April 28, 2006
js-4222
Remarks by Robert Carroll
Deputy Assistant Secretary for Tax Analysis
U.S. Department of the Treasury
Before a Conference at
The James A. Baker III Institute for Public Policy
Is it Time for Fundamental Tax Reform? The Known, the Unknown,
and the Unknowable
Tax Policy and the Dynamic Analysis of Tax Changes
Thank you very much for the opportunity to speak before you today.
Before I begin my remarks I especially want to thank the organizers of this conference at the Baker Institute. The subject of tax reform is a very important subject and the Baker Institute should be congratulated for bringing together such a fine group of individuals to discuss this issue. I also want to personally thank George Zodrow and John Diamond for the fine work they have been doing for the Treasury Department on tax reform. As some of you know, the Treasury Department provided a dynamic analysis of the options the President's Tax Reform Panel presented to the Secretary back in November. George and John were essential for getting that work done and we continue to work with them as we embark on creating a new Dynamic Analysis Division at the Treasury Department.
Often when I attend these types of conferences, I listen to folks from academia provide their perspective on how the latest research informs, or, in some cases, should inform, the analyses that government economists provide to policy makers in
In the decade and a half that I have been in
When folks were working on the Tax Reform Act of 1986 a shift had already begun in the academic literature suggesting that households and businesses responded to tax rates in important ways. The rate reductions and capital cost recovery provisions in the 1981 reflected this view.
Prior to this period, however, saving was generally thought by many to be unresponsive to changes in its after tax return and business investment unresponsive to changes in the cost of capital. Labor supply decisions were not thought to be responsive to changes in marginal tax rates until the work of Jim Heckman in the early 1980s. Many other dimensions of labor supply were still largely unexplored. The research finding large responses for capital gains and charitable giving dates back a bit earlier. But, I think it fair to say that prior to the early 1980s, many, including those in government providing analysis to policy makers, had the view that a variety of economic decisions were not particularly sensitive to changes in marginal tax rates and certainly did not view saving and investment as particularly responsive to changes in its tax treatment.
The 1986 Act probably came as close as we ever have to the economists ideal of taxing comprehensive or Haig-Simons income. The tax base was broadened, made more encompassing, and focused on taxing income. Tax rates on individuals came down, tax rates on capital income went up. Income generally was more comprehensively and uniformly taxed. The goal was to tax the return to investment more uniformly to improve the intersectoral allocation of capital, with little focus on intertemporal dimension.
Through the 1980s and into the 1990s, the research continued to move in the direction of finding that individual and business decisions were more responsive to changes in tax treatment than previously thought.
What was the catalyst at work here. Research began to use large micro data files – first cross-sectional data, then longitudinal data – to more carefully examine and quantify how tax rates affect various individual and business decisions. Econometric techniques became much more sophisticated. I already mentioned the early literature on the responsiveness of capital gains and charitable giving using, incidentally, primarily tax return data residing at the Treasury Department. The labor supply literature was finding greater responsiveness for secondary earners. The earlier literature on the responsiveness of savings relied on time series data and generally found that savings was relatively unresponsive to changes its after-tax return. Detailed analysis of IRAs and 401(k)s, although still hotly debated, revealed considerable responsiveness. Importantly, investment has also been found to be more responsive than initially thought.
This research has been followed be a growing literature on the taxable income response, a more reduced form approach encapsulating a variety of taxpayer responses. There is also an emerging literature on entrepreneurship that finds lower tax rates increase the probability of becoming an entrepreneur, and increase the likelihood that they hire workers, invest, and grow. Of course, this is not to say that all elasticities tend to one or that there still does not remain considerable uncertainty over the results reported in many of these areas. And, in some areas, more recent research has shifted in the other direction, and found smaller responses.
This growing empirically based literature on the responses to tax rates has coincided with the development of complex numerically-based general equilibrium models of the economy that emphasize the intertemporal aspects of consumption, investment, and labor supply decisions. The research using these models has shown that the tax system influences key intertemporal decisions and can have profound effects on longer term economic growth and impose substantial economic costs. A series of papers has also focused on the intertemporal aspects of taxing capital income and found that under certain assumptions the optimal tax on capital may be zero or even negative.
Almost as soon as the 1986 Act was enacted, it began to unravel. Over a period spanning more than a decade we have seen the tax base become more narrow as various tax preferences have been added back to the tax code and others expanded. But, this research has served as a basis for some of the shifts in policy. Tax rates on labor first rose, but more recently, in 2001 and 2003, were reduced. We have seen the tax on key aspects of capital income fall – first the maximum 20 percent rate on capital gains enacted in 1997 and more recently the 15 percent tax rate on both capital gains and dividends enacted in 2003. We have seen increasing consideration of faster write-off of investment – an expansion of section 179 expensing for investment in equipment for small businesses enacted in 2001. The enactment of bonus depreciation, albeit temporary and now sunset. These provisions helped to encourage investment and had as their intellectual underpinnings the notion that investment could be encouraged by reducing the cost of capital. There are continuing murmurs over corporate tax rates that are too high relative to our major trading partners.
The Office of Tax Policy at the Treasury Department and other policy shops in
Incorporating these responses is important because the failure to do so gives an inaccurate depiction of the effects of policy changes. This is a somewhat obvious point, but is worth highlighting. Consider a proposal to increase tax rates. Without incorporating the associated reduction in reported taxable income associated with the higher tax rates, the revenue gain from the higher tax rates would be over stated. Policy makers would be left with the mistaken impression that the policy would raise more revenue than would actually come into the Treasury. The deficit would be higher than they would have anticipated.
Equally important, policy makers would not have the information to properly compare and consider the tradeoffs between different policies. Not all tax cuts or tax increases are created equal. Some impose higher costs or lower gains to the economy than others. Failure to fully include the behavioral responses introduces biases in policy decisions. Policy makers might be less willing, for example, to accept tax rate increases if they raised less revenue than they initially thought.
But just factoring in behavioral responses, as we do now, still tells a story that is incomplete. As this audience knows well, the welfare cost of taxes is related to the square of the tax rate. Incorporating behavioral responses might provide a better estimate of the revenue cost, but it leaves out an important component of the cost of higher tax rates.
Now let me say a few words about the Treasury Department's initiative to expand its capability for dynamic analysis. Since this initiative was announced with the release of the President's FY 2007 Budget in early February, I have grown to better appreciate the different perspectives brought to this issue. There are some who have a long-standing interest in this subject, but who are very concerned that we at Treasury may not do this the right way. Then there are others who are concerned that dynamic analysis may politicize the work we do at Treasury. We at Treasury are well aware of the sensitivity of this issue and we take these concerns very seriously.
The real test for this endeavor – how it is received, whether it is a success or failure, and its longevity – depends crucially on its execution. So, I would like to outline some guiding principles that help inform our thinking and we are applying to this work.
First, we are starting with dynamic analysis – estimating how major changes in tax policy can affect the major macro-economic variables. We are not starting with dynamic scoring, although our work make evolve in that direction. Conventional revenue estimates will continue to be one of our bread and butter products.
Second, we plan to be as forthcoming and transparent with our approach as possible. The models used for this type of work are sensitive to key assumptions, so we need to clearly divulge those assumptions and release enough information regarding the models we are using and the results so that those outside of Treasury can understand and evaluate the work we are doing. This level of transparency might well be viewed as a departure from how we have proceeded with other types of analyses, such as revenue estimates, for example.
Third, we are very interested in long-run effects, as well as the effects along the transition path. One of the major benefits and motivations for dynamic analysis is focusing the attention of policy makers on the long-run benefits or costs of policy changes and the tradeoffs between different policies. This is in contrast to the five and ten year budget windows routinely used for evaluating the revenue effect of proposals today.
Fourth, sensitivity analysis is critically important and will be a central part of our work. As I mentioned, the models used for this type of analysis are sensitive to key assumptions, especially how a tax change is financed, so we need to carefully consider those assumptions and report how the results vary with different sets of assumptions. The Congressional Budget Office (CBO) and the Joint Committee on Taxation (JCT), in some respects, have dealt with sensitivity by using a portfolio of models that each emphasize different dimensions of the analysis. Treasury is very much playing catch up and will, at least initially, focus on a more narrow set of policies, so we are unlikely, at least initially, to use a broad set of models, but we will consider the sensitivity of the results to key assumptions.
Fifth, we very much want to have an open and continuing public dialogue on our work. This is simply an extension of transparency. Participation in conferences like this and those that Jim Poterba and Martin Feldstein have organized through the NBER over the past several years provide a very constructive environment from which we can openly discuss our work outside of the confines of Washington, DC.
This initiative should be viewed very much as building on the continuing evolution of the manner in which we have integrated the behavioral aspects of taxation on economic decision making in all our work at Treasury. And, it should be remembered that we already exercise considerable judgment, and I would say good judgment, in the work we do on conventional revenue estimates and a variety of other analyses.
The new dynamic analysis division is a very natural extension of the work we are already doing. The JCT and the CBO have already been doing this type of work for some time. In the case of CBO, they have been doing this work, in some manner, for many years. In many respects, we are playing catch-up to these other organizations.
This work is also very important to the next subject to which I would like to turn: tax reform. To be clear, tax reform is alive and well at the Treasury Department. As you know, the President's Tax Reform Panel delivered its final recommendations to Secretary Snow on November first of last year and Treasury is evaluating those options. They serve as a very firm foundation, a solid starting point for our work on tax reform.
But tax reform is a very difficult issue and it is an issue that comes along infrequently, perhaps every twenty years or so. Secretary Snow has directed us to proceed slowly and carefully, to thoroughly consider all options and avenues to reform the tax code with an eye for coming up with a plan that is practical and politically viable. We are not working under any specific public timetable. We very much want to get this right and will take the time we need to do so.
It is important to note that the reasons for reform remain. As you know all too well, the tax system is extremely complex imposing a compliance burden on taxpayers of $140 billion per year. Just reducing this compliance burden by one third would produce an annuity to the economy of nearly $50 billion each year, forever.
The
The benefit of reducing the compliance burden is small in relation to the benefit of reducing the economic costs of the tax system. Some estimates suggest that reform could ultimately increase output by 5 to 10 percent. In a 12 trillion economy, that would ultimately translate into an increase in output of $600 billion to $1.2 trillion.
To be clear, these estimates are for very fundamental reforms of our tax code. There are many issues that would need to be addressed in a legitimate and realistic proposal. How progressive would the rate structure need to be to address fairness and what is the extent of transition relief that will need to be provided are just two. But, just as an observation, in some of the options the tax panel put forward, they were able to provide substantial simplification, increase growth – increasing output by as mush as 6 percent -- while achieving the same or slightly more progressive distribution of the tax burden than we have today and providing at least some transition relief.
So long as the tax system continues to fall under its own weight and there are real, tangible benefits from reform, tax reform will remain viable both within
Again, I thank you very much for the opportunity to share some of these thoughts with you. I am happy to take your questions.