence and tax efficiency are inherently connected. If individuals and businesses do not change their behavior in response to a tax change, then no effi- ciency cost is created and the tax change is a burden only for those directly taxed. If individuals and businesses do change their behavior, then the tax change will have an efficiency cost, and determining tax incidence is more complicated.

Multimarket analysis is essential when dealing with state and local government taxes because the focus is often on the effect of a tax levied by one state or locality when there is mobility among states or local jurisdictions.

Perfectly elastic supply means that any amount of the product can be supplied at the market price, but that none will be supplied if the price falls below that mar- ket equilibrium. This situation is common in state-local government finance, with individual states or localities being small enough that they are price takers for goods sold in national (or world) markets.

SUMMARY