Statement of Laurence J. Kotlikoff,
Professor of Economics, Boston University, and
Research Associate, National Bureau of Economic Research
Testimony Before the House Committee on Ways and Means
Hearing on Fundamental Tax Reform
April 11, 2000
Chairman Archer and other distinguished members of the Committee on Ways and Means:
I'm honored by this opportunity to discuss with you the nation's need for tax reform and the role that consumption taxation, particularly a federal retail sales tax, could play in enhancing the economy's economic performance and improving its distribution of resources.
Our nation's economy has been performing remarkably well in recent years, but our economic success is no reason to be complacent about a tax system that is extraordinarily complex and highly distortionary and that plays a critical role in an overall fiscal system that is likely to visit enormous burdens on our children and grandchildren.
The complexity of the tax code doesn't just drive taxpayers crazy. It also costs them a significant amount of time - time that could be spent working or time that could be spent enjoying life. Having just spent three days doing my taxes, I have a refreshed sense of the substantial costs to the man in the street and the nation as a whole of complying with the federal income tax code.
The distortions of our tax system also diminish the nation's well being, but in ways that are less transparent. Today, almost all American households are in combined federal, state, and local marginal income tax brackets of roughly 50 percent. Because governments are collectively confiscating half of every dollar most workers earn, most workers work many fewer hours than they would were their tax payments independent of their labor earnings. And since the government is confiscating half of every dollar of income most savers earn on their non tax-favored retirement accounts, many Americans choose to spend today rather than save for tomorrow.
Tax Reform's Importance for Fiscal Sustainability and Generational Equity
Eliminating complexity and distortions would be cause enough for reforming the federal income tax, but there is a much more pressing reason: notwithstanding recent wishful projections about future government surpluses, our fiscal house is not in order. Indeed, getting it in order would require not cutting federal income taxes, as some in this chamber advocate, but immediately and permanently raising them by over 25 percent. That assessment comes not from academia, but from the Congressional Budget Office and the Federal Reserve Bank of Cleveland. A joint CBO-Cleveland Fed generational accounting study, to be published next month in the American Economic Review, shows that such a tax hike is needed to achieve generational balance - a situation in which our children and grandchildren will face tax rates that are no higher than those we face.
The 25 percent or greater requisite tax hike is derived under the assumption that growth in federal purchases of goods and services keeps pace with growth in the overall economy. This responsible assumption can be contrasted with the irresponsible one underlying the projection of very large surpluses over the next few decades. The irresponsible projection, whose surpluses are routinely cited by advocates of tax cuts and spending hikes, assumes that, as a share of GDP, federal spending will decline by 20 percent by the end of this decade and by 30 percent by roughly 2040.
Who am I to say that the federal government won't shrink to this extent relative to the economy? Just a parent of a two and a nine year old who knows that such shrinkage is highly unlikely and that basing policy on that assumption amounts to gambling with our children's future -- an enterprise worthy of neither this Congress nor this administration.
Our long-term fiscal position is bleak for one straightforward reason. Right now there are about 35 million older Americans. But waiting in the wings are 78 million baby boomers who will start collecting Social Security checks in just eight years and Medicare benefits in just eleven years. Over the next 30 years, the number of elderly will increase by 100 percent, while the number of workers will rise by only 15 percent. This enormous disparity in the growth of the number of elderly and in the number of those who will support them would be much greater still if the recently convened Technical Panel of the Social Security Advisor Board is correct in its assessment that the baby boomers will live substantially longer then the government's actuaries now predict.
What does tax reform have to do with addressing the generational imbalance in U.S. fiscal policy? Essentially everything. To see this, let's start with what a tax reform would tax. Since the federal government is currently taxing wages and capital income, the only meaningful reform would involve taxing consumption on a comprehensive basis (as opposed to levying, as it currently does, a few excise taxes). And each of the major tax reform proposals advanced in recent years does precisely that.
The retail sales tax clearly taxes consumption. But so does the Flat Tax. Just ask Robert Hall, one of the originators of the proposal, who describes his Flat Tax as, effectively, a Value Added Tax. A value added tax taxes output less investment (because firms get to deduct their investment.) Now investment equals saving, so taxing output minus investment is taxing output minus saving, which is taxing consumption, since output minus saving equals consumption.
The Flat Tax differs from a VAT in only two respects. First, it asks workers, rather than firm managers, to mail in the check for the tax payment on that portion of output paid to them as wages. Second, it provides a subsidy to workers with low wages. The first difference is one of form, not substance. The second is more important, but doesn't negate the basic fact that the Flat Tax taxes consumption.
So what does taxing consumption have to do with achieving a generationally equitable fiscal policy? Again, essentially everything. The reason is that the current elderly as well as the baby boomers, who will shortly retire, have one primary economic activity left to accomplish - consumption. And under a consumption tax, they will pay a lot more in future taxes than they would under the current tax system. Although the elderly as a group would share in the burden of a consumption tax, the poor elderly - those living exclusively on Social Security benefits - would not because their benefits are indexed to the consumer price level and are thus guaranteed in real terms.
To recapitulate, given the likely path of government spending and the inevitable aging of our society, our children and our children's children are in for extremely rough sledding. Indeed, the CBO-FED study suggests they will face lifetime net tax rates (1) that are 80 percent higher than those we face if nothing is done. This generational imbalance, rather than the treatment of the rich versus the poor within a generation, is the fundamental issue of economic justice facing us today. Consumption taxation can address that issue by asking the current and near-term elderly to do their fair share in helping to achieve generational balance.
Consumption Taxation and Economic Efficiency
Consumption taxation is needed not just to help our children. It is also needed to simplify the tax code and reduce effective marginal tax rates. The Fair Tax proposal is a case in point. This proposed reform would eliminate both the personal and corporate federal income taxes as well as the payroll tax, and replace them with a federal retail sales tax plus a rebate based on each household's demographic characteristics. Compliance costs would be vastly lower under a retail sales tax. So would, it seems, enforcement costs. The reason is that a broad based sales tax, with no exemptions for housing or any other forms of consumption, would feature much lower effective marginal tax rates than those we now face and, therefore, much smaller incentives to evade taxation.
The lower effective marginal tax rates under the Fair Tax would also mean much smaller economic distortions than currently exists. This reflects the proposition, which is well known to economists, that the welfare costs of distorting economic incentives rises with the square of effective marginal tax rates.
In addition to substantially reducing saving and labor supply distortions, a comprehensive retail sales tax would eliminate a myriad of other distortions such as the lock-in effect on the sale of appreciated assets, the subsidy to health insurance associated with the deductibility of premium payments, the differential tax treatment of investment in different kinds of capital, the tax advantage to debt over equity finance, the marriage penalty, and the subsidy to home ownership.
Is Consumption Taxation Regressive?
Consumption taxation has a bad rap among the general public. It's viewed as regressive when, indeed, it's nothing of the sort. To economists, consumption represents the primary measure of economic well-being. So it makes sense to compare households' taxes with their levels of consumption to determine whether those who are better off pay more than those who are worse off. But consumption is financed not just by current income, but by lifetime resources, which consists of lifetime earnings, lifetime inheritances and gifts received, and initial net worth. So comparing a household's taxes with its command of economic resources requires comparing its taxes with lifetime resources, not current income.
Since lifetime resources are either consumed or bequeathed and since bequests will, themselves, ultimately finance consumption, taxing consumption is like taxing lifetime resources. If consumption is taxed at a fixed rate, as in the case of the Fair Tax and the Flat Tax proposals, the consumption tax will be proportional to lifetime resources; i.e., the tax would be neither progressive nor regressive, but rather proportional.
So if there were no system of taxation to begin with and we introduced a consumption tax, someone with twice the level of lifetime resources as someone else would pay twice the amount of tax. But we aren't starting from scratch. Instead, we are starting from a tax system with some very progressive and some very regressive elements. When measured relative to lifetime resources, the personal income tax is highly progressive, while the payroll tax is highly regressive. And the corporate income tax is essentially proportional to lifetime income since it reduces the net returns to all households no matter the size of their lifetime resources. The fact that the current tax system is not strongly progressive and may even be regressive is the reason that moving from the current system to the Fair Tax, with its progressive rebate, could end up raising the overall degree of tax progressivity.
The lifetime resource perspective leads naturally to comparisons of tax burdens within a cohort, since the lifetime resources of the young and old will be quite different simply because of their ages. Among the elderly, the Fair Tax would be particularly progressive because a federal sales tax would lower the purchasing power of the rich elderly who live off their assets, but not the poor elderly, whose primary means of support - Social Security benefits - would be automatically raised in response to a sales-tax induced increase in the price level. Hence, the Fair Tax features not just a demographic rebate, but also, implicitly, a rise in Social Security benefits. If government transfers to the poor young were also effectively indexed to the price level, the adoption of the Fair Tax would also trigger a rise in those transfer payments as well. (2)
Were the very staid and well established businessmen and women who advocate the Fair Tax to proclaim that their tax reform 1) levies a tax on the holdings of wealth, 2) provides a highly progressive tax rebate, and 3) implies an increase in Social Security benefits and, most likely, transfers to the poor, they would probably be viewed as members of a vast left-wing conspiracy. But this is precisely what they are recommending.
The fact that a consumption tax is, in part, a tax on wealth is well know to public finance economists, but not to the general public. The reason is that when a consumption tax is levied, it lowers the amount of actual consumption that can be purchased with a given amount of wealth since some of the wealth must be spent on the consumption taxes. Stated differently, the imposition of a consumption tax visits an immediate real capital loss on wealth holders because their assets no longer have as large a claim on current or future consumption.
My sense is that the Fair Tax would be more progressive than the current system when assessed on a cohort-by-cohort basis and measured relative to lifetime income. However, knowing the actual degree to which the Fair Tax would enhance intragenerational progressivity requires additional empirical research based on lifetime models of consumption and saving. Such research is now underway, and I would expect that a year from now we'll have a pretty clear picture of the policy's potential impact on the distribution of resources within each generation.
The Long-Term Impact of Consumption Taxation on the Economy
In contrast to the limited empirical analysis of consumption taxation that has been conducted to date, consumption taxation has been studied extensively with large-scale life-cycle simulation models. My own research and that with Alan Auerbach, Cleveland Fed David Altig, Kent Smetters, and Jan Walliser indicates that the Fair Tax would raise the economy's living standard over the long term by roughly 15 percent. (3) This long-run increase in output is generated by a major long-run increase in capital formation and a modest increase in labor supply.
Part of the reason that consumption taxation stimulates saving and labor supply is its improved incentives to work and save. But the primary reason involves the shifting of fiscal burdens away from young savers and onto old spenders. It is a little know, but extremely important fact that the elderly in our country have much higher propensities to consume out of their remaining lifetime resources than do the young and certainly than do the unborn, whose propensities to consume in the present is, of course, zero. The fact that the elderly consume their remaining resources at a higher rate than other generations is precisely what the standard economic theory of saving - the life cycle model of Nobel Laureate Franco Modigliani - predicts. This explanation for this prediction is intuitive; the elderly are closer to the end of their lives than are the young and are, therefore, running short on time to spend their resources. To compensate, they have to spend at a faster rate.
As described in Gokhale, Kotlikoff, and Sablehaus (1996), essentially all of the decline in the rate of U.S. saving in the postwar period can be traced to the government's five-decade long policy of taking ever larger sums from the young and giving them to the old. (4) This intergenerational redistribution, carried out primarily through Social Security, Medicare, and Medicaid, has led to a dramatic rise in the absolute and relative consumption of the elderly. Since 1960, for example, the elderly's share of economy-wide consumption has increased more than four times fast than has their share of the population. Typical 70-year olds are now consuming roughly twice the amounts consumed by typical 30-year olds. In 1960, by contrast, 70-year olds consumed less than three quarters of the amounts consumed by 30 year olds.
In shifting to a consumption tax, the U.S. would shift more of the tax burden onto the current middle class and rich elderly and partly reverse the postwar process of taking from the young and giving to the old. In addition to depressing national consumption and raising national saving, the switch to consumption taxation would, as indicated above, ameliorate our grievous imbalance in generational policy.
The simulation studies also show substantial long-run welfare gains for all lifetime income classes from switching to consumption taxation. Indeed, under the Fair Tax, the initial upper income elderly are the only ones to suffer welfare losses during the transition.
Tax Rates
Simulation analysis and a variety of empirical calculations suggest that the retail sales tax rate needed for revenue neutrality under the Fair Tax, assuming no decline in the real value of government purchases, would be roughly 30 percent when measured on a tax-inclusive basis. This tax rate could be expected to decline by 3 or so percentage points over time as the economy expands. Moreover, if the Fair Tax were structured to include the consumption of existing housing services in its tax base, the initial Fair Tax rate would probably be about 3 percentage points lower. This could be accomplished by assessing the tax on the imputed rent on housing, where the calculation of imputed rent is based on a fair market valuation of housing real estate. This valuation could be done by local municipalities in the course of appraising houses for local property taxes.
A tax-inclusive consumption tax rate of 30 percent translates into a tax-exclusive consumption tax rate of 43 percent. While the 43 percent rate sounds very high, proper comparison of the Fair Tax tax rate with the current payroll and income tax rates requires evaluating the consumption tax rate on a tax-inclusive basis. Even a 30 percent tax rate may sound like a high rate. But one needs to bear in mind that middle and upper income households in America are typically in combined income tax and payroll tax marginal tax brackets of 40 percent or more and that low income Americans are typically in even higher tax brackets once one considers the phase out of the earned income tax credit. Hence, given the state of U.S. marginal taxation, 30 percent is a low number.
Transition Issues
Shifting to a consumption tax requires thinking carefully about transition issues. In the case of the Fair Tax, one would want to make sure that the vast sums that have been accumulated tax free in retirement accounts not avoid taxation. How this could be accomplished fairly and quickly is not altogether clear. But what is clear is that the large amount of revenue to be raised here could help limit the size of the Fair Tax Rate. Note that this problem doesn't arise under the Flat Tax because the Flat Tax maintains an explicit tax on labor income and retirement account withdraws are included in the labor income tax base.
While the Flat Tax deals with this transition issue much more easily than does the Fair Tax, the Fair Tax avoids the potential for special transition rules that would favor existing business capital under a Flat Tax and, thereby, dissipate the tax's implicit taxation of existing wealth.
Transparency and Perceived Fairness
The Fair Tax would be easily understood by the general public, and it would be clear to all that everyone - rich and poor alike - pays the tax. In contrast, under the Flat Tax, wealthy individuals who have no labor income will appear to be paying no tax when, in fact, they will implicitly do so through the revaluation downward of the market value of their assets, assuming no special transition rules in behalf of those assets.
Conclusion
The Fair Tax has a lot to recommend it. It would most likely help the poor more than the rich. It would substantially improve the economy's economic performance. It would save Americans enormous amounts of time complying with the bewildering provisions our current tax code. And it would redress the grave intergenerational imbalance America still faces with respect to its fiscal policy.
1. The lifetime net tax rate is defined as the present value (to birth) of taxes paid over the lifetime net of the present value (to birth) of transfer payments received over the lifetime divided by the present value (to birth) of lifetime labor income.
2. This sentence and the one preceding it assume the price level will rise with the adoption of the Fair Tax. If the Federal Reserve used its monetary policy to maintain the consumer price level, the adoption of the Fair Tax would entail a decline in the level of producer prices and, thus, the nominal wages and capital income received by productive factors. Under this scenario, government transfers, if they weren't reduced in nominal terms, would end up maintaining their purchasing power, while factor payments would not. I.e., the same real redistribution toward the poor would arise.
3. See Kotlikoff, Laurence J., "Replacing the U.S. Federal Tax System with a Retail Sales Tax - Macroeconomic and Distributional Effects," mimeo, December 1996 and Altig, David, Alan J. Auerbach, Laurence J. Kotlikoff, Kent Smetters, and Jan Walliser, "Simulating Fundament Tax Reform," forthcoming, The American Economic Review, 2001.
4. Gokhale, Jagadeesh, Laurence J. Kotlikoff, and John Sablehaus, "Understanding the Postwar Decline in U.S. Saving: A Cohort Analysis," The Brookings Papers on Economic Activity, no. 1, 1996, 315-90.