Perspectives on the Tax Stimulus Debate

William G. Gale


Mr. Chairman and Members of the Committee:

Thank you for the opportunity to present my views on the tax stimulus issues currently facing the Congress. Congress faces several key issues: whether to provide a stimulus; the size of any stimulus package; the composition between spending and tax cuts; and the design of the spending and tax cuts. This testimony focuses mainly on the most effective design and the potential impact of tax cuts.

The Economic and Budget Outlook

If a stimulus package is desired and if tax elements are a part of the package, the tax elements should be designed with the current economic and budget outlook in mind. The short-term economic outlook is uncertain, as the terrorist attack on September 11, 2001 having disrupted the workings of an already slowing economy. Despite the short-term uncertainty, there is widespread agreement that long-term economic prospects remain strong.

In contrast, the long-term budget outlook has deteriorated rapidly. Based on data from a recent bi-partisan statement from the Senate and House Budget Committees, the 10-year baseline budget surplus outside of social security appears to have shrunk by about $3 trillion (or 98 percent) since May, and currently stands at just $50-$100 billion, not counting any stimulus package or any other new proposals that Congress may wish to consider over the next decade.

Reasonable allowances for new spending that is widely supported (including defense, education, and prescription drugs) the extension of expiring tax provisions, and providing a fix for the AMT puts the non-social security budget in deficit to the tune of about $1.5 trillion over the next decade.

Principles for Design of an Economic Stimulus


Because the economic outlook suggests the possible need for a short-term boost, while the budget outlook suggests that the long-run revenue impact of any stimulus should be limited, the key guiding principle for a stimulus package, should be to maximize the “bang for the buck.” That is, the goal should be to get the most short-term boost at the least long-term cost. To achieve these goals, policy makers should focus on provisions that:

  • Encourage new household spending.
  • Encourage new business investment.
  • Minimizes long-term costs
  • Are temporary—that is, expire within about one year.

These principles are quite similar to those endorsed in a remarkable bi-partisan statement by the Democratic and Republican leaders of the Senate and House Budget Committees earlier this month. The principles represent sound policy in a number of key dimensions.

New household spending and new business investment provide direct stimulus to the economy. In contrast, raising household saving would reduce current aggregate demand. Likewise, subsidizing investments made in the past (and currently showing up as corporate income) does not provide any stimulus. To state what may be obvious, making tax cuts retroactive to before September 11, 2001, not only has no benefit whatsoever in stimulating the economy, it is plainly wasteful.

Minimizing the long-term costs serves both short-term and long-term purposes. In the short-term, lower budget costs translate into smaller boosts in interest rates. Increases in interest rates would serve to reduce business investment, housing purchases, and other interest-sensitive consumption and thus would choke off part or all of the direct effect of any stimulus package. Minimizing the revenue costs is also important for the long-term, even if government budgets have no effect on interest rates. It is clear that the underlying fiscal situation has changed dramatically over the past six months. Whatever one thought of the affordability of the tax cut last spring, it is evident that the nation now has to devote substantial resources to fighting terrorism over the next decade. The deteriorating budget outlook means that, after resolving the stimulus debate, policy makers will need to rethink the long-term fiscal picture and thus should not do anything now that will make the long-term situation significantly worse.

Keeping all items temporary is critical for several reasons. First, it raises the bang-for-the-buck. Temporary business incentives typically have larger short-run impacts that permanent incentives, and cost much less. Temporary household tax cuts may have somewhat smaller short-term impacts than permanent cuts given to the same people. Even so, the bang-for-the-buck for temporary household cuts is likely to be much larger than for permanent cuts, since temporary cuts cost much less. Second, focusing on temporary items is equitable (why should some groups receive permanent tax cuts as part of stimulus package, while others receive only temporary tax cuts?). Third, focusing on temporary cuts reduces the likelihood of enacting permanent changes that are not related to economic stimulus and that may not be good tax policy in any case.

It is also worth emphasizing that stimulus can be provided via new government spending on purchases of goods and services or via increased government transfer payments, such as unemployment insurance.

Applications to proposals for household tax cuts

A rebate to low- and middle-income households would be the most effective way to provide a tax stimulus. These households are more likely to spend any new income than high-income households are. The rebate checks this summer do not appear to have generated much in the way of new spending, perhaps because they were not targeted to low-income households.

Accelerating the previously enacted tax cuts is a poor way to provide stimulus. Most of the costs occur in years beyond 2002, when most commentators believe the economy will already have recovered. Almost all of the benefits go to higher-income households, who are less likely to spend the funds. The proposed accelerations would be expensive relative to the stimulus “budgets” proposed by President Bush, Fed Chairman Greenspan, and former Treasury Secretary Rubin. Finally, enacting the previously enacted tax cuts will have the effect of locking in a tax cut whose wisdom will need to be re-examined in the near future, in light of the substantial deterioration in the budget.

Capital gains tax cuts are at best an inefficient way to stimulate the economy in the short-turn, and could well be counterproductive. Capital gains tax cuts are usually advocated on the grounds that they will stimulate saving and long-term growth, not a short-term consumer spending boost.

Applications to proposals for business tax cuts

Temporary incentives for new business investment (e.g, accelerated depreciation or an investment tax credit) are the business tax cuts most consistent with the principles listed above. These tax cuts target new economic activity and, given their temporary nature, they would have a bigger short-term impact and a smaller long-term cost than permanent incentives.

Corporate tax cuts are, in general, a poor way to stimulate the economy. Almost all of the funds would go to providing windfalls to current corporate income, which is the return to old investment, rather than giving incentives to new investment.

Repeal of the corporate alternative minimum tax is mainly a subsidy to old investments, and even advocates of repeal have shown that it would generate virtually no stimulus in the first two years of its existence.

Even if the AMT is repealed, there is no reason why such a repeal should be retroactive, as discussed above.

Temporary expansion of the carryback provision for net operating losses is also mainly a subsidy to old investments. It is different from corporate tax rate cuts or AMT repeal, however, in that expanding the lookback provisions is better targeted to helping firms that are currently losing money.

Issues in the design of a temporary investment incentive

Although temporary investment incentives are much more appropriate than permanent incentives for stimulus purposes, temporary tax cuts raise a number of issues worth considering in detail and applying to the current situation.

First, the historical record suggests that it is difficult to time fiscal interventions accurately and that past interventions have destabilized rather than stabilized investment flows over the business cycle. In addition, discussion of temporary incentives by policy makers may create a decline in investment before the incentives come into existence, as firms wait to see when the incentives will take effect. For both of these reasons, Congress should announce immediately that any investment incentives that are eventually included in a stimulus package would be apply to investments made as of today (or as of September 11, 2001).

Second, temporary incentives may also create a decline in investment after they expire. This may create demand for extending the expiring provision, which adds uncertainty to the tax system. But if it is known (or suspected) that Congress will extend the incentive when it expires, then the temporary incentive loses some of its appeal. A third problem is if temporary incentives are placed to stay in operation for too long, they will provide little incentive to invest now, when the investment is most needed. For example, the 3-year period for partial expensing passed by the House Ways and Means Committee is much too long. It provides little reason for firms to invest more now, and gives them plenty of reason to hold off for a year or two to gauge the economic landscape before investing. Casual evidence suggests that businesses are putting off projects they already had in the works, and the immediate goal should be to encourage businesses to undertake those projects in the immediate future, not 2-3 years from now.

To address the second and third problems, Congress should (a) shorten the time period for partial expensing to 15 months and (b) adopt a sliding scale for the amount of the investment that can be partially expensed. Suppose, for example, that Congress stipulated that firms making qualifying investments could write off 50 percent of expenses immediately, if the investment is made before the end of calendar year 2001, 40 percent if made in the first quarter of 2002, 30 percent if made in the second quarter, 20 percent if made in the third quarter, and 10 percent if made in the fourth quarter of 2002. This would focus firms on making new investments now, rather than delaying. And by phasing down the expensing portion slowly over time, Congress would make much more credible its intention to keep the incentive temporary.

Other issues regarding a stimulus package

At least three additional issues bear on choosing the magnitude and nature of any tax stimulus program. The first, of course, is the likely depth and duration of the current economic decline. As a specialist in public finance issues, I will leave macroeconomic forecasts to the experts in that field.

Second, whatever the magnitude of the short-term problem, it is worth noting that a significant amount of economic stimulus has already been provided, including $40 billion in rebates during the summer, $40 billion in spending for defense, rescue and recovery, and the airline bailout. In addition, the federal reserve has cut interest rates 9 times and by several hundred basis points since the beginning of the year. More stimulus is on the way, even if Congress does nothing. The tax cut passed last spring provides for about $70 billion in individual tax cuts during fiscal 2002.

Third, it would be appropriate to be realistic about the likely economic impact of the tax stimulus proposals being discussed. A rebate to households of, say, $20 billion, would only amount to 0.2 percent of GDP in stimulus even if the entire amount were spent, which seems unlikely. The proposed 30 percent partial expensing of new investments would reduce the user cost of capital by about 4.5 percent. Applying investment elasticities of between -1/3 and -2/3 suggests that investment in equipment and software would rise between 0.15 percent and 0.30 percent of GDP. The effect may be even more limited during times like the present-when firms already have significant cash-on-hand and significant excess capacity.