The Trouble With Tax Cuts
By Bruce Bartlett
N RECENT YEARS, tax cuts have fallen sharply as an issue of concern to voters. Poll after poll puts cutting taxes well down the list of priorities. This is puzzling because taxes as a share of gross domestic product or personal income are at all-time highs and have risen very sharply during the Clinton administration. Taxes have risen much more sharply than during the 1970s, when the increases led to a tax revolt epitomized by Proposition 13 in California and the election of Ronald Reagan in 1980 largely on a promise to cut tax rates across the board.
Political conservatives lament the decline of tax reduction as a political issue, but as yet have offered no coherent explanation for why this should be the case. Many have merely fallen to repeating their calls for tax cuts in the apparent hope that repetition alone will make people more receptive. That is always a possibility. But on the theory that advocacy is most effective when advocates know not only what they want but also what others want, a better approach might be to take a hard look at why the case for tax cuts is failing.
Tax and poll data
T IS NOTORIOUSLY DIFFICULT to draw definitive conclusions from polling data, because questions are asked at different times by different organizations in slightly different ways. Nevertheless, although the particular numbers vary from place to place and time to time, as far as taxes are concerned the general trend is unmistakable. People simply do not rate tax reduction as a primary goal today. Consider the following data from September 1999 polls:
- Fox News/Opinion Dynamics. Registered voters were asked what are the two most important issues for the federal government to address. First was education (30 percent), second was Social Security (23 percent), third was health care (18 percent), with taxes coming up fourth (15 percent).
- Harris. Americans (not just registered voters) were asked the same question. First on their list of concerns was education (17 percent), second was crime (16 percent), and taxes (15 percent) were third.
- CBS News. Americans were asked to select the single most important issue they would like presidential candidates to address. First was education (27 percent). Health care, Social Security/Medicare and taxes were tied for second at 22 percent each.
This suggests that while a significant percentage of Americans view taxes as the most important issue, the vast majority of Americans put other issues higher on the list of their concerns. This attitude is in sharp contrast with prevailing views in the tax revolt era, when people seemingly wanted their taxes cut regardless of the consequences. Following is a New York Times reporter’s summary of a national poll in the wake of Proposition 13 in 1978:
Fifty-one percent of those interviewed in the Times poll said they would vote for a measure like Proposition 13, which cut property taxes and placed a limit on future increases in California.
A majority said they would prefer a smaller government and fewer services if it meant a lower tax rate, but many voters do not believe they really have to make a sacrifice. Time and again, they said that costs could be cut by "trimming fat" and that taxes could be reduced without reducing services.
Clearly, nothing like this sentiment exists today. Indeed, even self-identified Republicans believe that increasing spending for education and Medicare is a better use for the budget surplus than cutting taxes for businesses and individuals, according to a 1999 Wall Street Journal/NBC News poll.
The irony is that taxes are higher and rising faster today than they were at the time of Proposition 13. As Figure 1 illustrates, in the 26 quarters leading up to Proposition 13, federal taxes as a share of GDP were unchanged. They were 19.1 percent in the first quarter of 1972 and 19.1 percent in the second quarter of 1978. Over this entire period, taxes peaked at 20.1 percent of GDP in the third quarter of 1974, falling sharply thereafter as a result of the 1973-74 recession and the 1974 tax rebate.
By contrast, over the past 26 quarters — from the first quarter of 1993 to the second quarter of 1999 — federal taxes have increased significantly, from 19 percent of GDP to 21.9 percent. By a significant margin, in the entire history of the United States, including wartime, federal receipts have never taken a higher share of GDP. On a quarterly basis, the highest percentage ever recorded prior to the Clinton administration was 20.7 percent of GDP, achieved in the first quarter of 1969 as a result of the 10 percent surtax imposed that year, and in the first two quarters of 1981, just before the Reagan tax cut took effect. At the peak of World War II in 1943, taxes only consumed 19.9 percent of GDP, and at the peak of the Korean War in the first quarter of 1951, they only took 20.1 percent.
It seems clear that the sensitivity people feel toward high taxes is unrelated to the actual tax take, as conventionally measured. Clearly, there must be some other explanation for the tax militancy of the 1970s and the tax passivity of the 1990s.
Wealth and income
CONOMISTS SPEND a lot of time analyzing the so-called wealth effect. However, their analyses generally are limited to calculating how much additional spending will result from a given increase in household wealth. Many economists believe that the strength of consumer spending and, hence, economic growth over the past several years is due mainly to the wealth effect.
No analysis that I am aware of has looked at the impact changes in wealth may have on other economic variables — such as the sensitivity of individuals to taxes. But it may be that such changes help explain how people feel about the taxes they must pay, and thus the political consequences.
Generally speaking, taxes are paid out of incomes rather than wealth. (The estate tax is an exception.) Thus it makes sense to analyze taxes as a share of income, rather than as a share of wealth. But for most people, increases in wealth appear as more income. This is especially the case since more and more pensions have been converted from defined-benefit to defined-contribution plans, such as 401(k)s and iras. Under the former, increases in the stock market have no impact on people’s benefits, and, in any case, it is difficult for people to calculate from year to year the present value of any changes in their pension benefits. With defined-contribution plans, however, people see the direct effect of increases in the stock market in the balances of their retirement accounts, and they can determine the exact dollar amount almost instantly.
Furthermore, as corporate equities have grown as a share of household wealth, largely at the expense of housing, it is easier for people to see changes in their net worth on an annual basis. In just the past 10 years, corporate equities have more than doubled, from 10.5 percent of all household assets to 24.3 percent, and from 17.2 percent of financial assets to 34.6 percent. And the Securities Industry Association estimates that the percentage of households owning stock has risen from 19 percent in 1983 to 48 percent today.
With wealth changes easier to calculate, many people now view increases in net wealth as equivalent to increases in income. This is consistent with the definition of income that many economists have long advocated for tax purposes; they would treat changes in net worth, gains or losses, as ordinary income. Therefore, it may be more appropriate to look at taxes not just as a share of GDP, but as a share of GDP plus the increase in net worth over the previous year.
As Figure 2 shows, taxes as a share of GDP plus the year-to-year change in household net worth have fallen from 18.3 percent in 1990 to 15.1 percent in 1998. This is due mainly to the fact that net worth rose by $3.7 trillion last year, 43 percent of GDP, but only $431 billion in 1990, less than 8 percent of GDP.
Because a major share of net worth is held in forms that are not subject to annual income taxes — assets in pension funds, unrealized capital gains, home equity, etc. — taxes rise more slowly than the rise in net worth. By contrast, because of progressivity, taxes rise faster than incomes. That is why taxes as a share of GDP plus the rise in net worth have fallen, while taxes as a share of GDP have risen.
If people instinctively interpret changes in net worth as changes in their incomes, then total incomes are actually rising much faster than they appear in most government statistics. That has had the effect of reducing the effec-tive tax rate well below what is suggested by federal receipts as a share of GDP.
As long as net worth keeps rising at a healthy clip, people will probably remain relatively unconcerned about rising taxes. But if net worth should drop, then the effective tax burden will rise sharply. That is what happened in 1994, when the value of corporate equities and mutual funds held by households fell by almost $50 billion after rising more than $400 billion the year before. This may help explain the Republican takeover of Congress that year.
Change (for the better) in tax structure
NOTHER EXPLANATION for the decline in support for tax cuts has to do with the changing nature of our tax structure. Two facts stand out. First, the tax burden has shifted more and more heavily toward those with upper incomes. As a consequence, the effective tax rate on those with modest incomes has fallen even as the aggregate tax burden has reached record levels. Second, although effective tax rates on those with upper incomes have risen in recent years, marginal tax rates have fallen sharply from those prevailing during the tax revolt years of the 1970s.
Figure 3 presents data on the distribution of the total federal income tax burden by percentiles. From 1980 to 1997, the share of total federal income taxes paid by just the top 1 percent of taxpayers (ranked by adjusted gross income) rose from 19.05 percent to 33.17 percent. The top 5 percent of taxpayers now pay a majority of all federal income taxes. Their share has risen from 36.84 percent to 51.87 percent over the same period. And the top 25 percent of taxpayers have increased their share from 73.02 percent to 81.67 percent.
Ironically, the effective tax rate on those with high incomes is substantially lower today than it was when their share of total income taxes was much smaller. According to the Internal Revenue Service, the effective tax rate on the top 1 percent of taxpayers was 34.47 percent in 1980. It fell to a rate of 23.14 percent in 1990, but has risen since to a rate of 27.64 percent in 1997. For the top 10 percent of taxpayers, the effective rate was 23.49 percent in 1980, falling to a low of 18.43 percent in 1991, and now stands at 21.36 percent.
As a consequence, tax rates on those with low incomes have fallen, and have even become negative for many because of the refundable Earned Income Tax Credit. According to the Congressional Budget Office, the effective income tax rate for those in the bottom quintile has fallen from -0.2 percent in 1981 to -6.8 percent this year. For those in the second quintile, the effective federal income tax rate has fallen from 11.3 percent to 0.9 percent. For all families, the rate has declined from 12.6 percent to 11.1 percent.
Thus we see that even for the wealthy, effective tax rates are down from their peaks, although they have risen lately. But more important from an economic (and perhaps psychological) point of view, marginal tax rates (the rate on each additional dollar earned) are down very sharply. In 1981, the top federal income tax rate was 70 percent (50 percent on labor income). Today it is 39.6 percent (42.5 percent on labor income). Although this is up from the 28 percent top rate that existed from 1987 to 1991, marginal rates are still significantly below their peak even for those with the highest incomes.
Another indication of how marginal tax rates have fallen is that in 1980, only 9.2 percent of taxpayers faced a marginal tax rate of 16 percent or less; in 1995, 71.9 percent of taxpayers were in the 15 percent bracket, the lowest marginal tax rate. In 1980, 24.7 percent of taxpayers faced a marginal rate of more than 28 percent; by 1995, this figure had fallen to 4.2 percent.
Figure 4 illustrates the impact of changing marginal and effective tax rates on a family of four with twice the median income (about $110,000 this year). In 1981, such a family had an effective federal income tax rate of 19.11 percent and a marginal tax rate of 42.46 percent. According to the Treasury Department, this year that family’s effective rate is down to 14.11 percent and its marginal rate is 28 percent. The decline in the effective rate is much less when higher Social Security taxes are included, but because the payroll tax (except for the Medicare tax) cuts off at an income of $72,600 for each earner, its inclusion has only a small impact on marginal rates.
Finally, it is important to note that more of the tax burden today is borne by consumption and labor income, rather than more economically sensitive capital income. The top capital gains tax rate, for example, is just 20 percent today versus 49 percent from 1969 to 1978, and 28 percent from 1978 to 1981 and 1987 to 1997. Furthermore, inflation is far lower today than it was in the late 1970s and early 1980s, when the rate of consumer price increases reached double digits. Inflation has an extremely adverse impact on capital gains because taxes apply to nominal gains. Thus, real effective tax rates on capital gains can easily exceed 100 percent during periods of high inflation.
In conclusion, it may be that while the total burden of taxation has grown sharply, its impact is little felt by most taxpayers. As more of the burden is carried by those with upper incomes, average and even above average earners have seen declines in both their average and marginal tax rates. Although average rates and marginal rates for the wealthy have risen of late, top earners are still substantially better off on both scores than they were in the late 1970s and early 1980s. Thus it may be that the tax structure has changed so as to make the current level of aggregate taxation more bearable, thereby diminishing pressure for tax reduction.
Budget surplus as tax cut equivalent?
NE OF THE REASONS REPUBLICANS have had difficulty getting traction on tax and budget issues is because they have failed to update either their rhetoric or their economic model to account for an era of budget surpluses. If they are to make progress, they must do both.
At the rhetorical level, Republicans are prisoners of an unfortunate decision made long ago to paint deficits as unmitigated economic evils. The truth is that vast amounts of economic research have failed to show any consistent relationship between budget deficits and inflation, interest rates or economic growth. What really matters is the overall size of government. Big government generally is bad for growth. Inflation and interest rates, on the other hand, are almost entirely functions of Federal Reserve policy, which operates independently of fiscal policy.
Explaining why big government is bad, however, proved to be too difficult for Republicans. It was much easier to attack budget deficits, because people instinctively view them as bad. For many years, it mattered little that Republicans were attacking the wrong target. That is because deficits and big government tended to grow together. And indeed, insofar as people think that programs paid for with deficits cost nothing in terms of higher taxes, they may view such programs as costless. Hence, deficit spending may encourage growth in government.
A major challenge to this view was put forward in the 1970s by economist Robert Barro of Harvard. He postulated that people really are not fooled into thinking that programs paid for with deficits are costless. Instead, they view deficits as deferred taxes, reacting exactly as they would if the new programs were financed with higher taxes instead of government bonds. Barro called this theory the Ricardian equivalence theorem, after the great economist David Ricardo, who first suggested the idea.
Ricardian equivalence is still controversial among economists, but it can be tested. For example, if you believe that taxes will be higher in the future than today, then you will tend to reduce your consumption and increase your saving now in order to pay that bill when it comes due. Thus we observe that higher deficits are often associated with higher private saving. It also explains why deficits do not appear to be stimulative, as Keynesian economic theory supposes.
Interestingly, almost all analysis of Ricardian equivalence has been under conditions of budget deficits; very few have dealt with surpluses. But if the theorem holds at all, we should expect the opposite effects under surpluses. This means that people implicitly view surpluses as de facto tax cuts. That must be the case if they view deficits as equivalent to taxes. And it also means that people will tend to reduce their savings and increase their consumption — under Ricardian equivalence, the opposite of what deficits produce.
In fact, as Figure 5 shows, we now see people acting precisely in accordance with the predictions of Ricardian equivalence. The personal savings rate has collapsed to historically low rates. Meanwhile, consumers are spending like there is no tomorrow, and economic growth is exceeding all expectations. Furthermore, Ricardian equivalence explains why taxpayers seem so lukewarm to the idea of a tax cut and why the notion of paying down the national debt is popular. If Ricardian equivalence holds, then paying down the debt is exactly the same as getting a tax cut.
Polls support this hypothesis. Many polls show paying down the national debt to be growing in popularity. A February Associated Press poll found 49 percent of respondents favoring a tax cut and 35 percent wanting debt reduction. An abc News poll in July found 19 percent of Americans saying that the top priority for the budget surplus should be debt reduction, versus 22 percent favoring a tax cut. Another July poll by Hart/Teeter found 51 percent of Republicans more likely to support a presidential candidate whose first priority was paying down debt, with 42 percent supporting a candidate who would cut taxes.
An August 1999 ABC News/Washington Post poll found greater support for debt reduction over tax cuts by a 24 percent to 20 percent margin in one question and 19 percent to 14 percent in another. In September, an NBC News/Wall Street Journal poll found 25 percent preferring debt reduction, 20 percent cutting taxes.
It is ironic that for so many years efforts to cut taxes were thwarted by the existence of large budget deficits, but now that we have budget surpluses the prospects for tax cuts have not improved. Whether voters are reacting to Ricardian equivalence, it is clear that surpluses are a major barrier to tax reduction efforts. It may be — a further irony — that the federal government must return to deficits, perhaps as the result of a recession, before the prospects for tax reduction will improve. After all, the Reagan tax cut of 1981, one of the largest in U.S. history, took place at a time when the budget was in deficit.
F THERE ARE SOME compelling economic explanations for reduced support for tax cuts, there are political reasons as well. One may be that taxpayers simply do not believe they will get a tax cut. Politicians are widely viewed as willing to say anything to get a vote, including promising tax cuts, but unwilling to deliver once elected. Voters may also feel that even if a tax cut is enacted, it probably would not benefit them personally.
On the first point, a Fox News/Opinion Dynamics poll in March asked registered voters if they believed politicians who promise to lower tax rates. An overwhelming 87 percent said no, with just 9 percent saying yes. This result may be due to George Bush’s 1988 promise not to raise taxes, which he violated in the 1990 budget deal. Voters may also remember Bill Clinton’s 1992 promise of a middle class tax cut, which became a tax increase the following year.
On the second point, the same poll asked people whether they thought they would benefit from "targeted tax cuts." Only 18 percent thought they would benefit; 62 percent said someone else would gain. This perspective may result from the experience of the 1997 tax cut. The vast bulk of the tax reduction in this legislation was a $500 child credit, which by definition yielded no benefits to anyone without young children. Benefits were further limited by income tests that phased out the credit for families with high incomes — which also created high de facto marginal tax rates across some income ranges.
If these points are valid, it suggests that politicians favoring tax cuts must find a better way of ensuring that their promises will be kept, or at least that they will make the greatest possible effort to keep them. It also suggests that broad-based tax cuts are better politically than the targeted approach of both Republican and Democratic tax reduction efforts in the 1990s.
Waiting for a better case
HILE THERE IS NO QUESTION that support for tax reduction has declined since the 1980s, it would be wrong to conclude that opinion has shifted in favor of high taxes. Polls are clear that people still think taxes are too high. For example:
A CNN/Gallup/USA Today poll in July asked people if they thought their federal income taxes were too high. Sixty percent said yes, 37 percent said about right, and none said they were too low.
A Zogby poll in August asked likely voters a similar question. The answer was the same. Sixty percent of respondents said their taxes were too high and 37 percent said they were about right.
Reinforcing this point are poll data showing doubts people have about the value they are receiving from government for their tax dollars. A May 1999 poll by the Council for Excellence in Government asked people if they thought they were paying too much in taxes for what they get from government. Forty-six percent agreed strongly. Only 19 percent felt strongly that they got their money’s worth from their taxes.
Furthermore, support for fundamental tax reform, such as the flat tax, remains high and there is strong support for specific tax cuts, such as eliminating the marriage penalty. Consider:
A Harris poll in March asked people if they favored radical tax reform or the status quo. Forty-four percent said they favored a completely different tax system, 31 percent favored moderate changes, and only 21 percent supported the status quo. The same poll asked people if they would favor a flat tax in which everyone paid the same tax rate above some minimum. Sixty percent favored such a system and only 35 percent opposed it.
An April poll by Reuters/Zogby found 74 percent of likely voters agreeing with the need for a 10 percent across-the-board tax cut and only 25 percent disagreeing. The same poll found 60 percent supporting a flat tax and only 31 percent against.
An August poll by Wirthlin Worldwide listed all of the key provisions of the congressional tax bill and asked registered voters if they supported such a tax cut. Thirty-nine percent strongly supported it, 29 percent somewhat supported it, 14 percent somewhat opposed it, and only 17 percent strongly opposed it. The same poll asked specifically about eliminating the marriage penalty and the estate tax. Seventy-nine percent either strongly or somewhat favored abolition of the former and 70 percent favored getting rid of the latter.
A September poll by Market Strategies asked people about several specific tax proposals. Seventy-two percent favored expanding Individual Retirement Accounts, 64 percent favored a 1 percent cut in each income tax rate, 62 percent said they favored abolition of the estate tax, 61 percent supported a cut in the capital gains tax, and 57 percent said elimination of the marriage penalty was needed. (These are, of course, all of the major provisions of the tax bill that passed Congress in August and was vetoed by President Clinton.)
This analysis suggests that supporters of tax cuts need to be more sophisticated than they have been. They cannot just declare their support for tax cuts and expect hungry voters to react as if they are being thrown red meat. Voters have been burned by too many false promises, are generally more contented as the result of tax changes made in the 1980s that are still in effect, and want greater specificity in terms of exactly how taxes will be cut and who will benefit. Tax cuts that are too narrowly targeted leave too many taxpayers out of luck. Tax cuts that benefit broad classes of taxpayers appear to be more popular.
While the tax revolt may have gone into hibernation, it would be a gross error to think that people have swung in the opposite direction and now favor high taxes because they like big government. If one adds support for tax cuts to support for debt reduction, the resulting percentage dwarfs that for expanding government programs — even those with the greatest public support, such as education.
Currently, there is no reservoir of voter sentiment in favor of tax cuts — something politicians can simply tap into by proposing cuts. But there is no indication that the subject is permanently closed for people, either. Circumstances may change, or the effectiveness of the case politicians make for tax cuts may improve. The latter, however, must begin with an understanding of the economic and political forces that may be shaping people’s current attitudes.
Read Policy Review's Statement of Purpose
Subscribe to Policy Review or call 1-800 566-9449.
Send a Letter to the Editor or write our editorial office:
Policy Review, 214 Massachusetts Avenue NE
Washington DC 20002 Phone (202) 546-4400 Fax (202) 608-6136