Public Finance, 2nd Edition

Determining whether or not the tax is actually distortionary would

require a full analysis of all its design characteristics, as well as the underlying

market environment. For instance, some Wrms may be subject to borrowing

constraints that would change their investment margins. Also, estimated

depreciation allowances may not reXect true economic depreciation as as-

sumed above. All things considered, the tax is undoubtedly distortionary to

some extent. But in light of Stiglitz' analysis, assuming that the U.S. corpor-

ate tax is nondistortionary may be a good approximation to reality.

IMPORTANT MODIFICATIONS OF THE HARBERGER MODEL

The basic Harberger model makes a number of very strong assumptions that

need to be modiWed to extend the usefulness of the model. Three especially

useful modiWcations are variable rather than Wxed factor supplies, imperfect

rather than perfect competition, and heterogeneous rather than homogenous

consumers. We conclude the chapter with brief discussions of each of them.

Variable Factor Supplies

The assumption of Wxed factor supplies simpliWes an already complex analyt-

ical model, but it needs to be modiWed for the sake of reality. Factor supplies

are certainly variable in the long run, with the single possible exception of land,

and the supply responses to a tax matter in determining the incidence of the

tax. As a general rule, they tend to reduce the change in the price of the taxed

factors, which spreads more of the burden to the untaxed factors. For

example, a reduction in the supply of capital in the Harberger model would

raise the return to capital and thereby transfer some of the tax burden to labor.

The assumption of variable factor supplies is essential in long-run dy-

namic models of tax incidence, which we will consider in Chapter 17. Changes

in the supply of capital in response to a tax alter the time path of capital

accumulation, both physical and human capital, which in turn aVects the

marginal products of capital, labor, and all other factors of production. The

resulting changes in marginal products are often the most important deter-

minants of the ultimate incidence of a tax as the economy moves to its new

long-run steady state. For example, a tax on capital that slows the rate of

capital accumulation and reduces the capital/labor ratio can shift much of the

long-run incidence of the tax to labor as the marginal product of labor and

therefore the real wage fall over time.

Variable supply responses to taxation can be important as well in static

models of tax incidence. To see this, consider the opposite extreme from the

Harberger model. Assume that the supply of capital is perfectly elastic to the

taxing jurisdiction, as depicted in Fig. 16.8. This is a realistic assumption for

state (provincial) and local governments, and for all but the largest countries

566 IMPORTANT MODIFICATIONS OF THE HARBERGER MODEL

K

S

P

K

D

K

D'

K

K

K'

K

0

P'

K

FIGURE 16.8

with the most highly developed Wnancial markets. The supply price of capital,

P

K

, is a given for these smaller jurisdictions, determined in capital and

Wnancial markets whose scope extends far beyond their boundaries. As a

consequence, taxes on capital have very diVerent implications for them than

the Harberger model would suggest. The following three implications have

received attention in the tax incidence literature.

Mobile Versus Immobile Factors

In the Wrst place, a tax on the Wrms' use of capital in these jurisdictions

cannot be borne by capital (refer to Fig. 16.8). The tax shifts D

K

down by the

full amount of the tax. Since the supply price of capital, the required rate of

return, is set on the ``world'' market, the rate of return within the jurisdiction

rises by the full amount of the tax to P

0

K

. The capitalists escape the burden,

passing it on to other factors of production or to consumers. The general

principle illustrated here is this: If a jurisdiction contains both mobile factors

(e.g., capital) and immobile factors (e.g., land and possible labor), then the

immobile factors bear the burden of any factor tax levied on the Wrms. (If all

factors are perfectly mobile, with their prices Wxed outside the jurisdiction,

then consumers bear the entire burden in the form of higher prices.)

Taxing the Demand Versus the Supply Side

Another implication of the perfectly elastic assumption is that it matters

which side of the market is taxed. Earlier we had noted the principle that it did

not matter which side of the market the legislature chose to tax: All the eVects

of a tax on one side of a market can be duplicated by a tax on the other side of

the market. The case of perfectly elastic supply is the exception to this

principle, and an important one, if the government cannot tax all the suppliers.

16. THE THEORY AND MEASUREMENT OF TAX INCIDENCE

567