Do retirement saving programs increase saving?
Two saving programs introduced in the early 1980s were intended to encourage individual saving. IRAs rapidly became a very popular form of saving in the United States after they were made available to all employees in 1982. Any employee could contribute $2,000 per year to an IRA account, and a non-working spouse could contribute $250. The contribution was tax deductible. Annual contributions grew from about $5 billion in 1981 to about $38 billion in 1986, approximately 30 percent of total personal saving. Contributions declined precipitously after the Tax Reform Act of 1986, even though the legislation limited the tax deductibility of contributions only for families who had annual incomes over $40,000 and who were covered by an employer-provided pension plan. By 1990, less than $10 billion was contributed to IRAs. Whereas over 15 percent of tax filers made contributions in 1986, only 4 percent contributed in 1990.
The other program, the 401(k) plan, grew continuously and almost unnoticed, with contributions increasing from virtually zero at the beginning of the decade to over $51 billion in 1991. In 1991, almost 25 percent of families contributed to a 401(k). Deposits in 401(k) accounts are also tax deductible and the return on the contributions accrues tax free; taxes are paid upon withdrawal. But these plans are available only to employees of firms that offer such plans. Prior to 1987 the employee contribution limit was $30,000, but the Tax Reform Act of 1986 reduced the limit to $7,000, and indexed this limit for inflation in subsequent years. The contribution limit is $9,235 for the 1994 tax year.
By 1991, contributions to all personal retirement saving plans exceeded contributions to traditional employer-provided pension plans. It seems evident that if it were not for the 1986 tax legislation, personal retirement saving would have been much larger. Whether these programs increase net saving can be of critical importance to future generations of older Americans and to the health of the economy in general. The issue remains an important question of economic debate. In a series of papers based on very different methods of analysis, Steven F. Venti and I and James M. Poterba, Venti, and I have concluded that contributions to these accounts represent new saving in large part.
In determining the effect on saving of IRAs and 401(k)s, the key problem is saver heterogeneity: some people save and others don't, and the savers tend to save more in all forms. For example, families with IRAs have larger balances in all financial assets than families without IRAs. But this does not necessarily mean that IRAs explain the difference. Thus a continuing goal of the analysis has been to consider different methods of controlling for heterogeneity, or individual-specific saving effects. The methods that could be used when each analysis was conducted were largely dependent on the data available at that time. As new data became available, we used alternative and possibly more robust methods.
Early Parametric Analysis of Substitution at the Outset of the IRA Program
When we began this work in the mid-1980s, data were available for a limited time period; assets typically could be measured at only two points in time, one year apart. To use these data, Venti and I developed an econometric model for estimating the relationship between IRA saving and other saving. Within a framework that allowed for any degree of substitution between IRA and non-IRA saving, we asked: after controlling for age, income, other personal characteristics, and accumulated housing and financial assets, do persons who save more in the form of IRAs in a particular year save less in other financial asset forms? Given age and income, this approach used accumulated financial assets to control for "individual-specific" saving effects. The analysis accounts for the explicit limit on IRA contributions and places substantial emphasis on the change in non-IRA saving after the IRA limit is reached. The first results using this approach were based on early data from the 1983 Survey of Consumer Finances.(1) Subsequent analysis was based on the 1980-5 Consumer Expenditure Surveys and the 1984 panel of the Survey of Income and Program Participation (SIPP).(2)
The results suggested that the majority of IRA saving, even at the outset of the program, represented net new saving, and was not accompanied by large-scale reduction in other financial asset saving. According to these findings, increases in the IRA limits would lead to substantial increases in IRA saving and very little reduction in other saving. If the IRA limit were raised, one-half to two-thirds of the increase in IRA saving would be funded by a decrease in current consumption, and about one-third by reduced taxes; only a very small proportion - at most 20 percent - would come from other saving.
Following Individuals Over Time at the Outset of the IRA Program
As new data were released it became possible to use different methods to control for heterogeneity. We used two approaches based on traditional panel data.
Changes in Other Saving When Individual IRA Participation Status Changes
If non-IRA saving is reduced when IRA saving is increased, then when a household that was not contributing begins to contribute, that household should reduce non-IRA saving. Likewise, when a household that was contributing stops contributing, non-IRA saving should increase. But when the same families are tracked over time, there is little change in other financial asset saving when families begin to contribute to an IRA (or when they stop contributing). The SIPP panel data allow calculation of the change in non-IRA saving when IRA contributor status changes. This simple calculation controls directly for changes in saving behavior across families since it is based on changes over time for the same families. We considered the change in other (non-IRA) saving between 1984 and 1985 by IRA contributor status.(3)
These data reveal little substitution. The key information in this approach is the change in other financial assets when families began to contribute to an IRA. In particular, we consider the change for families that did not contribute in 1984 but did contribute in 1985. If the new IRA saving substituted for other financial asset saving, one would expect other saving to decline between 1984 and 1985. But the non-IRA financial asset saving of these new contributors declined by only $193 between 1984 and 1985. This decline in other saving is only a small fraction of the increase in saving from the typical family IRA contribution, about $2,300. Thus these data also suggest that even near the outset of the IRA program there was only a small reduction in non-IRA saving when IRA contributions began.
Following Households in the Survey of Consumer Finances
Using 1983 and 1986 Survey of Consumer Finances data, it is possible to compare the asset balances of the same households over time. Venti and I considered how the assets of IRA contributors changed over this time period.(4) Households that made IRA contributions over this period began the period with a median of $9,400 in other financial assets in 1983. Between 1983 and 1986, the median IRA assets of these families increased from $1,000 to $7,000. At the same time, the median of other financial assets increased from $9,400 to $13,500. These families ended the period with median total financial assets, including IRAs, of $24,000 - an increase of 100 percent over assets in 1983. We determined that the increase could not be accounted for by change in age, income, or rate of return. Thus we concluded that it was unlikely that the IRA contributions simply substituted for saving that would have occurred anyway.
Following "Like" Saver Groups Over Time
Another way to control for heterogeneity is to group families according to their saving behavior their "taste" for saving - and to follow the progress of the saving of these "like" families over time. Poterba, Venti, and I(5) followed this approach, grouping families according to whether they did or did not contribute to an IRA. We also grouped families according to whether they were eligible for a 401(k) plan and whether they contributed to that plan. Altogether, we considered eight different groups of "saver types." We focused on the change in the other saving of families in these groups who had been exposed to 401(k) and IRA plans for different time periods.
For example, data from the SIPP were available for 1984, 1987, and 1991. Random samples of family types are similar in each of these years, but the 1984 sample has had only about two years (1982 to 1984) to accumulate 401(k) and IRA balances, while the 1987 sample has had about five years, and the 1991 sample about nine years. The central question is whether longer exposure to these plans results in higher levels of saving by families who participate in the programs. The key test is whether non-IRA-401(k) assets declined as IRA and 401(k) assets increased. The answer is typically no. In each of the six different IRA or 401(k) participant saver groups, there was an increase in total financial assets between 1984 and 1991, but little change in non-IRA or non-401(k) financial assets.
It is sometimes suggested that these programs may affect households with limited assets but have little effect on wealthier households. We have addressed this issue by comparing the distribution of assets in 1984 and 1991. Again, we rely on the fact that by 1991 households had had much more time to contribute to the saving programs.(6) The higher levels of total financial assets held by IRA and 401(k) participant families in 1991 was not limited to families with large or small asset balances. Rather, the effect was evident across the entire distribution of households, from those with the least to those with the greatest assets. On the other hand, across the entire distribution, there was almost no change - between 1984 and 1991 - in the non-IRA-401(k) assets of contributors. At all points in the distribution there was a fall over time in the assets of noncontributors.
The 401(k) Eligibility "Experiment"
Another approach relies on the "experiment" that is provided by the largely exogenous determination of 401(k) eligibility. It considers whether eligibility is associated with higher levels of total saving, holding income and other demographic characteristics constant. In this case the key question is: did families who were eligible for a 401(k) in a given year have larger total financial asset balances than families who were not eligible, or, equivalently, did non-401(k) financial assets decline enough to offset the 401(k) contributions of eligible families?(7)
The inferences about the net saving effect of 401(k) contributions depend on the similarity of the saving behavior of families who are and are not eligible for a 401(k), controlling for income. It is important, for example, that the eligible group not be composed disproportionately of savers. The data show little evidence of this type of difference in saving behavior. The most compelling evidence is for 1984. In that year the two groups - eligibles and noneligibles - had about the same level of other financial assets. Thus, near the outset of the 401(k) program, families that were newly eligible for a 401(k) exhibited about the same previous saving behavior as families that did not become eligible.
Data for families with incomes between $40,000 and $50,000 illustrate the findings: In 1984, newly eligible and ineligible 401(k) families had almost identical non-401(k)-IRA assets of $5,027 and $5,082, respectively. By 1991, however, the median of total financial assets of eligibles families was $14,470, compared to $6,206 for ineligible families. But in 1991, the non-IRA-401(k) assets of the two groups were still about the same, $4,724 for eligible and $4,250 for the ineligible group. If families reduced saving in other forms when they became eligible for a 401(k) plan, the typical eligible family in 1991 would have accumulated less wealth in other financial assets than the typical noneligible family. This was not the case. Looking over all income intervals, the total financial assets of the eligible families by 1991 were typically two to eight times as large as the total financial assets of the noneligible families. Thus this approach shows no substitution of 401(k) contributions for other financial asset saving.
Indeed, for all income groups, eligible households have greater total financial assets than ineligible households at virtually all points across the entire distribution of financial assets. But there is virtually no difference across the entire distribution of the other financial assets of eligible and ineligible households.
Cohorts and the Effects of Retirement Saving Programs
In the United States
No single method provides a perfect control for all possible forms of heterogeneity. I believe that the method that comes the closest is cohort analysis. Essentially, this method compares the assets of persons who are the same except for age. Hence some groups - cohorts - had longer to contribute to special saving programs. For example, families that reached age 65 in 1984 had had only two years to contribute to an IRA or to a 401(k) plan. But families who attained age 65 in 1991 had had nine years to contribute.(8)
Venti and I find that younger families had consistently larger total financial assets than older families. The larger assets of the younger cohort are accounted for almost entirely by more assets in IRA and 401(k) plans. On average, there is no difference between the other financial assets of the older and younger cohorts. These results can be illustrated by comparing the assets of families who reached ages 60 to 64 in 1984 with the assets of families that attained that age in 1991.
To control for heterogeneity, the data for all families combined - both contributors and noncontributors - are the most compelling. The mean of total financial assets of all families that attained age 60 to 64 in 1984 was $32,807; the mean of those who attained this age in 1991 was $39,105 (in 1991 dollars and controlling for income, age, education, and marital status). The increase was accounted for almost entirely by personal retirement saving - $4,027 for the cohort that attained ages 60 to 64 in 1984 compared to $10,995 for the cohort that attained this age in 1991. There was essentially no cohort difference in other financial assets ($28,780 for the older cohort and $28,110 for the younger cohort). Because fewer than half of families participated in these programs, the median for all families is zero and therefore is not informative. The data for families who participated in personal retirement saving plans provide a better measure of the potential of the plans to augment the financial assets of retirees: the median level of total personal financial assets of contributor families that attained age 60 to 64 in 1984 was $29,847. Families that attained this age in 1991 had median total financial assets of $45,019. The median level of personal retirement plan assets of the families that reached this age in 1984 was $7,575. Those that reached this age in 1991 had $21,613. In contrast, the other financial assets of these families were about the same in 1984 and 1991 ($19,358 and $17,950, respectively). Thus there is little evidence of substitution of personal retirement saving for other financial assets. In contrast, the financial assets of families that attained age 60 to 64 in 1991 but did not participate in the personal retirement plans, were somewhat lower than similar families who reached this age in 1984.
This is the typical pattern for all ages that we considered, between 45 and 70. Indeed, the analysis suggests that if current patterns persist, families who reach retirement age 25 or 30 years from now will have much more in financial assets (at least $30,000) than families currently attaining retirement age; the difference will be due solely to assets in personal retirement accounts.
Registered Retirement Saving Plans (RRSPs) were first introduced in Canada in 1957. Like contributions to an IRA in the United States, contributions to an RRSP are deductible from income for tax purposes. Interest accrues tax free until withdrawal, when taxes are paid. The contribution limits were increased substantially in the early 1970s, and RRSPs were promoted widely. Since then, they have become a very prominent form of saving. Annual contributions grew from $225 million in 1970 to almost $3.7 billion in 1980 to $16 billion by 1992, when they accounted for about one-third of aggregate personal saving. In 1992 about 33 percent of families contributed an average of $4,180 to RRSPs. Now RRSP contributions exceed the total of employee and employer contributions to employer-provided pension plans.
Based largely on "cohort" analysis like the procedure illustrated above, Venti and I conclude that the data taken as a whole suggest that RRSPs have contributed substantially to personal saving in Canada.(9) In virtually no case do the data suggest substitution of RRSPs for other forms of retirement saving. In the two decades prior to the growth in popularity of RRSPs, the personal saving rate in Canada was typically below the U.S. personal saving rate. Since that time, the personal saving rate in Canada has become much higher than in the United States. Although it is difficult to make judgments about the RRSP effect on saving based only on the trends in U.S. and Canadian aggregate saving rates, a large fraction of the current difference can be accounted for by RRSP saving.
This series of papers suggests that the IRA and the 401(k) programs have had very important effects on saving and are likely to add substantially to the assets of future retirees. In addition, I believe that the research has another fundamental underlying theme: standard assumptions about the determinants of saving behavior leave important aspects of actual saving unexplained. For example, the Tax Reform Act of 1986 led to reductions in IRA contributions well beyond the effects that might have been predicted by models that rely on the aftertax rate of return. Even the contribution rate of persons whose aftertax rate of return was unaffected by the legislation was reduced by 40 to 50 percent. Higher-income families that lost the up-front deduction virtually stopped contributing, even though the return still accrued tax free. These results suggest that the promotion of saving plans may have an important effect on their use. The results also suggest that the up-front tax deduction may have been a critical determinant of the decision to make an IRA contribution.
Although the IRA participation rate was never more than 16 or 17 percent, the participation rate among persons eligible for a 4010(k) is above 60 percent, and is at least 50 percent even for low-wage earners. In addition to greater financial incentives provided by employer matching rates, a likely explanation is that contributions are typically by payroll deduction, and thus sheltered from current expenditure urges: the "out of sight, out of mind" feature of the payroll deduction arrangement may provide needed self control. More generally, the findings bring into question the extent to which different forms of saving are treated as perfect substitutes by real people. Thus in my view the results provide motivation to look more broadly for explanations of saving behavior.
1 S. F. Venti and D. A. Wise, "Tax-Deferred Accounts, Constrained Choice, and Estimation of Individual Saving," Review of Economic Studies 53 (1986); "IRAs and Saving," in The Effects of Taxation on Capital Accumulation, M. Feldstein, ed. Chicago: University of Chicago Press, 1987. and D. A. Wise, "Individual Retirement Accounts and Saving," in Taxes and Capital Formation, M. Feldstein, ed. Chicago: University of Chicago Press, 1986.
2 S. F. Venti and D. A. Wise, "Have IRAs Increased U.S. Saving? Evidence from Consumer Expenditure Surveys," Quarterly Journal of Economics (August 1990), pp. 661-698; and "The Saving Effect of Tax-Deferred Retirement Accounts: Evidence from SIPP," in National Saving and Economic Performance, B. D. Bernheim and J. B. Shoven, eds. Chicago: University of Chicago Press, 1991.
3 S. F. Venti and D. A. Wise, "Individual Response to Retirement Savings Programs: Results from U.S. Panel Data," Richerche Economiche, special issue, forthcoming.
4 S. F. Venti and D. A. Wise, "Government Policy and Personal Retirement Saving," NBER Reprint No. 1752, October 1992, and in Tax Policy and the Economy, Volume 6, J. M. Poterba, ed. Cambridge: MIT Press, 1992, pp. 1-41.
5 J. M. Poterba, S. F. Venti, and D. A. Wise, "401(k) Plans and Tax-Deferred Saving," in Studies in the Economics of Aging, D. A. Wise, ed. Chicago: University of Chicago Press, 1994; and "Do 401(k) Contributions Crowd Out Other Personal Saving?" Journal of Public Economics, forthcoming.
6 J. M. Poterba, S. F. Venti, and D. A. Wise, "Targeted Retirement Saving and the Net Worth of Elderly Americans," American Economic Review 84, 2 (May 1994).
7 Poterba, Venti, and I also used this approach in "401(k) Plans and Tax-Deferred Saving" and "Do 401(k) Contributions Crowd Out Other Personal Saving?" op. cit.
8 This is the approach that Venti and I followed in our analysis of "The Wealth of Cohorts: Retirement Saving and the Changing Assets of Older Americans," NBER Working Paper No. 4600, December 1993.
9 S. F. Venti and D. A. Wise, "RRSPs and Saving in Canada, "mimeo, 1995.
|Printer friendly Cite/link Email Feedback|
|Author:||Wise, David A.|
|Date:||Sep 22, 1995|
|Previous Article:||Behavioral economics.|
|Next Article:||Health care.|