The Hidden Entitlements
PART I
An Overview of Tax Expenditures
"Addressing the problems of the poor, Gingrich asked: 'Don't
we have to bring the poor with us? Maybe we need a tax credit for poor Americans
to buy a laptop.'"
-The Washington Post, Jan. 6, 1995, reporting on
House Speaker Newt Gingrich's Jan. 5 testimony before the House Ways and
Means Committee
On Thursday, January 5, 1995, his second day as Speaker of the House, Rep.
Newt Gingrich (R-Ga.) suggested a new federal spending program for poor
families one that would appear to cost as much as $25 billion. A
few weeks later, Gingrich admitted his off-the-cuff idea to buy every poor
American a laptop computer was "dumb." But why would a man who
usually rails against federal spending be so extraordinarily cavalier about
such an expensive new program? For one simple reason: in Gingrich's mind,
he hadn't called for more government spending, but instead had proposed
a tax cut! Rather than having the Department of Health and Human Services
buy computers for the poor, the checks would be written by the Internal
Revenue Service. That might look like a distinction without a difference,
but for Gingrich it appears to have made the potentially enormous cost of
his plan irrelevant.
"Tax expenditures" is the official term used to describe the vast
array of government spending programs that are implemented through the Internal
Revenue Code programs that will total $3.7 trillion over the
next 7 years. As the congressional Joint Committee on Taxation explains:
"Special income tax provisions are referred to as tax expenditures
because they are considered to be analogous to direct outlay programs .
. . . Tax expenditures are most similar to those direct spending programs
which have no spending limits, and which are available as entitlements."
1
What makes tax expenditures similar to spending programs is that they are
special tax provisions that are designed to accomplish some social or economic
goal unrelated to equitable tax collection. They are like "entitlements"
because they are not subject to annual budget appropriations, but are paid
out to any business or individual that meets the eligibility rules, regardless
of the total cost.
For instance, suppose the government wants to subsidize wages for low-income
workers. It could try to accomplish this goal in various ways. One might
be by regulation, to wit, by setting a minimum hourly wage that businesses
are required to pay. Alternatively, the Department of Health and Human Services
could provide direct wage subsidies to eligible workers. Or a wage subsidy
could be administered by the Internal Revenue Service, by reducing income
taxes for low-income workers, including tax "refunds" for those
who owe no income tax.
In fact, the government does all three of these things. First of all, of
course, there is a minimum wage. Second, many low-income workers have their
salaries supplemented by welfare, food stamps, unemployment compensation
and so forth. And third, the tax code provides a substantial "earned-income
tax credit" to low- and moderate-income working families.
Most government spending through the tax code is not targeted toward low-income
people, however. In fact, tax breaks tend to reward those with the most
lobbying muscle in Washington, especially organized business interests.
Many personal income tax subsidies offer much larger benefits to well-off
people than to average families, not just in dollars but even in percentage
terms. A deduction for $1,000 in mortgage interest, for example, is worth
$396 to a 39.6% top-bracket taxpayer, but only $150 to a 15%-bracket home owning
family of four earning $40,000. And it's worth nothing at all to the majority
of taxpayers who don't itemize. By law, the congressional Joint Committee
on Taxation and the Treasury Department must issue reports each year listing
tax expenditures and their estimated cost. (The two lists are almost identical.)
These "tax expenditure budgets" are designed to be informational
rather than prescriptive, so they include almost any tax provision that
can plausibly be characterized as the equivalent of a direct spending program.
In some cases, however, an item listed as a tax expenditure may not really
be a subsidy. Instead, it might be defensible on pure tax policy grounds
as a proper adjustment in computing ability to pay taxes.
For example, deductions for state and local income and property taxes are
included in the official tax expenditure budgets. Some might say, however,
that the state and local taxes a family pays reduce its ability to pay federal
taxes, and thus that deductions for those expenses should not be considered
a subsidy. In other words, it's argued, a New York family making $75,000
a year in total income has a lower ability to pay federal taxes than a Texas
family with the same income, because the New York family pays higher state
and local taxes. The deduction for extraordinary medical expenses has been
defended on similar ability-to-pay grounds, as has the charitable deduction
(at least for cash donations).
These arguments are not universally accepted, however. For example, in 1984,
the Reagan Treasury contended that (a) homeowner property taxes provide
direct, measurable benefits to families that pay them, such as better schools,
trash collection, etc.; (b) families living in areas with low property taxes
get fewer of these services or must purchase them privately; and therefore
(c) fairness dictates that property taxes should not be deductible. (Congress
rejected Treasury's argument.) Because there is honest disagreement over
whether deductions for state and local taxes, large medical costs and charitable
donations are proper adjustments in computing ability to pay taxes or are
instead subsidies, they are included in the tax expenditure budget for informational
purposes.
Although the rule of thumb is to include in the tax expenditure budget any
tax provision that is arguably a subsidy, there are some items that look
remarkably like subsidies but are not included on the official tax expenditure
lists. For example, the tax code allows a deduction for half of amounts
spent on "business meals and entertainment." Lobbyists for restaurants,
golf courses, and professional sports argue strenuously that such write-offs
are not only a necessary subsidy to their industries, but are also proper
deductions in computing net business income. Many people, however, wonder
why the government is targeting $6 billion a year to help pay for executives
to eat at nice restaurants and attend expensive sporting events, while most
people must pay for such things out of their after-tax earnings. Yet this
apparent subsidy is not included in either of the official tax expenditure
reports.
Despite some controversy at the margins over what should or should not be
termed a "tax expenditure," most of the items on the official
tax expenditure lists from mortgage interest deductions to capital
gains breaks are generally agreed to be deviations from normal tax
policy that are functionally equivalent to spending programs. These tax
entitlements loom very large in the overall budget picture. In fact, the
total tax expenditure budget comes to about $455 billion in fiscal 1996.
That's two and a half times as much as all means-tested direct spending
programs cost. In fact, it's almost as much as the government spends on
defense and interest on the national debt combined.
Size alone would seem to mandate that any serious analysis of possible ways
to cut federal spending and reduce the budget deficit must include tax subsidies
within its scope. And like other spending programs, tax expenditures ought
to be evaluated on the following grounds:
Before getting to the details of particular tax expenditures, one might
question whether the Internal Revenue Service is ever the appropriate agency
to administer a government spending program. After all, the IRS's expertise
is in tax collection, not housing or farming or business investment. Would
we ask the Energy Department to administer the Social Security system on
the side? Would we expect the Defense Department to do a good job running
the food stamp program? Does anyone think the Labor Department should be
in charge of securities regulation?
To be sure, handing a program to the IRS to run has advantages. The bureaucratic
overhead may be fairly low, since the IRS will inevitably devote most of
its attention to its main mission of collecting taxes. But the price for
that lack of attention may well be fraud and inefficiency in the administration
of the program. Tax-based subsidies for donations of artworks to museums,
for example, have been scandal-ridden for decades. Hugely expensive business
tax expenditures purportedly designed to encourage productive investment
have been perverted into tax shelters and corporate "leasing"
scams. Even the earned-income tax credit has been criticized over the years,
both for its failure to reach many of those who are its intended beneficiaries
and conversely for insufficient policing against fraudulent claims.
Another basic question about tax-based subsidies involves their "entitlement"
nature. As such, they run on auto-pilot once they are put into the tax code.
In contrast, so-called "discretionary" spending on defense, roads,
environmental protection, and other non-entitlement programs must be approved
every year, and it takes an appropriation bill passed by Congress and signed
by the President to do so. Under current law, discretionary spending is
subject to a hard freeze for most of the rest of this decade. In fact, under
the President's proposed 1997 budget and GOP budget plans, too, such spending
will be frozen at about $540 billion a year through fiscal 2001, unadjusted
for inflation or population growth, and will go up only with inflation in
fiscal 2002. That means an inflation-adjusted reduction in such programs
of 13% over that period and a 26% cut as a share of the economy. In contrast,
the total cost of tax expenditures is expected to rise from $434 billion
in fiscal 1995 to $609 billion in fiscal 2002. Thus, while discretionary
spending must shrink radically as a share of the economy, tax expenditures
will grow by 17% in constant dollars just about keeping pace with
the economy.
If a discretionary program turns out to cost more than expected, it
or something else must be scaled back in the annual budget. But if
the price tag on a tax break goes up, it continues anyway and the
process of curbing it is much more difficult.
For example, when tax expenditures for Individual Retirement Accounts were expanded
in 1981, the change was expected to cost only a few hundred million dollars
a year. Instead, IRA subsidies were soon costing the Treasury more than
$10 billion annually. Yet the subsidies continued for 5 years until they
were finally scaled back in 1986.
The 1986 Tax Reform Act illustrates, by the way, that tax expenditures are
not always untouchable. The 1981 Reagan tax bill expanded tax expenditures
so dramatically that by the mid-eighties our income tax system had quite
literally become more loophole than tax. That is, revenues foregone through
tax expenditures actually exceeded total income tax collections. Indeed,
corporate tax expenditures had skyrocketed to double the amount companies
paid in income taxes. In response to public outrage over corporate tax freeloading
and high-income tax shelters, Congress and President Reagan closed loopholes
and lowered income tax rates dramatically. As a result, today's tax
expenditures as a share of income taxes paid are back to about their
pre-Reagan level-although they are still higher than in 1970.
Still another fundamental question about many tax expenditures at
least those structured as personal deductions is their "upside-down"
nature. Who could imagine a direct government spending program that paid
an increasing share of, say, mortgage costs as peoples' incomes rose? Yet
that's exactly the effect of the current deduction for mortgage interest.
Unfortunately, lawmakers sometimes ignore the perversity of tax-deduction-based
subsidies. For instance, President Clinton has proposed a new tax deduction
for college tuition. This proposed tax subsidy would not benefit the truly
rich, because it would be phased out between $100,000 and $120,000 in total
income. But a family making $100,000 could get a 28% tuition subsidy under
the plan, while a family making $40,000 could get only 15% of its tuition
costs reimbursed. Thus, a family making $100,000 sending its child to Harvard
could save $2,800 a year in taxes under the plan, while a family making
$40,000 sending its child to the state university would save only about
$600.
Even tax subsidies structured as credits against taxes otherwise due can
produce strange results. Suppose a new entitlement program was proposed
to give families a $500 annual grant for each dependent child under age
18. Is it conceivable that families of four making less than $16,900 would
be made ineligible for the grant? Would anyone dare propose to limit families
with two children making $20,000 to only a $465 grant, but to give 2-child
families making $200,000 the full $1,000? Yet the Republican "Contract
with America" proposed exactly such a new subsidy program and
President Clinton has a similar proposal in the form of a child tax
credit that would be unavailable to lower-income families because it cannot
exceed income taxes due.
Poor administration, lack of cost controls, and perverse distributional
results are "features" that are far too typical of tax-based subsidies.
Yet despite these obvious drawbacks, many politicians, particularly those
who style themselves "conservatives," find tax expenditures extremely
attractive. One wonders: Do they actually favor the upside-down approach
to subsidies? Do they think poorly administered programs are a good idea?
Are they unconcerned about the impact of uncontrolled spending on the budget
deficit? Or do they simply see tax subsidies as a way to earn political
points with their backers and exert power over society and the economy without
having their efforts show up in the official spending budget?
This last point may be the key. Because of the way the government's budget
books are kept, unscrupulous politicians can have their cake and eat it,
too. To institute a new direct spending program that doesn't increase the
deficit, Congress must enact an offsetting tax increase to pay for it. Officially,
this will show up in the budget as higher federal spending and higher federal
taxes. But if an equivalent tax expenditure program is enacted and is paid
for with higher taxes on people and/or companies not benefited, the combination
will show up in the aggregate budget numbers as a wash. Neither net taxes
nor spending will appear to go up in the official budget.
In their proposed "Contract," GOP leaders in Congress talked a
lot about cutting spending. But among the most significant specific expenditure
changes they proposed in 1995 were more than $100 billion a year in increased
tax-based spending programs. Ironically, these huge new tax entitlements
mostly targeted to large corporations and the wealthy were
designed to show up in the budget not as additional spending, but as tax
cuts.
Ultimately, of course, tax entitlements are not free. If all current tax
expenditures were suddenly repealed, for example, the deficit could be eliminated
and income tax rates could be reduced across the board by about 25%. Such
a radical step is unlikely, of course, and in the case of some tax expenditures,
the net change in most people's bottom-line tax payments might not be very
significant. But there are many expensive tax subsidy programs whose benefits
are very heavily concentrated on the best-off people. Eliminating or scaling
back these kinds of tax entitlements could make a very significant difference
in improving tax fairness and easing most people's tax burdens. Such steps
would also be likely to improve economic growth to boot, by curbing wasteful
tax-sheltering activities and thereby increasing productive market-driven
investment.
Summary Cost TableTax Expenditures,
1996-2002
continue
back to contents