The Case for IRA Expansion

American Council for Capital Formation
August 1996

The American Council for Capital Formation submitted recommendations to the Platform Committees of both the Democratic and Republican national conventions for tax policy initiatives to promote saving, investment, and economic growth. What follows is the ACCF's analysis of the impact on the American economy of significant expansion of Individual Retirement Accounts.

Background Summary

As it seeks policy measures to encourage personal saving, the Platform Committees of the Democratic and Republican Parties should give serious consideration to a plank calling for a significant expansion of Individual Retirement Accounts (IRAs).

IRAs were developed in the 1970s to encourage individuals to save for retirement. The 1981 tax cut expanded IRAs so that any wage earner could contribute up to $2,000 annually to an IRA; a couple with two wage-earners could contribute $4,000. A couple with a nonworking spouse could contribute $2,250. Contributions were tax deductible and taxes on funds accumulating in an IRA were deferred until the funds were withdrawn. Individuals had to be age 59 1/2 or older to withdraw funds from IRA accounts without penalty.

The Tax Reform Act of 1986 curtailed IRAs so that IRA contributions are no longer fully deductible except for workers without employer pensions, families with annual incomes under $40,000, and individuals with annual incomes under $25,000. These income limits are not indexed.

Recent experience with IRAs suggests that lower tax rates on saving do have a positive impact on saving rates. A combination of tax relief in the form of deductions for saving and postponement of taxes on income generated by IRAs, well-structured market instruments, and aggressive marketing campaigns by financial institutions can raise the personal saving rate, according to well-documented academic studies.

The Case for IRA Expansion

A substantial expansion in tax-deductible IRAs would produce a significant impact on personal saving and would have important economy-wide consequences.

• Stimulate Sluggish U.S. Saving
  
  The U.S. personal saving rate dropped from 5.4 percent of GDP in 1960-1980 to 3.6 percent in 1991-1996 (see Table 1). Saving makes possible the productivity-enhancing investment that is critical to raising real wages for both skilled and unskilled workers. Investment in the United States averaged 8.9 percent of GDP from 1960 to 1980, but since 1991 it has averaged only 4.6 percent, and is unlikely to increase unless U.S. domestic saving increases. The downward trend in U.S. net private domestic investment must be reversed if real incomes are to increase. U.S. family income has been nearly stagnant since the mid-1970s and in recent years family income has actually fallen. For example, real median household income was $39,869 in 1989; income has declined each year to $36,959 in 1993.
  
  The U.S. net saving rate over the 1973-1993 period is also low relative to other countries, averaging 5.4 percent in the United States compared to 19.2 percent in Japan, 10.6 percent in Germany, and 8.3 percent in Canada (see Table 2). In addition, looming in the future is the need to finance the retirement of the baby-boom generation. Research suggests this generation's saving rate is only one-third the amount needed for secure retirement.
  
  Recent reports that the U.S. saving rate should not be cause for concern (for example, see the new study by the McKinsey Global Institute) seem to overlook several key factors. First, U.S. saving is low not only in comparison to our international competitors (see Table 2) but also in comparison with our historical record. U.S. net domestic saving averaged 9.3 percent over the 1960­p;80 period compared to only 3.3 percent since 1991. Second, the argument that rates of return on investment are higher in the United States than abroad and therefore we need not save as much as our less productive competitors seems questionable. If returns on investment are significantly higher in the United States than abroad, why is so much U.S. capital flowing into foreign investments? (Over the past decade, the flow of U.S. investment abroad has risen by over 600 percent.) In fact, new data from the Union Bank of Switzerland show that the average yield on equities over the past ten years was 17.4 percent per year in the Netherlands, 14.5 percent in Switzerland, 13.6 percent in the United States, and 13.2 percent in the United Kingdom. Finally, if capital were truly much more productive in the United States than abroad, we would expect that inflows of foreign capital would drive down returns on U.S. investment so that rates of return around the world would tend to equalize.
  
  In addition, scholarly studies by Stanford University Professor B. Douglas Bernheim and others conclude that U.S. households are saving at only one-third the rate needed to enjoy the same standard of living when retired as they now enjoy.
  

  ---

• Increase Retirement Saving Prominent public finance economist and scholars, including former Council of Economic Advisers Chairman Martin Feldstein, Treasury Deputy Secretary Lawrence Summers, and Professors David A. Wise of Harvard University, James M. Poterba of Massachusetts Institute of Technology, Steven F. Venti of Dartmouth College, Jonathan Skinner of the University of Virginia, and Richard A. Thaler of Cornell University, have concluded that IRAs--especially tax-deductible IRAs--do result in new saving. More than a dozen scholarly studies, using a variety of data sources and employing several differential statistical approaches, have examined whether targeted saving vehicles such as IRAs impact saving. For example, Professor Steven Venti's testimony before a Senate Finance Subcommittee examined saving data from a Survey of Income and Program Participation for three different age groups (families reaching age 60-64 in 1984, 1987 and 1991). Professor Venti found a striking increase in saving the longer the family has been exposed to the targeted retirement programs: IRAs, 401(k)s, and Keoghs (see Table 3). The growth in IRA asset balances is astounding, Professor Venti noted. The typical member of the youngest family--with nine years of exposure to targeted retirement saving programs--has nearly three times the targeted retirement assets of the oldest cohort. There is a comparable increase in total assets as well. In contrast, among families without IRAs, the youngest families have only about 75 percent the financial assets of the older families ($1,691 vs. $2,247). Professor Venti concluded that since total financial assets, including balances in IRAs, are much larger for the younger cohort in 1991 than for the older cohort in 1984, targeted retirement savings programs did stimulate new saving over the period.

• Benefit Middle-Class Taxpayers Most contributions to IRAs are made by middle-income families, according to Professor Venti. At the IRA program's peak in 1986, about 16 percent of tax filers contributed to an IRA and about 29 percent of all families with a head of household under age 65 had positive IRA asset balances. Of course, households that had low incomes or had a young household head were less likely than wealthier or older households to have an IRA. However, at the peak of the program 75 percent of all IRA contributions were accounted for by families with annual incomes less than $50,000. IRA participation is also closely related to age. Nearly 50 percent of all families in the 55-65 age interval had an IRA account. Thus, even though less than one-third of all families have an IRA at a single point in time, over their lifetimes at least half of all families could be expected to participate in an IRA program.
  
  
• Put U.S. Savers on More Equal Footing With Savers in Other Countries Many industrial countries, including Canada, Australia, France, the United Kingdom, the Netherlands, and Belgium, encourage saving through IRA-type programs. Most of these countries permit larger IRA contributions than does the United States, and most allow a tax deduction for the IRA. For example, Canada permits taxpayers covered by an employer's pension plan to contribute $3,500 per year to a Registered Retirement Saving Plan (RRSP); persons without an employer pension plan could contribute up to $13,500 in 1994. Canada's generous RRSP plan may help explain why that country has a higher personal saving rate than the United States (9.6 percent of disposable income over the 1986-1994 period, compared to 4.7 percent for the United States). Other countries with IRA-type saving programs also have personal saving rates which exceed that of the United States. For example, the personal saving rate was 6.3 percent in Australia, 12.5 percent in France, and 9.1 percent in the United Kingdom over the 1986-1994 period.
  
  
• Generate New Saving IRAs create new savings and do not simply shift monies from already existing accounts. A recent study by Capital Research Associates shows that American families simply do not have the money to keep shifting from other accounts into their IRAs. The study classified families by the age of the family head and found, specifically, that families headed by individuals 44 years of age and below have median net financial assets of just $700. Even families on the verge of retirement, headed by individuals aged 55-64, have median net financial assets of only $6,880. Overall, the median level of net financial assets for all U.S. families amounts to approximately $1,000, the analysis concluded, thus suggesting that American families cannot simply shift assets from other accounts to IRAs.
  
  
• Reinforce the Self-Control Needed for Saving The importance of self-control is usually ignored in the formulation of a national savings policy. According to Professor Richard Thaler, IRAs help people save by making it more difficult to access their saved money than if it were left in a bank savings or checking account. Professor Thaler explains that households allocate funds, implicitly or explicitly, into categories, or mental accounts. Some funds (for example, those in cash or the checking account) are designated for current consumption. Others (for example, those in the savings account) are for rainy days or special occasions. Self-control and mental accounts come together because mental accounts vary in how tempting they are to invade. Money in the wallet is more tempting to spend than money in the checking account, which in turn is more tempting than money in the savings account. Even less tempting are funds explicitly set aside for retirement, such as money in an IRA plan. Once this concept is factored in, IRAs can increase saving even if (as some critics charge) all the money put into IRAs would have been saved anyway. Thus, by putting funds into a less accessible account, the IRA increases long-term saving.