Effects of Pensions on Saving: Analysis with Data from the Health and Retirement Study
The wealth of the median households of those aged 50 to 60 in 1992 would finance an annuity equal to 79 percent of their final year's earnings, while still guaranteeing a surviving spouse two thirds of the basic benefit.
Some economists suggest that pensions do not add to "retirement wealth"-- which includes the value of private pensions, Social Security, homes, and other assets -- because, for a variety of reasons, higher pension values are offset by reducing saving held in other forms. However, according to a recent study by NBER Research Associate Alan Gustman and co-author Thomas Steinmeier, "the overall effect of pensions is to increase total wealth, probably by considerably more than half the value of the pension."
In Effects of Pensions on Savings: Analysis with Data from the Health and Retirement Study (NBER Working Paper No. 6681) , the authors explain that pensions cover about two thirds of the families approaching retirement, mainly those in middle and upper income brackets, and on average account for one quarter of retirement wealth. Much of this wealth is new saving, they write. Moreover, although there has been concern that those leaving "pension jobs" lose a substantial part of their pensions, this study finds that such pension losses are not severe, amounting to 12 percent or less of the total value of pensions held. The study also explores the other legs of the retirement stool: Social Security and saving held in the form of other wealth. It finds that each household with a member nearing retirement has on average about a half million dollars in retirement wealth, with Social Security accounting for about a quarter and other assets accounting for about half of retirement wealth.
Gustman and Steinmeier reckon that this retirement wealth is adequate to provide a reasonable standard of living in retirement for most households in their sample, with a member born in 1931-41. The wealth of the median households of those aged 50 to 60 in 1992 would finance an annuity equal to 79 percent of their final year's earnings, while still guaranteeing a surviving spouse two thirds of the basic benefit. Including the saving that will be undertaken and the added value of the pension that these households will enjoy in the seven years until retirement, the median household will have enough wealth to replace over 60 percent of final earnings (in real terms) throughout their retirement, again with two thirds of the benefit available for a surviving spouse.
Although some households are inadequately prepared for retirement, Gustman and Steinmeier suggest that we are far from the retirement crisis often pictured in the press. When pensions and Social Security are counted along with other wealth, total retirement wealth represents about 40 percent of lifetime earnings throughout almost the entire lifetime earnings distribution. Of those without pensions, a substantial subgroup holds business or property wealth. For some of the poorer households, Social Security replaces a substantial fraction of their earnings. However, about a quarter of households have no pension or business wealth and only limited Social Security, and thus are not well prepared for retirement.
Households will be in worse shape if Social Security benefits are sharply reduced, the authors note. Social Security, one leg of the retirement stool, is facing severe financial problems which are already apparent for this generation on the verge of retirement. Their Social Security payments will be about 10 percent less than what they would have gotten if the money deducted from their pay had been invested in U.S. Treasury securities. This negative return has come earlier than many analysts expected. All told, however, "[T]he data do not support the most dire views of retirement prospects, at least not for those now on the verge of retirement. Pensions are doing a better job providing support for households than they are given credit for."
These estimates of pension wealth are based on pension plan descriptions obtained from the employers of respondents to the Health and Retirement Study, a detailed nationally representative survey of some 7,600 families with at least one member born between 1931 and 1941. Lifetime earnings and Social Security benefits for survey respondents are estimated from earnings records obtained from the Social Security Administration.
-- David R. Francis