Skip Navigation

The Economic Effects of Comprehensive Tax Reform July 1997

Could a switch to a broad-based consumption tax stimulate national saving, domestic investment, and labor supply? If so, increases in economic output would be likely as well. The effects on market interest rates may be difficult to predict, but such effects are of secondary importance.

Any increases in capital accumulation and hours of work depend on certain features of the tax proposals. As described earlier, most proposals to replace federal income taxes make three fundamental changes: shift from an income tax base to a consumption tax base; move toward lower and more uniform tax rates; and broaden the tax base by eliminating both deductions and exclusions.

The switch to a consumption tax base is likely to encourage capital accumulation by removing the taxation of capital income and hence the incentive to consume sooner rather than later. More uniform tax rates will also eliminate some tax differences that encourage less productive uses of capital in the current system. A broader tax base could increase the labor supply if the broadening was sufficient to allow a reduction in overall marginal tax rates. Without broadening the base, a revenue-neutral switch from an income base to a consumption base would lower the tax on saving and investment but would increase the tax rate on labor.

The precise impact of switching to consumption-based taxes cannot, however, be predicted with accuracy. Simulation models using particular assumptions about changes in behavior yield some insight into what could happen if the assumptions of the particular models prove correct. Unfortunately, those models are not sufficiently advanced to provide a reliable set of bottom-line estimates.

Although comprehensive tax reform could lead to substantial revenue shortfalls or bring in more revenue than the current system, the analysis in this study assumes that all of the proposals are revenue neutral. That assumption implies that reform will not change the initial level of after-tax income of the private sector as a whole.
 

The Effect of Tax Reform on Saving

Saving is how a nation provides for the future. Through saving, it can build up a stock of assets and support higher levels of consumption in the future. The U.S. national saving rate--that is, the percentage of national income that is saved--declined substantially in the 1980s, and it has dropped even more in the early 1990s, raising deep concerns among policymakers and analysts. Although many factors, including those beyond the control of government policies, contribute to the low saving rate, replacing the current tax system with a more efficient one could boost saving.

Shifting toward any of the consumption-based taxes discussed in Chapter 2 would probably increase national saving and ultimately raise the living standards of future generations. But the magnitude of that response is highly uncertain. The proposals are complex, and economists still have a rather rudimentary understanding of the factors that influence saving. Nonetheless, the designers of a consumption-based tax would inevitably face a number of issues raised in the discussion below.

Why Would Saving Increase?

The proposals for a consumption-based tax could increase saving for two reasons. First, by eliminating taxes on new saving, they would reduce the price of future consumption compared with current consumption. In itself, that outcome would also raise people's lifetime resources, allowing them to consume more today and tomorrow. But if the reform is revenue neutral, the tax rate on consumption must be correspondingly higher. As a result, total lifetime resources will initially remain the same as they are under the current system. Thus, on average, the change in relative prices gives people an incentive to reduce current consumption and save more for the future.

Second, the proposals might redistribute lifetime resources from low savers to high savers. Among the various ways that tax reform might redistribute income, the effects of redistribution are particularly relevant to two groups: people of different ages, and people with different levels of income.

Consumption-based taxes impose a levy on existing business assets, which are largely held by older people (see Appendix A). At the same time, because the proposals are revenue neutral, the lighter tax burdens on younger people offset the added burdens on older people. Redistributing tax burdens from the young (including future generations) to the old could increase total saving because older people tend to consume a larger fraction of their lifetime resources than do younger people.

Two factors, however, lighten--and perhaps completely offset--that redistributive effect among generations. First, some of the proposals provide relief for holders of existing capital during the transition period. That relief eliminates much of the redistribution of resources between low and high savers. Second, holders of existing assets may be in a better position to take advantage of the investment incentives under a consumption-based tax. For example, ongoing firms have significant advantages over newcomers in most industries. If so, those advantage would tend to put upward pressure on the market value of firms. In principle, that effect could offset the negative impacts of the additional tax on the value of those firms' existing stock. Some economists suggest that stock prices might therefore rise.⁽¹⁾ But other analysts argue that for adjustment costs to be high enough to more than offset the levy on capital, share prices of firms would have to be implausibly high in relation to the per-share value of their capital. Thus, the argument continues, without transition relief the value of existing assets would fall.⁽²⁾

Some proposals might also end up redistributing income within generations if they flattened the rate structure. However, such a flattening would tend to increase the tax burden on lower-income families and reduce the burden on higher-income families compared with the present tax system. That type of redistribution would also encourage saving because higher-income families tend to be bigger savers than lower-income families, even on a lifetime basis.

How Much Would Saving Increase?

How much private saving would change is difficult to judge. Even a comprehensive version of a consumption-based tax, with no relief for the owners of existing capital, could produce a broad range of estimated responses in saving. Any proposal that provided relief to owners of existing capital would lead to a significantly smaller effect on saving.

Economists Alan Auerbach and Laurence Kotlikoff estimate that if a proposal completely eliminated the tax on existing capital, 70 percent of the increase in saving from tax reform could be lost.⁽³⁾ In the end, tax reformers confront a trade-off: if they want to ease the additional tax burden on existing assets and on the older Americans who own the largest portion of those assets, they must be willing to accept substantial reductions in the long-run economic benefits to younger people and future generations.

Direct Empirical Evidence. The most direct way to determine how saving might react to a switch to a comprehensive consumption-based tax is to examine how people have responded to changes in incentives to save in the past. Unfortunately, existing empirical studies provide a bewildering range of estimates. Some studies find that saving responded markedly to changes in after-tax rates of return; still others find no response.

One problem is the difficulty in measuring the key variable--the net real rate of return on saving. In principle, the rate of return depends on factors that cannot be observed, such as people's expectations of future inflation and effective tax rates. Early studies ignored those measurement problems altogether and thus do not provide reliable estimates.⁽⁴⁾ More recent studies make assumptions about how people anticipate future inflation. In one of the first papers to account for taxes and inflationary expectations, Michael Boskin reported a rather strong positive response by saving to increases in after-tax rates of return.⁽⁵⁾ In that paper, other things being equal, a 10 percent rise in the real after-tax rate of return would cause people to raise their gross saving by 4 percent. Moreover, if the real value of existing assets fell at the same time, saving would rise even more. But other empirical studies that also attempted to address the measurement problems found that interest rates seemed to have little effect on consumption and saving.⁽⁶⁾

Another problem with studies that use the direct approach is that they do not account for the demographic factors that can also influence saving behavior. For example, how workers respond to changes in the rates of return depends on how long they expect to live and when they expect to retire. Thus, the response of saving is not a fixed parameter, as assumed by the direct studies. Instead, it depends on the demographic structure of the population, which changes over time.⁽⁷⁾ The direct studies also ignore the reverberating effects that tax reform would have throughout the economy. To address both of those issues, economists use computer simulation models.

Simulation Models. Simulation models suggest that switching to a comprehensive consumption-based tax is likely to increase saving. The models reviewed for this study suggest that the saving rate could increase by as little as 3 percent or as much as 25 percent in the long run. The empirical evidence suggests, however, that consumer saving responds less to changes in interest rates than those models generally assume. Moreover, few of the results of the models account for the saving incentives in the current tax system or the effects of uncertainty on saving behavior. As a result, the rise in saving is unlikely to be in the upper end of that range. The response would be lower still if significant relief for owners of existing capital lessened the tax on existing capital. (See Appendix B for a more complete discussion of the simulation models.)
 

The Effect of Tax Reform on Capital Flows and Domestic Investment

The United States is part of the world economy in which savers are relatively free to invest in many different countries. Because saving can flow across borders, the level of domestic saving does not constrain domestic investment in the United States. If U.S. national saving fell short of domestic investment, capital from abroad would fill the gap. Conversely, if U.S. saving exceeded domestic investment, the difference would be invested abroad.

The free movement of capital across borders has significant implications for the effects of tax reform. If the economy was closed to the outside world, saving would always have to equal investment. In that case, no fundamental difference would exist between a tax policy that stimulated investment and one that stimulated saving. In the end, the real effects of the two policies on the economy would be the same. But in an open economy, incentives for investment are not equivalent to incentives for saving. Incentives for saving encourage domestic savers whether they invest at home or abroad. In contrast, incentives for investment encourage both foreign and domestic savers to invest in domestic firms.

Compared with current policy, all of the proposals described in the previous chapter encourage both saving and investment. The saving incentives come from the increase in the after-tax return from postponing consumption. The investment incentives stem from a reduction in the cost of capital to businesses.

Tax Reform and the Cost of Capital

Businesses will undertake new investments if those investments yield a sufficient return after taking account of all costs, including operating expenses, depreciation, and taxes. The cost of capital is the pretax rate of return that is necessary to yield the prevailing after-tax return, once all costs are paid. Current proposals to replace the income tax make several changes that, when taken together, would probably reduce the overall cost of capital.

First, the full expensing that takes place under cash-flow consumption taxes (such as the value-added tax, Armey-Shelby flat tax, and Unlimited Saving Account tax) allows businesses to deduct any purchases of capital immediately. That deduction directly lowers the cost of capital and stimulates demand for investment. Second, integrating personal- and corporate-level taxes would remove the double taxation of corporate equity and hence tend to reduce the cost of corporate capital. Finally, broadening the tax base implies lower overall marginal tax rates. Consequently, the overall cost of capital could fall.⁽⁸⁾

However, the proposals would probably have bigger effects on the costs of some forms of capital than of others simply because the current tax system does not treat all forms of capital equally. (See Chapter 4 for a closer examination of the effects on the costs and allocation of different types of capital).

Risk Taking

Economic activities do not all receive the same after-tax return, even after adjusting for differences in risk. Some do better than others; some do worse. Taxing unexpectedly high returns from investment can have seemingly paradoxical effects on investment decisions.

Under current law, taxes on capital income generally discourage investment because they reduce the expected net return from investment. They also increase risk taking (the riskiness of a given dollar invested) because they reduce the variation of net returns. If the investment yields an unexpectedly high return, the government receives higher revenues. Yet if yields fall below expectations, revenues do also. In effect, the government bears some of the risk of the investment by sharing unexpectedly good or bad outcomes, the marginal tax rate being the government's share.

The incentive for taking risks would change under a consumption-based tax. Because the expected net returns from capital are untaxed under a consumption tax, the level of investment should be higher than under current law. However, the excess of actual returns above expected returns would be subject to tax. Moreover, any shortfall of actual returns below expected returns would reduce tax liability. The government still bears some risk. Thus, although overall investment should be encouraged after switching to a consumption tax, the riskiness of investments could increase or decrease depending on whether the tax rate on unexpected capital income was higher or lower than under the current income tax system. Decreases in marginal tax rates, accomplished by broadening the base and flattening the overall rate structure, would make it more likely that a replacement for a consumption tax would discourage risk taking while encouraging overall investment.

Investment and Economic Rents

Some investments deliver economic rents (or supernormal returns) because they can take advantage of the power of their monopoly in the marketplace. Those monopolies sometimes arise when product markets do not work perfectly. Yet they also come from innovations that create new products and markets. In general, those economic rents are taxed under current law.

All of the proposed consumption-based taxes would continue to tax economic rents, although to varying degrees. Expensing would eliminate any tax on the normal returns from an investment. Under a consumption tax, the cost of an investment is fully deductible (expensed) at the time of purchase. The future income from an investment would be subsequently taxed. But for a normal investment (one without economic rents), the present value of that income stream is the same as the cost of the investment. Making the initial investment deductible is thus equivalent to not taxing the income from the investment.

The supernormal returns earned by firms with market power would still be subject to tax, however. Any tax collected on supernormal returns would raise revenue but would not affect investment decisions at the margin (that is, for each additional dollar of investment).

International Capital Flows and Trade

The balance between saving and investment in the United States will determine the amount borrowed from abroad. If tax reform stimulates more investment than saving, capital flows into the United States will initially rise. But if saving increases more than investment, capital inflows will fall.

The MSG multicountry (or MSG2) model is a macroeconomic forecasting model that offers some insights into the possible effects of switching from an income tax to a value-added tax in an open economy. The model reflects the key interactions among the major industrialized countries. After the change in tax policy, the MSG2 model predicts that rising demand for investment would outweigh the increasing supply of saving for almost 20 years. As a result, capital inflows would rise, which initially would make the dollar appreciate and net exports fall. The way in which reform treats imports and exports would not, however, affect the trade balance appreciably. Despite the apparent difference between destination- and origin-based value-added taxes, they have about the same effect on the trade balance in the long run (see Box 2).
 

If interest rates rise, the openness of the United States economy will ease the transition to an economy with greater capital intensity. Access to international capital markets means that the nation can build up its capital stock without cutting consumption as much. In essence, U.S. residents will be borrowing from abroad to smooth their consumption over time.⁽⁹⁾ In the long run, however, the debts to foreigners must be serviced, and the income from imported capital will largely accrue to foreigners.

Although analyses that treat the U.S. economy as a closed one may understate the increase in capital formation that would occur following a switch to a consumption tax, the simplest specifications of open-economy models tend to err in the opposite direction in a more extreme way. Models that treat the United States as a small open economy (in which changes in the U.S. capital stock have no effect on world interest rates) often suggest implausibly large increases in capital formation. Acknowledging that the United States is really a large open economy moderates the expected increases in the capital stock. In fact, the evidence on international capital flows does not generally suggest that they are large given the magnitude of the capital stock.⁽¹⁰⁾

Relaxing certain unrealistic assumptions under an open-economy framework tends to bring results closer to those of a closed economy.⁽¹¹⁾ Moreover, openness introduces the possibility that the U.S. capital stock might contract because debt capital might flow out as equity capital flows in.
 

The Effect of Tax Reform on Labor

The effect of reform on labor depends to a significant extent on the details of the tax proposal. Some proposals could increase the supply of labor; others could reduce it. The more that policymakers broaden the tax base by eliminating current tax preferences, the lower will be marginal tax rates on labor income, and hence the greater the likelihood that the supply of labor will increase under tax reform.

Labor Supply

Decisions regarding how much labor to supply ultimately depend on calculating the relative benefits of work over leisure. In simple terms, people work in order to finance consumption both today and in the future. Thus, the return from work is how much current and future consumption the person can obtain by giving up an hour of leisure today. In that view, the incentive to work depends not only on the after-tax wage, but also on the relative price of current versus future consumption. Consumption taxes will affect both after-tax wages and the relative price of future consumption. In addition, tax reform may redistribute income in ways that affect the total supply of labor. Sometimes those effects will work in the same direction; sometimes in opposite directions.

Marginal Tax Rates on Labor. Marginal tax rates on labor directly reduce the after-tax wage and hence reduce the relative cost of leisure. Thus, people have an incentive to work less.

The effect of tax reform on the marginal tax rate depends critically on the details of the proposal. If reform does not broaden the tax base by eliminating various preferences, the marginal tax rate on labor must increase in order to raise the same amount of revenue as the current system. The reason is that the base for a consumption tax is smaller than that for an income tax, the difference being saving. In such a situation, tax reform might lead to a decrease in the supply of labor. In contrast, if reform also eliminated tax preferences, it might be able to broaden the tax base by enough to permit a reduction in the tax rate on labor. In that situation, reform could increase the incentives to work.

Further, what matters for the labor supply is the marginal tax rate on total compensation, not just the marginal tax rate on wages received. Total compensation includes fringe benefits, such as health insurance and pension benefits, as well as payroll taxes paid by the employer. Fringe benefits are excluded from tax under current law, and thus the effective marginal tax on compensation is much less than the statutory tax rate on wages. That aspect of current law should be taken into account when estimating how reform changes incentives to work. If a proposal expands the definition of taxable compensation, labor supply is more likely to decline.

The various proposals for a consumption-based tax also differ in the extent to which they provide transition relief to holders of existing capital, which will also ultimately influence marginal incentives to work. As discussed earlier, a pure consumption tax imposes a tax on people who hold existing capital. Because that tax burden is effectively a lump-sum tax on wealth, it does not distort economic choices. In contrast, relief for transition reduces the size of that lump-sum tax. But nothing is free, and higher taxes on labor must inevitably finance transition relief.

Empirical studies indicate that workers are modestly responsive to revenue-neutral changes in after-tax wages.⁽¹²⁾ For the workforce as a whole, a 10 percent rise in after-tax wage rates could increase the labor supply between 2 percent and 4 percent.⁽¹³⁾ About half of that increase results from people joining the labor force; the remainder reflects an increase in average hours worked.

The effect of tax changes on the labor supply varies significantly among different groups of workers. For example, a decrease in tax rates would have more impact on the labor supply of married women than of men, single women, or female heads of households. In response to a revenue-neutral policy that increased after-tax wage rates by 10 percent, evidence suggests that men and single women would increase their hours of work by 1 percent to 2 percent. In contrast, second earners in two-worker families, who are mostly married women, could increase their labor supply by 6 percent to 9 percent. Thus, if tax reform prompts any large change in labor supply, it will probably occur among married women.

Changes in the Relative Price of Consumption. The second effect on labor arises because a switch to a consumption-based tax would reduce the price of future consumption in terms of current consumption. With a higher after-tax rate of return from saving, people get more future consumption for each current dollar they save (do not consume). That change effectively increases the after-tax return from work and stimulates labor supply.

The idea that changes in the price of future consumption could affect incentives to work today may seem puzzling at first. But most people work because they want to be able to consume goods and services. Moreover, although most income is consumed in the same year that it is earned, some of it is saved to finance future consumption. Thus, when the price of future consumption falls, workers can effectively purchase a larger basket of future goods and services. In essence, the decline in the price of future consumption encourages people to work more now because they can consume more later--that is, they substitute future consumption for current leisure.

Models that account for substitution over time can capture that "intertemporal" effect on labor supply. If that effect is significant, a consumption-based tax is likely to stimulate labor supply to a greater degree than an income-based tax would. For example, one analysis suggests that labor supply is encouraged slightly under a switch to a flat-rate consumption tax but discouraged under a switch to a flat-rate income tax.⁽¹⁴⁾

Redistributive Effects on Labor Supply. The proposals may initially redistribute resources (lifetime income) among different types of people. An increase in such income, with given prices and wage rates, will lead people to decrease labor supply (increase leisure). With more lifetime resources, people can consume more without having to work as much. For instance, assume resources are shifted from people whose labor supply is less sensitive to changes in lifetime income to people whose supply is more sensitive. Other things being equal, the supply of labor will initially fall, even though total resources are unaffected when the reform is revenue neutral.

Productivity, Wages, and Unemployment

Although tax reform could cause labor supply to increase, most estimates suggest that capital will expand even more in the long run.⁽¹⁵⁾ As a result, the overall capital intensity of the economy will rise, which will in turn push up productivity and wage rates. On average, labor should eventually benefit from increased capital accumulation.

In fact, simulation models provide a range of results for the effects of tax reform on real wages. Using a life-cycle model, Alan Auerbach finds that the before-tax real wage rate increases by 4 percent to 6 percent in the long run, depending on the proposal. Using a different life-cycle model, Don Fullerton and Diane Lim Rogers find that the real wage rate rises by 2 percent to 7 percent, depending on how responsive households are in their timing of consumption. Finally, using a precautionary-saving model that shows smaller increases in the ratio of capital to labor, Eric Engen concludes that the pretax real wage rate increases by just 1 percent in the long run.⁽¹⁶⁾

In the transition, however, some workers are likely to experience unemployment as the economy adjusts. In addition, the reallocation of resources to capital- intensive production may well lead to the permanent displacement of certain types of workers. For example, such displacement can occur if growing industries use laborers who have different skills from those used in shrinking industries. In other words, although labor overall is apt to benefit in the long run from increased capital accumulation, the overall gain may consist of a larger number of gainers and a smaller number of losers.

Moreover, because labor markets tend to adjust more quickly to incentives than capital markets do, labor supply may increase more than capital supply in the short run. That development would lower the ratio of capital to labor and thus reduce productivity and real wages during the early years of the transition.

Human Capital

Improving the skills and knowledge of workers, or "human capital," is one of the most important elements influencing the long-run performance of the economy. More questions are being raised about how changes in tax policy could affect the accumulation of human capital.

The accumulation of human capital involves both direct and indirect costs to individuals and firms. The direct costs are cash expenses, such as tuition, fees, books, and other out-of-pocket costs. The indirect costs are the income that people lose (or output that firms lose) when people spend time at school or in training rather than working.

The indirect costs are, in effect, deducted from taxable income under the current income tax system because the lost income reduces the taxable income of the individual (or firm) dollar for dollar. In addition, firms can also write off their direct training costs. In contrast, individuals receive no preferential tax treatment under the current system for their direct cash expenses, although the government tends to subsidize some of those costs.

All proposals for consumption-based taxes would preserve the current treatment of indirect costs and the direct costs of firms. They differ, however, in their ability to relieve the current tax on the direct costs of human capital for individuals. The USA tax would give individuals a deduction for a portion of those direct costs. Under a national sales tax, schools and other educational institutions could be exempted. However, in the case of a value-added tax, how the taxes on those direct costs for human capital could be eliminated completely is far from clear. Although schools and other educational institutions could be exempted from paying the tax on their own value added, they would still end up paying some of the tax indirectly in their purchase of goods and services from nonexempt firms. Moreover, at the base of any such effort lies a significant administrative problem in defining what constitutes a legitimate investment in human capital.

For those reasons, and because human capital already is largely taxed on a consumption basis, tax reform is unlikely to have much effect on the supply of human capital. But reform will significantly reduce the tax rate on tangible capital investments--and thus change the relative incentive to invest in physical capital. Consequently, physical capital would become more attractive than human capital.

Alternatively, accumulating human capital may be encouraged if higher levels of skill are necessary for employees to work with new, perhaps higher-tech, physical capital. In such a case, human capital might be considered more of a complement to physical capital than a substitute for it.
 

The Effect of Tax Reform on Economic Output

What would be the effects on output of replacing the current income tax with a comprehensive consumption-based tax? The outcome would hinge on what happens to the supply of inputs and how efficiently those inputs are used. Evidence seems to support the prediction of positive effects on saving and capital accumulation. Even though the effect on labor supply depends on the details of the policy, most economic models also predict a positive effect on the level of national output from an increased labor supply.

Simulation models can provide some insights into the effects of fundamental tax reform on total output and consumption. This section will consider results from two classes of models. The first class is represented by general-equilibrium models, which assume that the economy is always at full employment. The second class of models has so-called "structural features" in which workers can become unemployed if total demand is insufficient.

General-Equilibrium Models

Recent studies using general-equilibrium models suggest long-run increases in output of 1 percent to 10 percent. One study finds that output could increase between 2 percent and 9 percent, depending on the particular details of the policy.⁽¹⁷⁾ Another study predicts increases in output of between 1 percent and 6 percent, depending on how sensitive consumers would be to changes in the rate of return from capital.⁽¹⁸⁾ Still another study predicts an increase in output of about 3 percent.⁽¹⁹⁾ An earlier study found that reform could raise output by as much as 19 percent. However, that study was based on tax law in 1980, which had a narrower base and higher rates than current law. In addition, that study assumed a saving response higher than what recent empirical evidence supports.⁽²⁰⁾

A large part of saving, including all saving through pensions, is already exempt from taxation under current law. Most earlier studies omit that feature of the current tax system and thereby miss an important feature of the starting point for tax reform. However, a recent model explicitly takes into account the hybrid nature of the current tax system and finds an increase in output of 7.5 percent in the long run.⁽²¹⁾

Structural Macroeconomic Models

Although tax reform raises the long-run level of gross domestic product (GDP), the economy could experience some short-term increases in unemployment. Analysts at Data Resources, Inc., used their model to investigate the impacts of flat-tax legislation assumed to be in place in 1996.⁽²²⁾ Their starting point was a 15 percent decline in housing prices that would occur when payments for mortgage interest and property taxes were no longer tax-deductible. Growth would slow sharply in the first few years because the reductions in consumer and housing spending would occur promptly, whereas the capital spending boom would take time to develop. In the period from 2000 through 2005, however, real GDP would be almost 1 percent higher on average than in the baseline.

Another model with structural features, the MSG2 model, also predicts that tax reform would lead to short-run losses in economic output. With the changes beginning in 1998, the MSG2 model predicts lower output than in the baseline for 1999 through 2002, but higher output thereafter. In the long run, the model predicts, output would be about 10 percent above the baseline level. But consumption would fall more than 10 percent below its baseline value in the early years and would not return to the baseline until 2022.

In all of those models, although tax reform ultimately increases the level of GDP, it does not permanently raise the growth rate of the economy. To be sure, the economy has to grow somewhat faster during the transition period in order to reach that higher level. However, once the economy reaches that higher level, it grows at the same rate that it would have if policy had not changed. Although some recent research has suggested that some policy changes might be able to raise long-term growth rates, support for those theories is weak.⁽²³⁾

Whatever the increase in output, long-run consumption would increase by less. Some of the gains to gross domestic product would have to be used to cover the additional depreciation of the higher level of capital, and people would consume a smaller fraction of net income than they do now.
 

The Effect of Tax Reform on Interest Rates

Studies of tax reform have differed wildly in their estimates of the effect of tax reform on interest rates. Some researchers predict that interest rates would fall; others see them going up. Much of the confusion about the effects of tax reform on interest rates is a product of an abundance of studies that focus on different measures of the interest rate, use different models of the saving response, and make different assumptions in their calculations.

Despite that confusion, researchers agree that the switch to a consumption-based tax would raise the after-tax return from saving and would lower the marginal product of capital (the amount of output produced by the last unit of capital invested). But the effects of reform on other rates of return--such as the market return from equity or the interest rate on corporate debt--are ambiguous (see Box 3).
 

In any case, how tax reform would affect rates of return is much less important than how it would affect the ratio of capital to labor. On that point, most researchers agree: moving to a consumption-based tax would increase capital intensity, which would boost the real wage rate and increase the standard of living in the long run. The effects on market interest rates are of secondary importance.

Results of Model Simulations

Because of the uncertainty about the effect of tax reform on rates of return, using model simulations to analyze the role of various factors can be helpful. The simulations show that if saving is assumed to be more elastic, significant reductions in rates of return in the long run become more likely. But the results crucially depend on assumptions about the relative tax rates on debt and equity and the size of the risk premium. Moreover, comparing the results of different analyses is difficult because they focus on different rates of return.

Using a precautionary-saving model with a small saving elasticity, one study found that the before-tax return from capital (essentially a weighted average of the rates on debt and equity) could decline by as little as zero or as much as 11 percent in the long run, depending on the details of the proposal.⁽²⁴⁾ Even bigger effects--18 percent to 24 percent in the long run--came from one study using a life-cycle model that has a large saving elasticity.⁽²⁵⁾ In addition, Robert Hall and Alvin Rabushka, the architects of the consumption-based flat tax, estimate that the market rate of interest would fall by approximately 20 percent if the United States adopted a consumption tax similar to the Armey-Shelby flat tax.⁽²⁶⁾ (Those results are stated in percentage terms, not percentage points. A 20 percent change in an interest rate that was initially 10 percent would move rates by 200 basis points.) In the short run, however, Alan Auerbach found that tax reform would stimulate labor supply more than capital supply. That outcome would cause the marginal product of capital--and the pretax rate of return from capital--to increase temporarily. Other models do not predict such a response.⁽²⁷⁾

Unfortunately, only a few studies distinguish between the rates on equity and debt. One such study, by Martin Feldstein, suggests that interest rates on debt would nearly double in the near term and remain higher than prereform interest rates over longer horizons.⁽²⁸⁾ His predictions are based on the assumption that reform would not alter the premium between equity and debt and thus would create significant upward pressure on interest rates, as discussed above. He also assumed that the marginal product of capital would not decline by a large amount and that the prereform tax rate on interest income would be much lower than that on equity income.

In the previously cited paper, Auerbach comes to opposite conclusions about the interest rate, noting that current tax rates on equity (including housing) are much lower than those on interest income--and thus interest rates should fall by more than equity rates.⁽²⁹⁾ But he also assumes that debt and equity are perfect substitutes and therefore ignores the risk premium that plays such an important role in Feldstein's analysis. In summary, those two studies illustrate the tremendous uncertainty that plagues predictions about the effects of reform on rates of return. The results critically depend on the underlying assumptions.

International Considerations

The flows of capital across national borders will also influence how tax reform affects interest rates. In some open-economy models, interest rates (adjusted for exchange-rate expectations) are assumed to be fixed by world capital markets. Thus, tax reform would not affect interest rates. Those models, however, apply only to small economies in which capital is highly mobile.

In reality, the U.S. economy is quite large compared with the rest of the world. In 1992, gross domestic saving by the United States represented 17 percent of the world total; gross domestic investment totaled 18 percent. Such magnitudes suggest that the much larger world capital market is likely to absorb much, but not all, of the pressure on interest rates in the United States.

The MSG2 model provides channels for the United States to influence the rest of the world. In that model, tax reform would cause an increase in demand for investment that would outweigh the increasing supply of saving for almost 20 years. Hence, the real rate of interest would rise above the baseline level during that period. In the long run, however, the real rate would fall below the baseline.
 

Conclusion

The effects of fundamental tax reform on the economy are highly uncertain. Results depend on the type of models used and the assumptions built into those models. However, some broad conclusions stand out about capital accumulation, labor supply, and economic output.

Capital Accumulation

Results from economic simulation models indicate that consumption-based tax proposals are likely to increase national saving. The empirical evidence on the intertemporal response, the hybrid nature of the current tax system, and uncertainty about saving behavior suggest that the effects of tax reform on capital accumulation are less likely to be at the higher end of the range of estimates presented in this study. Transition relief would reduce the effects still further. The reductions in overall costs of capital under a consumption tax should lead to an increase in investment and capital intensity, but might reduce risk taking. The switch to a consumption-based tax is likely to affect international capital flows as well, although the short-term patterns will differ from those in the longer term.

Labor Supply

A switch to a comprehensive consumption-based tax could cause the supply of labor to increase or decrease, depending on the significance of reductions in marginal tax rates and a lower price of future consumption in relation to present consumption. The effects of tax reform on human-capital accumulation depend largely on whether human capital is more of a substitute for or a complement to new physical capital.

Output

The probable increases in the capital stock, coupled with smaller changes in labor supply (in either direction), indicate that the level of national output would be likely to increase in the long run. As with the saving response, the magnitude of the change in output is uncertain. Simulation models suggest increases ranging from 1 percent to 10 percent, and other factors suggest that the upper end of that range is less likely. Nonetheless, even with significant changes in the level of output, tax reform is unlikely to raise the growth rate of the economy permanently.

1. David M. Cutler, "Tax Reform and the Stock Market: An Asset Price Approach," American Economic Review, vol. 78, no. 5 (December 1988), pp. 1107-1117; and Andrew B. Lyon, "The Effect of the Investment Tax Credit on the Value of the Firm," Journal of Public Economics, vol. 38 (March 1989), pp. 227-247.

2. Alan Auerbach, "Tax Reform, Capital Allocation, Efficiency, and Growth," in Henry J. Aaron and William G. Gale, eds., Economic Effects of Fundamental Tax Reform (Washington, D.C.: Brookings Institution, 1996), pp. 29-73.

3. Eliminating the tax on existing assets would effectively convert a consumption tax to a wage tax. See Alan J. Auerbach and Laurence J. Kotlikoff, Dynamic Fiscal Policy (Cambridge, England: Cambridge University Press, 1987).

4. Paul David and John L. Scadding, "Private Saving: Ultra-Rationality, Aggregation, and Denison's Law," Journal of Political Economy, vol. 82, no. 2, part 1 (March-April 1974), pp. 225-249.

5. Michael Boskin, "Taxation, Saving, and the Rate of Interest," Journal of Political Economy, vol. 86 (April 1978), pp. S3-S27.

6. Alan S. Blinder and Angus Deaton, "The Time-Series Consumption Function Revisited," Brookings Papers on Economic Activity, no. 2 (1985), pp. 465-511.

7. Douglas W. Elmendorf, The Effect of Interest-Rate Changes on Household Saving and Consumption: A Survey, FEDS Working Paper 96-27 (Federal Reserve Board, July 1996).

8. The cost of corporate and overall capital could fall even under an income-based tax replacement. The Gephardt proposal, however, might not be as likely to cause reductions in the cost of capital. That proposal does not integrate personal and corporate taxes and eliminates some business tax preferences.

9. For a more extended discussion of these effects, see Enrique G. Mendoza and Linda L. Tesar, Supply-Side Economics in a Global Economy, Working Paper No. 5086 (Cambridge, Mass.: National Bureau of Economic Research, April 1995).

10. See possible explanations discussed in Roger H. Gordon and A. Lans Bovenberg, "Why Is Capital So Immobile Internationally: Possible Explanations and Implications for Capital Income Taxation," American Economic Review, vol. 86, no. 5 (December 1996), pp. 1057-1075.

11. Jane G. Gravelle, "Simulation of Economic Effects for Flat Rate Income and Consumption Tax Proposals" (draft, Congressional Research Service, 1996).

12. For a more extended discussion of these issues, see Congressional Budget Office, Labor Supply and Taxes, CBO Memorandum (January 1996).

13. Because the proposal is revenue neutral, this estimate is based on the compensated wage elasticity.

14. Alan J. Auerbach and others, Fundamental Tax Reform and Macroeconomic Performance (paper prepared for the Joint Committee on Taxation, January 1997).

15. This is the message suggested by most of the simulation models featured in a symposium on tax modeling held by the Joint Committee on Taxation on January 17, 1997. Labor supply would increase between zero and 7 percent, while the capital stock would increase by as much as 30 percent.

16. Numbers from these three models come from unpublished reports and from results appearing in Aaron and Gale, eds., Economic Effects of Fundamental Tax Reform.

17. Alan Auerbach, "Tax Reform, Capital Allocation, Efficiency, and Growth," in Aaron and Gale, eds., Economic Effects of Fundamental Tax Reform.

18. Don Fullerton and Diane Lim Rogers, "Lifetime Effects of Fundamental Tax Reform," in Aaron and Gale, eds., Economic Effects of Fundamental Tax Reform; and Diane Lim Rogers, "Assessing the Effects of Fundamental Tax Reform with the Fullerton-Rogers General-Equilibrium Model" (paper prepared for the Joint Committee on Taxation, January 1997).

19. Dale W. Jorgenson and Peter J. Wilcoxen, "The Long-Run Dynamics of Fundamental Tax Reform" (paper prepared for the American Economic Association annual meeting, January 1997).

20. Lawrence H. Summers, "Capital Taxation and Accumulation in a Life Cycle Growth Model," American Economic Review, vol. 71 (September 1981), pp. 533-544; and Owen J. Evans, "Tax Policy, the Interest Elasticity of Saving, and Capital Accumulation," American Economic Review, vol. 73 (June 1983), pp. 398-410.

21. Auerbach and others, "Fundamental Tax Reform and Macroeconomic Performance."

22. Roger Brinner, Mark Lasky, and David Wyss, "Market Impacts of Flat Tax Legislation," in DRI/McGraw Hill, Review of the U.S. Economy (Lexington, Mass.: DRI/McGraw Hill, June 1995), pp. 29-37.

23. Congressional Budget Office, Recent Developments in the Theory of Long-Run Growth: A Critical Evaluation, CBO Paper (October 1994).

24. Eric M. Engen and William G. Gale, "The Effects of Fundamental Tax Reform on Saving," in Aaron and Gale, eds., Economic Effects of Fundamental Tax Reform, pp. 83-112.

25. Auerbach, "Tax Reform, Capital Allocation, Efficiency, and Growth."

26. Robert E. Hall and Alvin Rabushka, The Flat Tax, 2nd ed. (Stanford, Calif.: Hoover Institution Press, 1995).

27. Auerbach, "Tax Reform, Capital Allocation, Efficiency, and Growth."

28. Martin Feldstein, The Effect of a Consumption Tax on the Rate of Interest, Working Paper No. 5397 (Cambridge, Mass.: National Bureau of Economic Research, December 1995).

29. Auerbach, "Tax Reform, Capital Allocation, Efficiency, and Growth."

---