Introduction.

The United States' personal saving rate as measured in the national income and product accounts has declined for nearly two decades, and it is currently near historic lows. During the same period, however, real asset values have risen sharply, particularly for corporate stock and owner-occupied real estate. The result is that household net worth relative to GDP or disposable income is currently higher than during much of the last two decades. This situation is almost tailor-made for economic controversy. If one is interested in personal saving because of the close links between national saving and national investment, and because a nation's long-term productivity growth is closely linked to its investment rate, then the low personal saving rate is a potential source of concern. If one is worried about whether future retirees will have adequate resources to support their lifestyle after they leave the labor force, however, then the accumulated stock of wealth is a focus of attention. Not just the aggregate amount of wealth, but its distribution across households, is critical. It is possible for total wealth to be more than adequate to support the retirement consumption of an aging cohort, but for the distribution to be unequal enough to result in many retired households living in economic hardship. For many economists and policy observers, the distribution of wealth is also a variable of independent interest. Recent stock price appreciation has increased the degree of wealth inequality in the United States. Some have called for policy reforms to reign in wealth inequality, while others have argued that much of the economic expansion during the 1990s was due to a low tax rate environment, and that further reductions in the taxation of capital income are warranted.

Controversy within economics about whether the U.S. saving rate is too low, and whether the Baby Boom generation and cohorts that follow it will be adequately prepared for retirement, translates into an active public policy debate about stimulating private saving and influencing the distribution of wealth. There are many policy instruments that can in principle affect both the national and the personal saving rate, including the federal budget deficit, regulatory policies toward credit markets that affect household borrowing and credit access, and public policies that affect corporate pension funds. Yet most of the policy debate on encouraging personal saving, and on channeling personal saving to particular objectives, such as retirement or financing higher education, focuses on the tax system.

The taxation of capital income has been an active subject of research and debate in public finance for at least a century. A long-standing debate focuses on the choice between an income tax and a consumption tax as the preferred foundation for a tax system. The heart of this debate, which has attracted many distinguished economists and several blue-ribbon government commissions, concerns whether capital income should be taxed at all. In the last two decades, however, a number of new issues in tax policy toward saving have emerged. Beginning with the introduction of Individual Retirement Accounts in the early 1980s, there has been a move toward designing specialized tax incentives that encourage saving for particular activities. These have included saving for retirement, with IRAs, 401(k) plans, and 403(b) accounts; saving for education, with 529 plans and "Education IRAs;" and saving for future medical expenditures, with the recently-introduced Health Saving Accounts. These accounts do not prohibit savers from using accumulated balances for purposes other than the targeted objective, but they typically impose penalty taxes on such withdrawals. The rise of these accounts has introduced a complicated set of new provisions into the tax code. These provisions govern eligibility, required minimum distributions, and the tax treatment of contributions and withdrawals.

While targeted saving incentives have become an active topic of policy debate, they are not the only element of tax policy and saving that has attracted attention from policy analysts. Recent shifts in the perceived distribution of wealth have generated new interest in the taxation of estates and gifts. The estate tax was scaled back in the tax bills passed in 1997 and in 2001. The 2001 legislation eliminated the estate tax for deaths that occur in 2010, but the budget guidelines governing the legislation made it impossible to enact a permanent repeal. The estate tax is, therefore, currently scheduled to return to its pre-2001 form in 2011, the year after the one-year repeal. This situation suggests that estate tax reform will continue to attract attention from tax legislators in coming years.

Stimulated by the ongoing public policy interest in taxation and saving, the National Bureau of Economic Research undertook a multi-year research project on how the tax system affects saving and wealth accumulation. The project focused on the impact of targeted saving programs on household behavior, and on the links between saving, the distribution of wealth, and tax policy. The project involved researchers from many different universities as well as the U.S. Treasury Department, and it touched on a wide range of issues in saving policy. A core element of this project was upgrading NBER's TAXSIM program, a large computer program for calculating federal as well as state income tax liability for a randomly selected subset of households in the universe of tax fliers. The TAXSIM program can be used to calculate tax burdens on capital income historically, back to 1960, and it can also be used in tandem with plausible assumptions about future population and productivity growth to project future capital income tax burdens.

The researchers who participated in the NBER project on taxation and saving are widely dispersed in geographical terms. This makes it very useful to bring all the researchers together to discuss their findings, their research strategies, and their future agendas. To this end, the NBER research group gathered on August 1-2, 2003, in Chatham, Massachusetts. Each research team had an opportunity to present their research findings and to discuss potential improvements with the other experts gathered at the meeting. This special issue of the National Tax Journal includes revised versions of six of the studies that were presented at that meeting.

The six papers in this volume address issues that bear on the current, and ongoing, tax policy debates with regard to personal saving and wealth accumulation. The papers can be divided into three groups: three studies that focus on targeted saving accounts, two that examine issues related to the estate tax, and one that analyzes how the Alternative Minimum Tax, a rapidly-expanding "parallel tax system" to the ordinary income tax, affects the incentives for taxpayers to save. Since summaries of these papers appear elsewhere in this volume, they are described very briefly in this introduction.

The three papers that address targeted tax incentives focus on saving for educational expenses and for retirement. Susan Dynarski's study of "Who Benefits from the Education Saving Incentives" presents new evidence on the net benefits that accrue to households that save through 529 plans and Coverdell accounts. The paper explores the interaction between household wealth accumulation in these accounts and eligibility for need-based college financial aid. It thereby estimates the net rate of return to saving in a tax-preferred education saving account, and evaluates the net incentive for saving through these programs. For some households in income and wealth ranges over which financial aid is scaled back in response to higher family income or financial assets, the net rate of return to saving through targeted educational saving programs can be low. For other households, however, the tax-deferred accumulation offered by these accounts provides a powerful way to save for college expenses and related outlays. This heterogeneity suggests that these tax incentives are likely to have different behavioral effects at different places in the income and wealth distribution.

The second paper on tax-deferred saving programs is Gary Engelhardt and Brigitte Madrian's study of "Employee Stock Purchase Plans." ESPPs are employer-sponsored programs that permit employees to save by purchasing company stock. Contributions to these plans, and accumulations within them, are subject to favorable tax treatment. After describing the tax incentives for saving in these plans, the paper uses employment records from a large firm that sponsors an ESPP to analyze how potential savers respond to this program. The evidence suggests that many workers fail to participate in these programs in spite of their substantial tax benefits. The authors argue that factors other than maximization of after-tax investment returns must be at work to explain this. Liquidity constraints or other factors that produce a short-term demand for cash could potentially account for this pattern. The results offer an important caveat in analyzing the behavioral effects of tax incentives for saving. After-tax returns do not appear to be the only factor driving household saving decisions.