[Economic Report of the President (2004)]
[Administration of George W. Bush]
[Online through the Government Printing Office, www.gpo.gov]


 
Chapter 4


Tax Incidence: Who Bears the Tax Burden?

The study of tax incidence is the economic study of which taxpayers
bear the burden of a tax. This question is of considerable importance
to policy makers, who want to know whether the distribution of the tax
burden (between rich and poor, capital and labor, consumers and
producers, and so on) meets their criteria for fairness.

Distributional tables showing the tax burdens borne by different
income groups are an important application of incidence analysis.
The Joint Committee on Taxation (JCT) and the Department of the
Treasury prepare distributional tables for the existing tax system
and for some proposed and adopted tax changes. The Congressional
Budget Office (CBO) prepares such tables for the existing tax system.
In addition to these official analyses, some private groups also
publish distributional tables.

When used properly, distributional tables can contribute to informed
decision making on the part of citizens and policy makers.
Unfortunately, mainstream economic analysis suggests that these tables
do not always accurately describe who bears the long-run burden of
certain taxes. This problem does not arise from bias or lack of
Instead, it reflects resource and data limitations, uncertainty about
some of the economic effects of taxes, and variations in the time
frame considered by the analyses. Nevertheless, the shortcomings of
distributional tables can lead to misperceptions of the impact of tax
changes.

This chapter discusses some of the ways in which distributional tables
can be improved. The key points in this chapter are:

  The actual incidence of a tax may have little to do
with the legal specification of its incidence. Official distributional
tables recognize this fact in many contexts, but not in all of them.

  In the long run, a large part of the burden of
capital taxes is likely to be shifted to workers through a reduction
in wages. Analyses that fail to recognize this shift can be
misleading, suggesting that higher income groups bear an
unrealistically large share of the long-run burden of such taxes.

To begin, it is useful to review the basic economic principles of tax
incidence and apply them to different types of taxes.


Theory of Tax Incidence


One crucial finding in the study of tax incidence is that the
economic incidence of a tax (the identity of the person who bears the
burden of the tax) can be completely different from its statutory or
legal incidence (the identity of the person upon whom the law
officially imposes the tax). In other words, the person who is
legally responsible for paying the tax may not be the one who
actually bears the burden of the tax. As explained below, the
incidence of a tax depends upon the law of supply and demand, not the
laws of Congress.

Another crucial principle is that only people can pay taxes.
Businesses and other artificial entities cannot pay taxes. Although
the corporate income tax is legally imposed on firms that are
organized as corporations, the actual burden of the tax can fall only
on people--perhaps the firm's owners, or its employees, or its
customers--but certainly not on a legal artifact such as a
corporation. Similarly, although the estate tax is legally imposed on
the estate, the burden of the tax can fall only on people--perhaps
the decedent who left the estate, perhaps the heirs, perhaps other
people--but not the estate, which is merely a legal construct
established to sort through the ownership of the decedent's assets.

It is simplest to first discuss the incidence of a simple excise tax,
a tax levied on a specific good or service. As explained below, the
key insights from this analysis can be extended to apply to other
types of taxes.

Incidence of an Excise Tax

Consider a tax on apples. Suppose that when there is no tax, the
price of apples is $1. Now, suppose that the government imposes a
10-cent excise tax on apples and that the producers are legally
responsible for paying this tax. Do producers actually bear the
economic burden of the tax?

The answer depends on what happens to the price of apples. If the
price remains unchanged, producers bear the economic burden
(the economic incidence of the tax is the same as the legal
incidence). Consumers pay $1, the same as before, and suffer no
burden. Producers, after collecting $1 from the consumers, must pay
10 cents to the government, so they clear only 90 cents.
Alternatively, if the price rises by the amount of the tax, from $1
to $1.10, consumers bear the burden. Although they do not send any
money to the government, they pay 10 cents more per apple than they
did without the tax. The producers bear no economic burden, even
though they are legally responsible for paying the tax. After
collecting $1.10 from consumers and sending 10 cents to the
government, they still clear $1, as they did without the tax.
In this case, economists say that the producers shift the burden of
the tax to consumers. To consider another possibility, if the price
of apples rises by 5 cents, to $1.05, consumers and producers share
the burden equally. Consumers bear a 5-cent burden because they pay
$1.05 for each apple, compared to the $1 that they paid without the
tax. Producers bear a 5-cent burden because they clear only 95 cents
per apple, compared to the $1 they cleared without the tax: they
collect $1.05 from consumers, but send 10 cents to the government.

As these examples show, the division of the tax burden between
consumers and producers depends on what happens to the price of
apples. When prices are free to adjust, they are likely to be
determined by the law of supply and demand. If the price of apples
was $1 with no tax, then the number of apples consumers wanted to buy
at that price must have equaled the number of apples that producers
wanted to sell at that price.

What happens when the 10-cent excise tax is imposed? It depends
on how responsive consumers and producers are to changes in the
prices they pay or receive. The relevant questions are: How many fewer
apples do producers sell if the amount they clear per apple declines?
How many fewer apples do consumers buy if the amount they pay per
apple rises?

For example, suppose that producers are four times more responsive to
price changes than consumers. Then, producers face a price change
that is one-fourth as large as that faced by consumers. The 10-cent
tax causes the price to rise from $1 to $1.08, putting an 8-cent burden
on consumers and a 2-cent burden on producers. At that price, the
number of apples consumers want to buy falls by the same amount as
the number that producers want to sell. Alternatively, if consumers
were four times more responsive than producers, then producers would
bear 8 cents of the burden and consumers would bear only 2 cents.

The group that is less responsive bears more of the burden of the
tax. The group that is more responsive escapes much of the burden
because it responds to the tax, abandoning the taxed activity when
threatened with a tax burden. The price-responsiveness of each group
depends upon its flexibility. Do producers have good alternatives
(in the form of other industries in which they can produce)? Do
consumers have good alternatives (in the form of other products they
can buy)?

The answers vary across products, types of producers (such as workers
and owners of capital), and time frames. If the excise tax applied
only to Granny Smith apples, consumers could switch to other,
untaxed, kinds of apples. If it applied to all apples, consumers
would have somewhat less flexibility. Some workers may have skills
specific to the apple industry. Other workers may be more flexible
because their skills are more general; they could avoid bearing the
tax burden by finding a job in another industry. The owners of
capital employed in a taxed industry may bear a significant short-run
burden because the buildings and equipment in the industry may be
designed specifically for its use and the owners may have little
ability to move those resources elsewhere. In the long run, though,
capital can leave the taxed industry: as buildings and equipment
depreciate in the taxed industry, new buildings and equipment are
constructed in other industries.

A similar logic applies if the product is subsidized rather than
taxed. The group that is more responsive receives the smaller benefit
because the subsidy prompts new members of that group to enter the
market and compete away the benefits of the subsidy. Conversely, the
group that is less responsive receives the greater benefit from the
subsidy because little entry occurs.

Because the incidence of an excise tax depends upon the relative
flexibility of consumers and producers, the burden may not always
fall where the Congress intends. When the Congress imposed a "luxury"
tax on yachts in 1991, for example, it intended the wealthy
purchasers of yachts to bear the burden. Such purchasers, however,
may be quite responsive to price because there are many alternative
goods that they can purchase (expensive cars and jewelry,
for example). If this is so, then a significant part of the burden of
a yacht tax may fall on workers in the industry, who may be less
well-off than owners of yachts. Indeed, after the tax was introduced,
production and employment in the boat industry fell, leading some
observers to claim that workers were bearing much of the burden of
the tax. Although the validity of this claim cannot be conclusively
determined (the industry's decline may have been caused by the
1990-1991 recession rather than the tax), the Congress responded to
these concerns by repealing the tax in 1993.

Legal Incidence Is Unimportant

As long as prices can freely adjust, the economic incidence of a tax
does not depend on the legal incidence. Suppose that, in the above
example, the government imposes the 10-cent excise tax on apple
consumers rather than apple producers. Consumers then must make the
tax payment to the government, in addition to the price they pay to
producers.

Because producers are four times more price-responsive than
consumers, the price received by producers must still fall by 2 cents
and the price paid by consumers must still rise by 8 cents. Despite
the legislative change, that is still the only outcome that keeps the
want to buy. If the tax is legally imposed on producers, they shift
8 cents of the burden to consumers. If it is legally imposed on
consumers, they shift 2 cents of the burden to producers.

Given that the price can freely adjust, it should not be surprising
that the final outcome is unchanged. It is irrelevant whether the
tax collector stands next to consumers and takes 10 cents from them
when they buy an apple or stands next to producers and takes 10 cents
from them when they sell an apple. It does not matter whether the
producer who puts it in the same bowl.

Applied Distributional Analysis of Excise Taxes and Subsidies

The legal incidence of Federal excise taxes is sometimes placed on
consumers, sometimes on manufacturers, and sometimes on other
producers or importers. In most cases, this legal incidence rightly
receives little attention. In accordance with the economic theory of
tax incidence, the JCT and Treasury economists preparing
distributional JCT generally allocates excise tax burdens to
consumers. Treasury follows a similar, but more elaborate, approach.
These approaches are reasonable, since consumers are likely to bear
much of the long-run burden of most excise taxes. In the long run,
most producers are flexible, or price-responsive, because they can
switch to other industries. Consumers are likely to have less
product being taxed.

The theory of incidence also applies to more-subtle excise subsidies,
such as those included within the individual income tax. The income
tax law grants tax reductions for purchasers of various products--for
example, an itemized deduction for medical expenses, a credit for
electric cars, and the Hope and Lifetime Learning credits for the
costs of higher education. The economic benefits of these provisions
are likely to be divided between consumers and producers, with the
greater benefit going to the group that is less price-responsive. The
long-run benefits are likely to go largely to consumers, because they
are likely to be less price-responsive than producers. Official
consumers.

The basic insight that tax burdens fall more heavily on groups that
are less flexible can be applied to a wide range of taxes. The
remainder of this chapter applies this framework to payroll taxes,
taxes on capital, and estate and gift taxes.

Payroll Taxes


The largest Federal payroll tax, earmarked to finance Social Security
and Medicare Part A, is imposed at a 15.3 percent rate on the first
$87,900 of earnings and at a 2.9 percent rate on earnings above that
amount.A much smaller Federal payroll tax, earmarked to finance
unemployment compensation, is imposed at a 0.8 percent rate on the
first $7,000 of earnings. The legal incidence of the Social
Security-Medicare tax is divided equally between employers and
employees. The legal incidence of the Federal unemployment
compensation tax is placed entirely on employers.

With a payroll tax, the product being taxed is labor and its price is
the wage rate. Applying the insights obtained from the analysis of
excise taxes, the relevant question is whether firms' demand for
labor or workers' supply of labor is more responsive to changes in
the wage rate. In the long run, it is likely that firms are more
responsive, or flexible, particularly in a global economy in which
they can relocate abroad. This conclusion implies that employees bear
most of the payroll tax burden, a result supported by empirical
studies. In other words, wages paid to employees are lower by an
amount roughly equal to the employers' part of the payroll tax. In
accord with this conclusion, official distributional analyses
generally assign the full burden of payroll taxes to employees.
The primary controversy in this area concerns whether the
distributional analysis should also include the Social Security
benefits that are financed by the payroll tax (Box 4-1).

Much of the individual income tax is also imposed on labor income.
Based on the above discussion, the burden of the individual income
tax on labor, like that of payroll taxes, should also fall on
workers. Official distributional analyses generally allocate the
individual income tax on labor income to workers.

Some taxes on and subsidies to labor income are more subtle. The
income tax laws deny firms their normal business-expense deductions
for some payments of labor income. For example, under certain
circumstances, firms cannot deduct salaries greater than $1,000,000
per year paid to senior executives or some "golden-parachute"
payments made to executives in connection with corporate takeovers.
Because of this denial of deductibility, the firm pays a tax on
these labor income payments, in addition to the regular tax on its
owners' net income. This tax operates as an additional payroll tax
legally imposed on employers, although of a much narrower scope than
the payroll taxes discussed above. On the other hand, the income tax
laws allowfirms to claim tax credits for some other payments of
labor income. Examples include the work opportunity tax credit,
the welfare-to-work credit, the empowerment zone employment credit,
and the Indian employment credit. (The work opportunity and
welfare-to-work credits expired on December 31, 2003, but may be
reinstated by future legislation.) In economic terms, these credits
are subsidies to labor.

The fact that these taxes and subsidies are implemented as changes
in the employer's (rather than the employee's) income tax does not
change their economic incidence. The fact that they apply only to
employees in specific jobs or in specific locations or to those
receiving specific forms of compensation, however, may change their
incidence. Because employees can, to some extent, change their jobs,
locations, and forms of compensation, the flexibility of the employee
may be greater than was assumed in the discussion of general taxes on
labor income. As a consequence, the division of the burdens

---------------------------------------------------------------------
Box 4-1: Social Security and Transfer Payments in Distributional Tables
In addition to collecting taxes, the government makes transfer
payments to households. The net burden that the fiscal system imposes
on households is better measured by looking at tax payments minus
transfer payments received rather than by looking at tax payments
alone. Official distributional tables, however, usually show only
tax payments. They do not tabulate the distribution of transfer
payments, except sometimes the refundable tax credits that are
administered through the individual income tax, such as the Earned
Income Tax Credit. For example, if a household has $20,000 of wage
income, pays $5,000 in taxes, and receives transfer payments of
$2,000, the distributional table would report that the household
bears a $5,000 tax burden, overlooking the fact that its net burden
imposed by the fiscal system is only $3,000. In some tables, transfer
payments are included in the income measure that is used to classify
households into different income groupsï¿½in this example, the
household might be classified as having income of $22,000 rather than
$20,000.  But, the transfer payments are not netted against the taxes in
measuring the householdï¿½s burden.
This practice induces a potential political bias because policy makers
receive ï¿½distributional creditï¿½ for helping the poor only if they do
so through the tax system rather than through transfer payments.
The omission of government benefits from distributional tables may
provide a misleading picture of Social Security. Official
distributional tables generally show that the Social Security payroll
tax imposes a smaller burden, as a fraction of income, on high income
groups than on lower and middle income groups. However, if the analysis
were expanded to include the Social Security benefits financed by the
payroll tax, it would likely reveal that high income groups bear a
larger net burden, as a fraction of income, than some other groups.
Thus, distributional tables might be more accurate if these benefits
were included in some manner. One possibility would be to treat the
present value of the future benefits accrued by a worker each year
as an offset to his or her payroll tax liability.
---------------------------------------------------------------------



or benefits between employees and firms is not clear. Official
distributional analyses generally allocate the burdens and benefits
of these provisions in the same manner as firms' other income tax
payments. (As discussed below, these analyses differ in their
treatment of the corporate income tax.)


Taxes On Capital Income


The Federal tax system imposes taxes on capital income. Capital
income generated by corporations is generally subject to the
corporate income tax. Capital income received by individuals is
generally subject to the individual income tax.

Many observers view capital income taxes as highly progressive
because capital income is highly concentrated. However, economic
analysis suggests that capital income taxes are particularly likely
to be shifted, especially in the long run. Taxes imposed on owners of
capital in one sector of the economy may be shifted to the owners of
capital in other sectors. More importantly, capital income taxes may
be partly shifted to workers through a reduction in wages. The
extent of shifting differs across time horizons because savers
(who provide capital and earn capital income) are more flexible in
the long run than in the short run.

Shifting Across Sectors

Even if a tax is imposed on capital income in one sector of the
economy, it is likely that owners of capital in all sectors bear the
same economic burden in the long run. To see why, note that if
capital is mobile across sectors, after-tax rates of return must be
equalized across sectors, after adjustment for risk. Suppose that an
economy contains two sectors and that, when there are no taxes,
capital earns a 6 percent rate of return in each sector. Now,
suppose that a 50 percent tax is imposed on capital income in one
sector, while no tax applies in the other sector. In the very short
run, capital in the taxed sector earns an after-tax return of only
3 percent, while capital in the tax-exempt sector earns an after-tax
return of 6 percent. At this point, only the owners of capital in
the taxed sector bear the burden.

This state of affairs cannot continue. Owners of capital in the
taxed sector will move their money out of that sector and begin
investing in the tax-exempt sector. As they do so, two things happen.
First, the before-tax rate of return rises in the taxed sector as
capital becomes more scarce. Second, the before-tax rate of return
falls in the tax-exempt sector as capital becomes more plentiful.
This movement continues until investors are indifferent between the
two sectors, which happens when after-tax rates of return are once
again in balance. For example, after a certain amount of capital has
relocated, the before-tax rate of return in the taxed sector may
rise from 6 to 8 percent, while the before-tax rate of return in the
tax-exempt sector may fall from 6 to 4 percent. At this point,
investors in both sectors earn the same 4 percent after-tax rate
of return. Because all investors initially earned 6 percent and now
earn 4 percent, they all bear the same burden from the tax, even
though the tax legally applies to only one sector.

For example, the corporate income tax is likely to be shifted across
sectors. This tax applies only to the corporate sector, but the above
analysis suggests that the burden is shared by owners of capital in
both the corporate and noncorporate sectors. Similarly, tax
provisions that apply to only a single industry are likely to
ultimately affect owners of capital in all industries.

Shifting to Workers

Shifting across sectors may not be the most important way in which
the burden of capital income taxes is shifted. In the long run, much
of the burden of capital income taxes (whether imposed at the
firm or individual level) is likely to be shifted to workers. The
reason is that such taxes reduce investment, which diminishes the
capital stock. With a smaller capital stock, the before-tax rate of
return to capital is higher, offsetting part of the burden that the
owners of capital would otherwise bear. Also, workers are less
productive because they have a smaller capital stock to work with
and earn lower real wages. Part of the tax burden is therefore
shifted to workers.

In accordance with the insights obtained by studying the incidence
of excise taxes, owners of capital bear less of the burden if the
supply of capital is more responsive to changes in its after-tax
rate of return. This responsiveness, and hence the extent to which
capital income taxes are shifted, depends upon several factors,
including the amount of time that has elapsed since the tax was
imposed, the willingness of consumers to substitute between current
and future consumption, and the extent to which capital can escape
the tax by relocating abroad.

The time frame is very important. The shifting of the tax burden to
workers is likely to occur slowly because it takes time for large
changes in the capital stock to occur. In the short run, the tax
causes little change in the capital stock, because most of the
capital on hand was already in existence when the tax was adopted.
With little change in the capital stock, very little of the burden is
shifted from owners of capital to workers. Over time, however, the
tax has a greater impact on the capital stock as it discourages the
accumulation of new capital. As a result, more of the tax burden
falls on workers and less falls on owners of capital.

Under certain assumptions, the entire burden of the capital income
tax is shifted to workers in the long run, although owners of capital
bear much of the burden in the short run. A textbook model of
economic growth, called the Ramsey model, provides an illustratio
of this effect. (The Appendix to Chapter 5, Dynamic Revenue and
Budget Estimation, explains the basic features of this model.) Using
plausible values for the key inputs to the Ramsey model demonstrates
that the economy adjusts only gradually to a capital tax increase.
Initially, 100 percent of the burden of a capital tax increase is
borne by the owners of capital, since they have already invested
in the capital currently in place. Five years after the tax
increase, about a quarter of the tax burden has shifted to workers.
Ten years after the tax increase, workers have taken on over 40
percent of the burden. It takes 50 years for the burden to shift
nearly completely--by that time, capital owners bear only 6 percent
of the burden and workers bear 94 percent.

If consumers are more willing to substitute between present
consumption and the future consumption made possible by their
savings, saving is more responsive to the after-tax rate of return
and more of the capital income tax is shifted. The responsiveness of
saving to the after-tax rate of return also depends on consumers
planning horizons. The Ramsey model assumes that consumers consider
the impact of their saving decisions on their descendants. If,
instead, consumers plan only for their own lifetimes, saving is less
responsive to changes in its after-tax rate of return and less of
the capital income tax burden is shifted to workers.

International capital flows also play a role. If the tax applies
only to capital located in the United States and capital is mobile
across international boundaries, the tax is more likely to be shifted
to workers. The above example assumes that there are no international
capital flows; incorporating such flows would increase the speed at
which the tax is shifted.

Empirical work provides some evidence that capital income taxes are
shifted to some extent: studies find that the before-tax return to
capital income is higher when the tax rate on capital income is
higher. However, the picture is not entirely clear, because other
factors may cause tax rates and before-tax rates of return to move
together.

The belief that a large portion of the capital income tax burden is
shifted in the long run is common in the economics profession. In a
1996 survey, public finance economists were asked to state "the
percentage of the current corporate income tax in the United States
that is ultimately borne by capital." The average response was 41
percent, and three-quarters of the respondents gave answers of 65
percent or less. This survey indicates that the average public
finance economist believes that more than half of the tax is
eventually shifted from the owners of capital to workers or other
groups.

Because labor income is more evenly distributed across taxpayers than
capital income is, recognizing that part of the burden of capital
income taxes is shifted to workers reveals that high income
taxpayers bear a smaller share of the burden than is often assumed.
Chart 4-1 classifies households by their levels of total income and
tabulates the share of national labor income and national capital
income earned by different groups. The chart shows, for example,
that the 10 percent of households with the highest total incomes
receive 37 percent of labor income and 62 percent of capital income.
If half of capital taxes are shifted to workers in the long run,
the fraction of the burden falling on this high-income group is
percent to 49 percent; if all capital taxes are shifted to workers
in the long run, the high-income share of the burden falls to 37
percent.


[Graphic Not Available in Tiff Format]


Applied Distributional Analysis and the
Choice of Time Frame

Official distributional analyses differ in their treatment of the
corporate income tax. The JCT previously distributed the burden to
owners of corporate capital, but now does not distribute it on the
grounds that the incidence of the corporate income tax is uncertain.
The CBO and Treasury now distribute the corporate tax burden to
owners of all capital. None of these analyses currently recognizes
the shifting of the tax to workers. The CBO previously presented
analyses that allocated half of the burden to workers and Treasury
did the same in its January 1992 corporate integration study.
Official analyses generally allocate individual income taxes on
capital income to the persons who bear the legal incidence of the
taxes.

The time frame plays a key role in how tax incidence is treated.
When the JCT adopted its former practice of allocating the corporate
income tax to corporate capital, it stated that its analysis was
intended to refer to the very short run, when little shifting of
any kind would occur. Similarly, Treasury has justified allocating
the burden to owners of all capital by stating that this is the most
reasonable assumption for incidence over a 10-year horizon. These
analyses serve the useful objective of informing policy makers of
how the current tax burden is divided between current workers and
current owners of capital.

Nevertheless, presenting estimates only for short time frames leaves
an incomplete picture. If a tax change is intended to be permanent,
it is important to also inform policy makers how its long-run burden
will be divided between future workers and future owners of capital.
Answering that question requires additional distributional tables
that recognize the significant shifting to workers that is likely to
occur in the long run.

Estate and Gift Taxes

Capital can also be subject to estate and gift taxes when its
ownership changes hands due to an inheritance or gift. The lessons
from the analysis of capital income taxes can therefore be applied to
estate and gift taxes.

The estate and gift taxes apply on a cumulative basis to an
individual's lifetime gifts and to the estate the individual
bequeaths at his or her death. An individual may make up to $11,000
of gifts to any recipient per year, without counting them against the
lifetime total. Bequests to surviving spouses are exempt, as are
gifts and bequests to charitable organizations.

The taxes apply only when lifetime gifts plus the estate exceed an
exemption amount, which was $675,000 in 2001. Under the laws in place
at the beginning of 2001, the exemption amount was scheduled to
increase to $1,000,000 starting in 2006. The taxes applied at rates
of up to 55 percent.

The tax law adopted in June 2001 provides for further reductions in
the estate and gift taxes for 2002 through 2009. This law increases
the exemption amount to $1 million for 2002 and 2003 and gradually
increases it to $3.5 million for 2009. The law reduces the top tax
rate from 2002 to 2009, with top rates of 50 percent in 2002 and 45
percent in 2007 through 2009. For 2010, the law completely repeals
the estate tax, but retains the gift tax with a top rate of 35
percent. It also increases, in some cases, the capital gains taxes
paid by heirs who sell property that they inherit.

Because the 2001 tax law is scheduled to expire at the end of 2010,
the estate and gift taxes are scheduled to return in 2011, at the
levels specified by the previous laws. The President has proposed
permanently extending the provisions of the 2001 tax law that are in
effect in 2010, including the repeal of the estate tax.

The issue of who benefits from estate tax repeal has been a
prominent one in the debate over repeal. Treasury allocates the
burden of estate and gift taxes to the decedents (the individuals who
have died) and donors. The JCT used to do the same, but has now
stopped distributing them due to uncertainty about the taxes'
incidence. The CBO's recent distributional analyses have not included
estate and gift taxes. Allocating the estate tax burden to decedents
supports the common view that the tax is highly progressive, since
(at the current exemption amount) the tax applies to only the
largest 2 percent of estates.

It is virtually certain, however, that little of the economic burden
of the estate tax is borne by the decedents. The burden of the estate
tax is borne by them only if the tax prompts them to reduce their
lifetime consumption and accumulate a larger estate, so that the tax
can be paid without reducing the after-tax bequests left to their
heirs. In other words, the estate tax must reduce lifetime
consumption and promote estate accumulation for it to be borne
by the decedents.

This condition is unlikely to hold. Because the estate tax makes
estate building less attractive, it probably reduces the size of
bequests. Empirical research confirms that the estate tax reduces
the amount that decedents accumulate and pass on to their heirs.
As a first step, it would make more sense to distribute the burden
of the tax to heirs rather than to decedents.

Despite what one might expect, the heirs of wealthy decedents are
not always wealthy. Economists have found that the correlation
between the long-term labor earnings of successive generations
is around 0.4 or 0.5. The correlation between long-term incomes
(which includes the inheritances themselves) or between long-term
consumption levels of successive generations has been estimated to
be around 0.7. (Correlation is a number, ranging from -1 to 1, that
measures the strength of the relationship between two variables. A
correlation of 0.4 or 0.7 indicates that one variable tends to
increase when the other increases, but that the relationship is
not perfect.) Some bequests are left to grandchildren or nephews
and nieces where the correlation between the incomes of decedents
and the incomes of heirs may be even lower. Because heirs can be
less wealthy than decedents, recognizing that the estate tax burden
is more likely to fall on the former reveals that less of the burden
is borne by the very wealthy.

A more important point, however, is that the reduction in estate
building induced by the tax is likely to take the form of a
reduction in capital accumulation. Because the estate and gift taxes
are taxes on capital, part of their long-run burden is likely to be
shifted to workers through a reduction in wage rates, as discussed
above. Part of the burden is therefore likely borne by ordinary
workers who never receive a bequest or taxable gift.


Conclusion

Distributional analysis can be a useful tool for policy makers. It is
important, however, to recongnize the limitations of existing
analyses.  Current analyses can be misleading, particularly with
respect to the estate and gift taxes and other capital taxes.  These
taxes are likely to be shifted substantially to workers in the long
run, reducing the extent to which their burden falls on high-income
groups.