Economic analysis, at heart, involves the study of how a change in one
segment of the economy is diffused throughout the rest of the economy. Those
who have studied in the field of public finance have long recognized that
the person from whom a given tax is collected is not necessarily the one who
ultimately pays the tax.
It should be recognized that any tax levied directly on a business will
ultimately be paid by real, live people–if not consumers via higher prices,
then business owners via reduced profits or employees via reduced wages. In
the first instance, the tax is considered to be shifted "forward," and in
the second and third instances it is considered to be shifted "backward" to
the factors of production. (Taxes may also be exported out of state, thereby
relieving the burden in state. Of course, other states’ taxes may end up
being imported into Texas as well.) In any case, or in any combination where
the tax burden is borne jointly, the old cliché is true: "Only people pay
taxes."
Governments levy taxes, for the most part, to cover the costs of their
expenditures. In and of itself, a tax will have two direct economic effects.
First, it will alter the relative prices of goods and services, affecting
what is produced and how. Second, to the extent that virtually every tax
takes more income from some groups than from others, it will alter the
distribution of income.
Incidence analysis attempts to identify who bears the ultimate burden of
a given tax. The analysis can be conducted on two levels: first, measurement
of the initial direct "impact" of the tax in terms of the shares borne by
consumers and/or different business sectors; and second, measurement of the
ultimate "incidence," frequently represented by translating the initial
impacts in terms of their effects on different household income groups.
The analysis is complicated because it is difficult, if not impossible,
to isolate a change in one tax without taking into account the effect on
other taxes or expenditures. For example, eliminating an exemption in one
tax would imply either an equal decrease in another tax (to compensate for
the increased revenues) or an equal increase in spending–either of which
would have its own incidence implications above and beyond the incidence of
the exemption being repealed.
The study of tax incidence is also made difficult because of competing
policy goals. That is, while some taxes are justified on the basis of
fairness or equity (the "ability to pay" principle); others are justified as
user fees (the "benefits received" principle). The former is best
exemplified by the federal income tax; the latter, by federal and state
motor fuels taxes, which are earmarked for highway construction and
maintenance and other public transportation.
Most incidence analyses reflect a concern for how well the tax or
tax/expenditure system meets the "ability to pay" principle, which holds
that those with higher incomes should bear higher tax burdens. Here, it is
useful to distinguish among three different degrees of tax equity:
regressive, proportional, and progressive.
A tax is considered "regressive" when the tax burden as a share of income
increases as income decreases; "proportional," when the share of tax burden
relative to income remains constant for all income groups; and
"progressive," when the share of tax burden relative to income grows larger
as income increases. As such, taxes on alcohol and tobacco are considered
regressive (because consumption levels remain relatively flat as income
rises); a "flat" single-rate income tax (without any deductions or
exemptions) is considered proportional; and the federal tax on luxury
automobiles is considered progressive. Note that under either a proportional
tax or a progressive tax, the ability to pay principle may be satisfied,
because people with higher incomes pay more under either tax.
For practical purposes, most empirical incidence analyses are reduced to
measuring the effects of a single tax in isolation of all others, without
taking into account the effects of other taxes or any government
expenditures or transfers. Even here, however, economists must confront the
thorny problem of accurate income measurement. That is, the results can vary
depending upon whether income is measured at the individual or household
level, in terms of "current" or "lifetime," and whether it is "gross,"
"adjusted gross," or "taxable" income. This problem becomes particularly
difficult at the lower end of the income scale, where transfers–which are
not always susceptible to accurate quantification–make up a significant
portion of the income stream.
In addition, and equally complicated, is the problem of determining the
proper "shifting" assumptions–what portion of the tax is shifted to
consumers, what portion is shifted to labor, what portion is shifted to
capital, and what portion is exported out of state.
The shifting effects will depend on many things, including how producers
and consumers respond to price changes and whether a particular market is
competitive or monopolistic. In general, most tax burdens are believed to be
borne jointly by producers and consumers–raising the price paid by consumers
and reducing the revenue received by producers, with the share of the burden
depending upon the level of competition and the price elasticity of demand
for the item being taxed. The more inelastic the demand, the greater the
burden shifted to the consumer (consider the tax on cigarettes). The more
elastic the demand, the greater the burden borne by the producer (consider a
tax on milk in glass milk bottles but not on milk in paper cartons, each a
close substitute for the other in the eyes of most consumers).
Similarly, when the producer enjoys a monopoly over the good being taxed
(consider a tax on local telephone service), the greater the ability to
shift the tax forward onto consumers by raising prices; but when the
individual producer has no ability to set prices (consider the world oil
market), the less the ability to shift the tax burden.
Finally, the answer to who bears the tax burden can vary depending upon
whether the analysis focuses on the short term or the long term. For
example, imposition of an increase in the fee for a liquor license or an
annual occupation tax would not be expected to be shifted forward in higher
prices in the short term because the fee would be considered part of the
firm’s fixed costs, whereas prices are determined by marginal costs (the
cost of producing one incremental unit of the item sold). In this instance,
the fee would be borne entirely by the producer.
In the long run, however, when all costs are taken into account,
resources would shift and prices would adjust to take the tax into account
in determining price, and as such the producer would be able to shift at
least a portion of the burden forward onto consumers.
Recognizing the impracticality of developing an incidence model that
satisfies all the demands of pure economic theory, the tables in the
following section reflect the necessity of making certain basic assumptions,
which are described in the beginning of that section. Perhaps key among
these assumptions is that consumers will bear the ultimate burden of any
taxes levied directly upon them.
While the following tables may be of great interest for policy makers, it
nevertheless must be recognized that the results depend not upon hard
science but upon subjective assumptions–and that the only thing that can be
said with certainty is that no one really knows how taxes (particularly
those levied on property and business) are shifted.
