Tax Policy
Len Burman[*]
Comments published in
American Economic Policy in the 1990s,
edited by Jeffrey Frankel and Peter Orszag
(MIT Press: Cambridge MA, 2002).
Gene Steuerle has done a fine job of documenting what happened in tax policy in the 1990s. Gene is ideally suited to the task, having literally written the book on tax policy in the 1980s—The Tax Decade. This first installment on the sequel is particularly appropriate and revealing.
I agree with Gene’s analysis, but I’d like to emphasize a few points and elaborate on several others. I will start by talking a bit about the big picture—in particular, the new budget paradigm that distorted tax and budget policy in the 1990s. I will then summarize some of the good and bad things that came out of tax policy in the 1990s. Finally, I will conclude by reflecting a little bit on the process.
The Importance of Budget Rules
Gene talked about the effect of budget rules in tax policy. I think that may be one of the most important features of tax policy in the 1990s. The budget rules were designed to constrain spending, and the overall budget caps clearly helped accomplish that.
But they also altered the terms of trade between direct expenditures and tax expenditures. They required legislators to pay for a spending increase (above legislated discretionary caps) by cutting other spending. New tax expenditures, however, could be paid for either by a cut in spending or by an increase in other taxes. That meant that, on the margin, new spending was easier to accomplish on the tax side of the ledger than on the expenditure side.[1]
Consequently, new programs were more likely to be designed as inefficient tax expenditures than as direct expenditures, even if the spending program would be more effective and efficient. Much of the debate in the Clinton administration surrounded that tradeoff.
That process had a perverse effect on public policy. If one assumes that the budget caps were effective, then the overall size of the government budget was fixed. Moreover, the amount of discretionary spending was also fixed (assuming that the discretionary caps were binding—a safe assumption). The only variable was the amount of tax expenditures, which were permitted to be financed by offsetting taxes. Thus, for any given size of government, a larger share of spending had to be done through inefficient tax expenditures rather than through more efficient direct expenditures. In this model, the rules that distinguish tax expenditures from direct expenditures reduce the efficiency of government without affecting its size or the overall tax burden.
If the budget constraint were not fixed, limiting marginal policies to ones that don’t work might reduce the odds that those policies would be enacted. Conceivably, that could reduce the overall budget. Under the further assumption that, unconstrained, government would be too large, that trade-off might be efficient in a second best sense.
I am reminded of an argument made by Michael Best and William Connolly (1976)—two socialist political scientists—in The Politicized Economy. They complained that our two-party system would never work effectively because Republicans hate government and Democrats love any new government program. So they compromise on programs that don’t work. That makes the Republicans happy because it validates their view that government doesn’t work. And it makes the Democrats happy because they can claim credit for all the new programs.
Arguably, Republicans offered Democrats a larger public sector so long as all the growth was on the tax side of the budget. That is, they compromised on programs that don’t work very well.
Good and Bad Features of Tax Policy in the 1990s
Fiscal discipline was, I believe, the great accomplishment of the 1990s. It clearly was both a result of tax policy and also a major factor in the development of tax policy.
In the beginning of the decade, deficits were large and growing. By 1993, debt reached fifty percent of GDP. Because of the 1990 and 1993 tax increases, spending restraint, and the longest economic expansion in history, we managed to eliminate the deficits and also produce a humorous debate about the risk of surpluses that no one could have imagined ten years ago. (Burman 1999)
President Clinton deserves a lot of credit for that accomplishment. He stood up to people in his own party who wanted to increase spending and to Republicans (and some Democrats) who wanted to cut taxes when he committed to balance the budget. Of course, that commitment was not new. Ronald Reagan had made a similar pledge. What was new was the resolve to make the tough choices necessary to actually improve the budget situation, and then to close the deal when favorable budget projections showed that surpluses were actually in sight. (Waldman 2001)
By the time I got to Treasury in 1998, surpluses were a reality, and it seemed certain that nothing could prevent a large tax cut. President Clinton and his economic team, however, managed to maintain the political momentum for fiscal discipline by inventing and endlessly repeating the mantra, “Save Social Security First.” Orszag, Orszag, and Tyson (this volume) point out that the Social Security plan probably did not advance the prospects for meaningful reform, but it did fundamentally alter the tax and budget debate. Members of both parties eventually agreed that the portion of the surplus attributable to Social Security should be considered off limits both for new spending and tax cuts. That made it possible for President Clinton to sustain a veto of several significant (and sometimes politically popular) tax cuts in the last two years of the Administration, and to deter other tax cut proposals.
In addition, some worthwhile tax policies were enacted in the 1990s. Most notably in my view, taxes were cut substantially for low-income families. The earned income tax credit was expanded twice—in 1990 and 1993. The child credit, enacted in 1997, also helped millions of lower middle-income families.
The Clinton Administration also made some proposals that would have been very positive had they been enacted. When I was at Treasury (from 1998 to 2000), we proposed real relief from the alternative minimum tax (AMT) that effectively would have kept children from being treated like tax shelters for purposes of the AMT. It would have cut the number of people on the AMT in half.[2] By comparison, the tax bill that was just enacted would effect a tiny and temporary reduction in the number of people on the AMT. Moreover, it sunsets after 2004. By the end of the decade, the Joint Committee on Taxation projects that the number of people on the AMT will double. That’s a serious problem.
But, the tax system also changed in many ways for the worse. First of all, we continued the pattern of frequent tax changes. Major tax bills were enacted in 1990, 1993 and 1997. Martin Feldstein (1976) pointed out that frequent tax changes are costly because they make it hard for businesses to plan, and the transition from one set of rules to another is costly. Tax changes also affect individuals. Politicians’ compulsion to do something all the time—and especially right after presidential elections—means it’s very hard for taxpayers to actually keep track of what the tax system is at any point in time.
Unfortunately, the situation does not look any more sanguine for this decade. The bill that was just enacted basically legislates a major tax change every year for the next nine years. And then it’s repealed. It’s safe to predict that taxpayer confusion will be the norm for some time.
In the 1990s, there was a proliferation of hidden taxes. Although they weren’t invented in the last decade, the technology clearly advanced. Most notably, the phase-out of itemized deductions, enacted in 1990, adds over one percentage point to effective tax rates for most taxpayers with incomes over $100,000. The phase-out of personal exemptions, enacted the same year, adds a surtax based on family size.[3] These are hidden taxes that almost nobody understands.
There was a strange migration of education policy from the spending side of the budget to the tax side. And there were dozens of proposals for new tax expenditures. Most weren’t enacted, but unfortunately those proposals are going to remain a threat to sensible tax policy for a long time. Indeed, President Bush embraced several in his budget and energy proposals.
Nonetheless, I don’t share Henry Aaron’s total dismay about the unraveling of the Tax Reform Act of 1986 (TRA). At the end of the decade we had lower tax rates than we did before TRA. And much of the base broadening enacted in 1986 survived the 1990s. But we definitely were moving in the wrong direction.
The deal in 1986 was broad base and low rates, but rates increased and new loopholes were created in the 1990s. As Steuerle noted, the new loopholes were mostly on the individual side of the budget, which is important. In contrast, the loopholes in the 1980s, which were eliminated by the TRA, were mostly on the business side.
The narrow base and high rates that we were trending toward is probably the worst alternative from a tax policy perspective. For example if we had a broad base with high rates, there wouldn’t be many opportunities for tax shelters, even though people would want to engage in them. We would raise a lot of revenue. As a result, the additional government savings could undo some of the damage done by having high tax rates on individuals. Although a narrow base with low rates wouldn’t raise much revenue, there wouldn’t be much incentive to engage in tax shelters and avoidance if rates are low. But a narrow tax base with high tax rates, as Steuerle points out, violates every principle of tax policy. It creates all sorts of inequities. Probably more important, it makes people feel that the tax system is unfair and that they are being treated unfairly. It’s inefficient and it’s complicated.
From my perspective, one of the worst things that comes from the proliferation of targeted tax incentives is that they create a huge constituency for tax complexity. The next tax reform will be a lot harder to attain than the last one. We can’t replace a lot of targeted tax preferences with lower tax rates without raising taxes on some people.[4] And many who benefit from a special provision oppose reform in that area even though they’d like to reform other people’s tax breaks.
Politicians certainly treat this perception as a constraint. The most recent tax bill illustrates the case. Congress had an opportunity to eliminate one of the most complex provisions affecting lower income people: FRED. FRED refers to Full Refundability for Excess Dependents. (It is actually partially refundable.) Under FRED, a family with three of more children could get a refundable child credit to the extent that the employee share of Social Security taxes plus individual income taxes exceeded its earned income tax credit up to the amount of the full child credit. It was very complicated. Not many people benefited from it, but it did have a cute nickname.
Largely because of very effective analysis and lobbying by people at this conference, including Bob Greenstein, Jeff Leibman, and Gene Steuerle, the Senate added a new, much more generous, provision for refundability of the child tax credit to the 2001 tax bill. When it is fully phased in, the child credit will be refundable up to 15 percent of earnings above $10,000. Although it’s not totally transparent, it’s a lot simpler than FRED. It was obviously time to pull the plug on FRED. The new refundable credit would cut taxes for more than 90 percent of those who currently benefit from FRED—and for millions more who don’t. It would make the tax system fairer and simpler.
But repealing FRED would violate the rule that nobody can pay higher taxes. A very small number of people would be worse off under the new credit. To avoid this, Congress agreed to let taxpayers calculate their refundable credit under the new rules and compare that credit to the amount that they could claim under the old rules (FRED) and take the larger credit. In other words, they get the complexity of FRED plus the complexity of the new proposal. We now have not only an alternative minimum tax, but also an alternative refundable child credit. This is supposed to be something that people with a high school education or less could figure out.
At the same time that targeted tax incentives make reform less likely, they also make it much more important. People come to view the tax system as hopelessly complex and unfair, which has bad effects. It undermines support for fair progressive taxation. And it even undermines the new tax incentives themselves because people can’t respond to an incentive that they don’t understand.
The Tax Policy Process in the Clinton Administration
I’m going to talk a little bit about the process. I saw the process more than the results because no significant tax legislation was enacted while I was at Treasury. The interesting thing about the process was that it was at least a pretty good caricature of enlightened economic decision making. Public finance principles say that you shouldn’t alter market prices unless there is a market failure. Many of the discussions that we had within the Administration were about whether externalities or other market failures existed that could justify a new tax incentive.
Unfortunately, that didn’t help quite as much as one might imagine, because advocates of tax incentives were incredibly innovative in finding market failures. But some really dumb proposals—such as the tax credit for corporate responsibility alluded to in Orszag, Orszag, and Tyson (this volume)—were ultimately rejected because they could not meet minimal standards of public finance.
Another thing that was interesting and novel about this last decade was that Larry Summers, an economist, was the Treasury lead on tax policy for the last half of the Clinton administration. That was an unusual and historic event. And it indicated economics was on the table as part of tax policy making.
The NEC process, as described in Orszag et al, was another innovation aimed at improving economic policy making. The NEC process was good in the sense that it greatly facilitated communication and coordination. Lone rangers were not generally running off doing their own thing. The NEC also did a great service by effectively selling the policies that we put together. That would have been hard for people at Treasury to do in a lot of cases. But the NEC played a much more active role at the end of the Clinton Administration than simply a neutral honest broker. It was clearly a leader in initiating ideas for new tax incentives, many of which would not have made the tax system better were they enacted.
There was a political perception that there had to be a tangible policy response to every problem that faced the nation.[5] Every summer Gene Sperling and the NEC staff would make a list of ideas that could address these problems. Proposals were grouped into broad theme areas such as education, economic development, housing, health care, worker training, bridging the digital divide, encouraging philanthropy, energy, and climate change. (Larry Summers added one on vaccines.) And program agencies would add numerous additional ideas.
The NEC was incredibly effective at creating ideas that sounded like they might work. But the problem was that the new tax incentives would have added great complexity to a system that people already don’t understand. Even though most of the really bad ideas were ultimately rejected as a result of the NEC process (informed by terrific staff work by the nonpartisan career staff in Treasury’s Office of Tax Policy), the constant flow of complicated proposals tied up Treasury staff and got in the way of necessary improvements to the models used for revenue estimating and distributional analysis, basic research that might have informed future policy development, and regulations to implement tax laws already passed.
There also was a sense that we had to have rapid reactions to issues that are on the table. I think tax policy is an extraordinarily poor instrument for rapid reaction to short-term policy problems.
And then there were some policy accomplishments that wouldn’t be worth winning. For example, after lengthy negotiations with auto manufacturers, the White House’s Council on Environmental Quality proposed a tax credit for highly fuel efficient “hybrid” vehicles. A hybrid automobile stores much of the energy typically lost in braking in a battery, which is then used to boost the car back to cruising speed after a stop. If widely adopted, such cars would clearly improve the environment and reduce greenhouse gas emissions. The problem was that American auto manufacturers would not be able to produce hybrid vehicles until at least 2003, while Honda and Toyota were ready to bring theirs to market in 2000. The US car manufacturers did not want to subsidize Japanese companies, so they insisted that the credit be delayed until 2003. Had the credit been enacted, it would have given American consumers an incentive not to purchase the most fuel-efficient cars in the market.