Madam Chairwoman and members of the subcommittee, it is a great privilege to have the
opportunity to appear before you today. I am an economist who works at the Washington-based think tank, the American Enterprise Institute. I have spent a good deal of my research time since I completed my dissertation studying the effects that tax credits have on the behavior of firms. I come to you today to focus on the provision of H.R. 2086 that would make permanent the Research and Development Tax Credit-to give you my reading of what the scholarly literature has to say about such a move.

Firms engage in R&D with the hope of developing new products that will generate healthy profits. R&D is a classic example of an activity that has external benefits: when a firm uncovers something new, the knowledge will help some other firm perform its own R&D. It is often hard to predict where these spillovers will occur, but it is generally accepted by those who study R&D that breakthroughs often create a kind of snowball effect, spreading innovation and productivity increases to the far reaches of the economy. Think, for example, of how many everyday appliances have been changed by computer technology. The benefits to society of R&D are likely to be higher than the benefits to individual firms doing research, since firms tend to look only at their own payoffs. Thus, without government subsidies, there would be too little R&D from society's point of view.

Currently, the R&D tax credit expires frequently and is renewed frequently. This state of affairs exposes firms to a great deal of uncertainty, uncertainty which likely leads them to respond less than they might otherwise to the credit-fearful, for example, that a costly research investment would become too much of a burden if the credit is not renewed. Despite this, the successive temporary credits have likely had a large positive effect on R&D spending and economic growth. A recent literature survey conducted by the National Bureau of Economic Research concluded that every dollar of revenue lost because of the credit added about a dollar to aggregate R&D spending; a panel of experts convened by the Office of Technology Assessment recently reached the same conclusion. Most readers of the literature agree that the benefit-cost ratio for the existing R&D credit is somewhere between 1 and 2, making it one of the most successful government tax provisions on the books.

If the credit were to become permanent, then the benefits could well be higher, since the uncertainty surrounding its renewal would be removed. Exactly how much stimulation would come from the change is uncertain. A recent study by Coopers and Lybrand estimated that making the credit permanent would stimulate an additional $41 billion of R&D spending between 1998 and 20 10, relative to a baseline in which there is no credit. Since this spending has a high rate of social return, the benefits to society of the change

would likely be impressive. Coopers estimates that the economy will produce $58 billion worth of new goods over the period between 1998 and 2010, if the credit is made permanent. The study concludes that the credit will actually pay for itself, given the long run growth effects, and seems, to me, to be quite consistent with the conclusions any informed observer would draw from the literature as a whole.

On the other hand, the actual economic benefits might be fairly small relative to a baseline that assumes we pass a sequence of short-term credits in the coming years. While this might seem like an alternative system that makes little sense, it seems, sadly, to be the system we have adopted in the past. My own informal survey of tax professionals found that nearly everyone believes that the credit is effectively permanent, even though it technically expires almost every year. If firms already believe that, then the effects of formally making the credit permanent would be much smaller.

The best argument for making the credit permanent is probably that it would help, in a small way, restore faith in our tax system. It is certainly unusual, given the general agreement that R&D subsidies are a sensible way for government to spend revenues, that the current R&D credit expires frequently and is renewed frequently. Since it was first enacted about eighteen years ago, it has expired ten times. This odd state of affairs feeds cynicism about the tax code, in addition to undermining the effectiveness of the credit itself. For example, Steven Moore of the Cato Institute made the following observation about the credit in the Wall Street Journal: "Members of Congress like to use the tax code as a way to milk money from the business community ... By making it temporary, they can milk them year in and year out."

My own sense is that the R&D credit has been an odd byproduct of the budget process, with its illusory temporary nature making it easier to meet deficit targets in the out years. Given the current surplus situation, members have a historic opportunity to make a sensible and effective credit permanent.

Thank you for your attention.