Testimony | Retirement

Testimony before the Subcommittee on Select Revenue Measures of the House Committee on Ways and Means

Chairman Neal, Ranking Member English, and distinguished members of the Subcommittee, I appreciate the opportunity to appear before you today to discuss ways to expand retirement security. The opinions I will express are my own and not necessarily those of the Economic Policy Institute.

Before I begin, I would like to clarify that the issues I plan to address are relevant to all types of individual savings accounts, not just IRAs. The distinction between IRAs and defined-contribution plans is often immaterial, because most funds in IRAs were rolled over from defined-contribution plans, and some IRAs, like SIMPLE IRAs, are very similar to defined-contribution plans.

In recent years, the focus of retirement experts and policymakers has been on proposals to increase retirement savings through payroll deductions into savings accounts like IRAs or 401(k)s. The latest such proposal is the Automatic IRA Act, which has been introduced with bipartisan support in both houses of Congress.

These proposals are designed to overcome behavioral obstacles to participating and contributing to retirement accounts—by, for example, requiring workers to opt out of a plan rather than opting in. The Automatic IRA Act, for example, would require any employer with more than 10 employees who does not have a retirement plan to offer automatic deduction to an IRA.

The consensus is that this approach cannot hurt. The problem is that it will not help much either. This is because these proposals do not make 401(k)s or IRAs a better deal for ordinary workers, they just make it easier for them to put money into an account.

These proposals are a distraction from the real problem, which is that most workers have not been well served by the shift from traditional pensions, which are truly automatic, to IRAs, 401(k)s and other individual savings accounts that not only require workers to sign up for an account, but also shoulder most or all of the cost and the risk.

One might argue that the shift from traditional pensions to individual accounts is simply a reflection of market forces. But tax incentives for savings accounts represent an enormous subsidy from the federal government, with little to show for it.

There are three big problems with these supposed incentives, which Automatic IRAs would do little to change. First, only households that owe income tax are eligible for the subsidies, the size of which also depends on the household’s tax bracket.1 Even if that were not the case, high-income households have more disposable income to set aside, so subsidies would still disproportionately benefit them. Finally, the incentive effect is weak or non-existent, since there is no way to ensure that tax incentives encourage new saving.

The present value of tax expenditures for 401(k) and IRA contributions in 2007 amounted to nearly $135 billion.2 According to the Urban-Brookings Tax Policy Center, roughly 70% of these subsidies go to those in the top 20% of the income distribution, and almost half go to the top 10%.3

These tax breaks are not just unfair, they are ineffective, because they mostly cause wealthy households to shift savings to tax-favored accounts rather than increase overall savings—thus the paradox that taxpayers are giving up more and more revenue to promote retirement savings while retirement security declines.

Tax incentives are supposed to do more than lower the taxes paid by wealthy households. They are supposed to help workers save for retirement. Yet enrollment in employer-based retirement plans has remained stagnant at around 50% of full-time workers. Due to inadequate contributions, cash-outs and other leakages, Federal Reserve data show that the median 401(k) and IRA account balance of workers approaching  retirement was $60,000 in 2004, not even enough to buy a $400 per month annuity.4

The implicit assumption behind the “Auto IRA” (and “Auto 401(k)”) approach is that the problem lies with the worker, not with the retirement options she faces. But even in the best-case scenario—a participant who contributes regularly and does not touch the savings until retirement—high fees may erode a quarter or more of her nest egg compared to savings pooled in a cost-efficient pension fund.

Meanwhile, the worker is likely to be getting little help from the federal government, due to an upside-down incentive structure that gives a wealthy family in a 35% tax bracket a tax break three and a half times more valuable than a family in a 10% tax bracket, even if each family contributes the same dollar amount to a tax-favored account. In other words, those who need the least help saving get the most. Thus it is not surprising that only about three out of ten households received a tax break for contributing to a defined contribution plan or IRA in 2004.5

These problems are compounded by the problem that individual investors feel they must choose between low, fixed returns and gambling with their nest egg. Retirement experts often bemoan the tendency of many 401(k) participants to invest in money market funds, but it is hard to argue against conservative investments when you consider that bear markets can last for a decade or longer. Other people, of course, take the opposite approach, investing all their retirement savings in risky stocks in a desperate attempt to catch up. In contrast, traditional pension funds invest in diversified portfolios and pool the savings of people who retire at different times, smoothing investment returns across generations.

Incidentally, the decision-making problem is not limited to those with little formal education. Los Angeles Times reporter Peter Gosselin found several Nobel Prize-winning economists willing to admit that they could not decide how to allocate their retirement savings.6

Finally, even a worker who saves steadily and has good luck with investments may outlive his or her savings. Theoretically, individuals can insure themselves against longevity risk by purchasing life annuities, but an adverse selection problem makes annuities expensive on the individual market, and people are often stymied by the difficulty of choosing among investment products.

Some of these problems can and should be fixed. Congress has begun to address the issue of hidden 401(k) fees, for example. But there are inherent advantages to traditional pensions, because pooling allows employers or the government to insure workers against most financial and longevity risks while taking advantage of economies of scale. Thus, the shift from traditional pensions to individual accounts has increased administrative costs while saddling workers with risk that would be easy to insure against in a group plan.

In other words, closing the retirement gap is not simply a question of increasing contributions, but also ensuring that benefits are broadly shared and retirement savings and income are secure.

We need a whole new approach. We need to replace IRAs and 401(k)s with something better. And though traditional pensions work well for large, stable employers, others are not in a position to take on long-term pension liabilities.

Last year, the Economic Policy Institute asked retirement expert Teresa Ghilarducci—who unfortunately could not be here today—to come up with a replacement for the current system of individual accounts. The resulting Guaranteed Retirement Account plan7  is a hybrid that combines the best features of defined-benefit and defined-contribution plans, including steady and predictable employer a
nd employee contributions, low administrative costs, and guaranteed lifetime benefits.

The GRA plan would reapportion federal subsidies, which now disproportionately go to high-income families, while insuring participants against financial and longevity risk. It would start by converting tax expenditures for defined-contribution plans and IRAs into flat refundable credits. A Tax Policy Center analysis of the GRA plan found that this by itself would make 58% of taxpayers better off and only 16% of taxpayers worse off, most of them in the top income quintile. And unlike high-income households, low- and middle-income households would not fully offset this increase in savings with dis-saving in other forms.

Unlike “Auto IRA” and “Auto 401(k)” plans that focus on increasing voluntary contributions to savings accounts, the GRA plan squarely addresses the issue of adequacy through mandatory contributions, efficiency gains, and plugging cash-outs and other leaks. Like “pay or play” healthcare plans, the plan calls for all workers not enrolled in an equivalent or better pension plan to enroll in a Guaranteed Retirement Account. Contributions equal to 5% of earnings up to the Social Security earnings cap would be deducted along with payroll taxes and credited to individual accounts, though the funds would be pooled and invested together.

The cost of these contributions would be split equally between employers and employees. However, employee contributions would be offset in whole or in part through an inflation-indexed $600 refundable tax credit that would take the place of tax breaks for defined-contribution accounts and IRAs.

GRA accounts would be administered by the Social Security Administration, and the funds managed by the Thrift Savings Plan or similar body, which in turn would outsource investment functions to an outside provider. Though the funds would be invested in financial markets, participants would earn a fixed 3% rate of return adjusted for inflation and guaranteed by the federal government. If the trustees determined that actual investment returns were consistently higher than 3% over a number of years, the surplus would be distributed to participants, though a balancing fund would be maintained to ride out periods of low returns.

Workers would be able to track the dollar value of their accumulations, the same as with 401(k)s and IRAs. However, account balances would be converted to inflation-indexed annuities upon retirement to ensure that workers would not outlive their savings.

The result is that participants would be guaranteed a secure retirement after a lifetime of contributions. A prototypical worker could expect a benefit equal to roughly 25% of pre-retirement income after 40 years. Since Social Security provides such a worker with a benefit equal to roughly 45% of pre-retirement income at age 65, the total replacement rate would be approximately 70% of pre-retirement income, which is considered the minimum necessary to avoid a drop in living standards upon retirement.

The GRA plan gives workers what they want—a simple, fair and effective way to save for retirement. According to the Rockefeller Foundation’s American Worker Survey, Americans are equally concerned about having access to health care and pension benefits, and they are about three times more likely to want a job that guarantees health coverage and a pension rather than one that pays more.

Americans are seeking financial security after flirting with day trading, stock options, and house flipping. Quasi-free market solutions relying on inequitable and ineffective tax breaks have lost much of their appeal. Instead of tax breaks for a lucky few, the government would be telling all workers, “we’ll throw in the first $600, and the rest is up to you and your employer.”

Admittedly, Americans remain leery of government solutions and convinced that retirement is unaffordable. The GRA plan addresses these concerns through advance funding, shared employer-employee contributions, and a revenue-neutral reallocation of government subsidies. Thus, it is important to emphasize that the GRA plan would not increase the federal deficit and would reinforce the link between work and retirement benefits, encouraging people to work longer.

Current economic conditions highlight the need for a new plan. In contrast to Social Security and defined benefit pension plans, individual accounts like 401(k)s and IRAs are not insulated from the effects of economic downturns, since asset markets tend to move pro-cyclically. In a recession, participants are often forced to delay retirement, which has a spillover effect on unemployed workers as vacancies shrink. A recession is also likely to reduce contributions and increase leakages, and some fund managers have already reported an increase in hardship withdrawals and loans.

The role of housing as a conduit to savings and financial security has eroded as the housing market has slumped and homes have been transformed into speculative investments or collateral for loans. Even ignoring the immediate problems associated with sub-prime loans and foreclosures, two long-term trends – increased mobility and home equity withdrawals – point to a greater need for more leak-proof and secure savings vehicle than housing, which represented nearly 40 percent of total assets held by households, according to the last Survey of Consumer Finances conducted by the Federal Reserve. (That figure may, of course, be somewhat lower today.)

I would like to say that if Automatic IRAs are the answer, you are asking the wrong question. The question should not be, “how can we make a bad system a little better,” but rather, “how can we make sure Americans have adequate and secure retirement incomes after a lifetime of work?” The answer, I think, is the Guaranteed Retirement Account plan.

Before I conclude, I should add that one of our state affiliates—the Economic Opportunity Institute in Washington State—has come up with an Automatic IRA plan that would be administered by a state agency.8 The Institute has worked closely with Mark Iwry and other architects of the Automatic IRA approach.

However, the Washington State plan would do more to protect workers than the current federal legislation. Besides expanding coverage, the plan would keep costs down by, among other things, using an existing administrative structure, pooling funds to take advantage of economies of scale, and negotiating fees with providers.

The plan is a step in the right direction, but because it is designed to work within the existing federal framework, it cannot correct the failures of this system, such as the fact that IRAs currently function more as tax shelters for the wealthy than retirement vehicles for the rest of us.


1. The Saver’s Credit is designed to address this problem, but most people with incomes low enough to qualify cannot take advantage of it because they do not owe income tax and the credit is non-refundable (William G. Gale, J. Mark Iwry, and Peter R. Orszag, “Making the Tax System Work for Low-Income Savers: The Saver’s Credit,” Urban-Brookings Tax Policy Center Issues and Options, July 2005).
2. Office of Management and Budget, Analytical Perspectives, FY 2009 Budget, Table 19-4.
3. Leonard E. Burman, William G. Gale, Matthew Hall, and Peter R. Orszag, “Distributional Effects of Defined Contribution Plans and Individual Retirement Accounts,” Urban-Brookings Tax Policy Center, 2004.
4. Survey of Consumer Finances, as cited in Alicia H. Munnell and Annika Sundén, “401(K) Plans Are Still
Coming Up Short,” Center for Retirement Research Issue Brief, March 2006.
5. Burman et al., 2004.
6. Peter G. Gosselin, “Experts Are at a Loss on Investing,”  Los Angeles Times, May 11, 2005.
7. Teresa Ghilarducci, “Guaranteed Retirement Accounts: Toward retirement income security,” EPI Briefing Paper, November 20, 2007. http://www.sharedprosperity.org/bp204/bp204.pdf
8. Information about Universal Voluntary Retirement Accounts is available at the EOI website a http://www.eoionline.org/washington_voluntary_accounts/voluntary_accounts.html

Ross Eisenbrey is vice president of the Economic Policy Institute in Washington, D.C.


See related work on Retirement

See more work by Ross Eisenbrey