Book reviews for The Journal of Pension Economics and Finance reflect reviewers' own views and in no way represent the views of the institution(s) with which they are affiliated.
With the retirement of the baby boomers well under way, it is high time to focus on the economic issues associated with population aging. One question is whether Americans save enough for retirement. Another question relates to the fiscal pressures associated with population aging: How should we change our entitlement and tax systems to adjust to the realities of an older population?
In “How America Supports Retirement,” the Investment Company Institute's Peter Brady offers some timely analysis that addresses both of these questions. His book has three main parts. In the first and largest part, he provides an accessible but detailed look at the economic benefits of the main programs used to encourage retirement savings: Social Security and tax deferral programs like individual retirement arrangements and employer pensions. In the second, he argues that too much emphasis is put on the distributional consequences of such programs, and that a progressive tax and spend system does not require that every policy be progressive. Finally, he provides a different framework for analyzing the merits of our current retirement programs.
Brady disputes the common view that tax deferral primarily benefits higher income taxpayers and that, as a result, we have an “upside down” retirement system that provides saving incentives to high-income taxpayers rather than to the low-income taxpayers who need them. He contends that, to understand the distributional aspects of tax deferral programs, it is necessary to examine them in conjunction with Social Security, because the distributional benefits of tax deferral programs are inextricably linked with those of Social Security.
Brady analyzes both tax deferral and Social Security under a “tax expenditure” framework, in which the benefits of the programs are calculated as their net present value (present value of future income less contributions) relative to what would be received if the investment were fully taxable. Brady shows that the reduction in taxes associated with a contribution to a retirement account is not a valid measure of the benefits, because most of this simply represents a deferral of taxes. Taxes are lower when the investment is made, but higher when the assets are withdrawn in the future. The true benefit of tax deferral comes from the fact that investment earnings are untaxed. Brady also shows that the benefits of deferral are larger when future tax rates are lower than current ones.
He then compares the benefits from Social Security and tax deferral for six representative households, each with different income. He shows that Social Security is highly progressive, and that while tax deferral benefits do accrue more to the wealthy, the retirement system as a whole is still progressive. Brady argues that the wealthy benefit more from tax deferral primarily because Social Security replaces a larger share of income for lower-income households than for upper-income households. Thus, lower-income households have less need to save for retirement, and so don't contribute much to private retirement accounts.
While this is undoubtedly true, I think Brady overstates the point. He argues, for example, that the benefit of tax deferral does not rise linearly with the tax rate, so that deferral is not necessarily more valuable for wealthier people. However, as he shows, the benefit of tax deferral is the fact that investment income is never taxed. The higher the tax rate, the more valuable this exemption. Brady's calculations make an arguably false comparison by asking how the benefit of investing an additional pre-tax dollar differs by income. Because higher income people have higher marginal tax rates, a pre-tax dollar of pension contribution is a much smaller investment for a high-income person than a low-income person. If, instead, he had analyzed the benefits of an after-tax dollar of contribution, he would have found that, as common sense suggests, the benefits of tax deferral do increase with the tax rate.
Brady then makes the case for not focusing on “microprogressivity”, the effect of specific policies on progressivity. Instead, he argues, policymakers should focus on the progressivity of the tax system as a whole. I wholeheartedly agree with this view. Nonetheless, it is important to examine the progressivity of different elements of the tax system in order to know how to do this. If a policymaker wants to raise taxes in a progressive manner, where should she look? It is clear even in Brady's analysis that the benefits of tax deferral go disproportionately to the rich, and that limiting these benefits would increase progressivity.
In the final chapter of the book, Brady argues that tax deferral should be maintained. He argues that tax deferral increases fairness because it promotes “income averaging.” He also argues that tax deferral promotes economic growth, because it removes the disincentives for saving inherent in a pure income tax system (and hence, rather than providing an upside-down incentive for saving, tax deferral simply removes the larger disincentive faced by high-income people under an income tax.) Finally, he argues that many of the proposals to change tax deferral would increase complexity. Of these three arguments, only the second is compelling. (Income averaging doesn't require exempting investment income from tax, and there are many very simple proposals to limit retirement plan tax benefits.)
Many economists might accept the view that lower taxes on saving promotes investment and hence growth. But I don't think this is at all clear. As Brady notes, under the current system, investments in tax-preferred accounts are capped for many rich people; for these taxpayers, tax deferral has no effect on the incentive to increase saving. Furthermore, for any given level of desired tax revenues, lower taxes on saving mean higher taxes on other things—most likely labor income—which also distorts behavior. It is not obvious that cutting back on saving incentives instead of raising marginal tax rates in a distributionally-equivalent way would be bad for growth.
Despite my disagreements with some of Brady's conclusions, this book is well worth reading for anyone interested in thinking about the nuts and bolts of retirement policy. It is well written, full of information displayed in clear charts, and thought provoking.
