401(k) Plans, Lifetime Earnings, and Wealth at Retirement
In 2000, per capita assets in these personal retirement plans for people retiring at age 65 averaged about $29,700. By 2020, they are projected to average nearly three times that amount in year-2000 dollars, and by 2040, $269,000 year-2000 dollars.
In 1980, 92 percent of contributions to retirement saving plans in the United States were directed to traditional employer-controlled defined benefit pension plans. By 2000, 87 percent of such contributions went into to plans controlled by employees - either defined contribution plans or one of the many tax-deferred accounts that have been created in the last 30 years, including 401(k), 403(b), and section 457 plans, as well as Individual Retirement Accounts.
In 2000, per capita assets in these personal retirement plans for people retiring at age 65 averaged about $29,700. By 2020, they are projected to average nearly three times that amount in year-2000 dollars, and by 2040, $269,000 in year-2000 dollars. Most of the increase in the next two decades is the result of an ongoing rise in the percentage of their career for which retirement-age workers have been eligible for 401(k) plans, rather than the further diffusion of 401(k) plans.
In short, over the next 35 years the individual retirement savings of the average 65-year-old will rise sharply. Self-directed retirement assets will become more important than defined-benefit pension plans by 2010, even though traditional defined benefit pensions remain the most important source of retirement assets for federal, state, and local employees.
This analysis is based on projections detailed in two papers: Rise of 401(K) Plans, Lifetime Earnings, and Wealth at Retirement (NBER Working Paper No. 13091) and New Estimates of the Future Path of 401(K) Assets (NBER Working Paper No. 13083), by James Poterba, Steven Venti, and David Wise. The authors identify a number of factors that will be key determinants of future retirement assets, including the diffusion rate of 401(k) plans to employees who are not currently covered by such plans and the rates of return on stocks and bonds. The benchmark case in both studies assumes that past rates of 401(k) diffusion slow prospectively, that investments in corporate stock generate an average return 3 percentage points below their historical average return, that bonds generate their historical returns, and that retirement savers direct 40 percent of their contributions to bonds and 60 percent to stocks
The authors use the Survey of Income and Program Participation (SIPP) to estimate the eligibility for and participation in 401(k) plans for individuals in detailed age-earnings categories. The SIPP data suggest that in 1984, only 5.8 percent of 44-years-old had 401(k)-type accounts. Nineteen years later, in 2003, 44-year-olds had a participation rate of 44.3 percent.
In 1984, 55.5 percent of eligible families with heads aged 30-34 participated in a 401(k)-type plan. By 2003, the participation rate was 80.7 percent. Participation rates vary as a function of earnings. On average, 17.1 percent of those in the lowest earnings decile in 2003 participated in individual retirement savings plans; for the highest earnings decile, the average was 67.2 percent.
The authors combine the SIPP data with information from the Health and Retirement Study (HRS), which includes Social Security earnings records and makes it possible to estimate 401(k)-type contributions by age and earnings cohorts. The authors assume that the combined employer and employee contribution rate to these plans is 10 percent of earnings. They also recognize the possibility that individuals withdraw assets from the 401(k) system with a lump sum distribution when they change jobs. Both the contribution rate and the rate of "leakage" from the 401(k) system through such withdrawals play an important role in determining future accumulations in self-directed retirement plans.
-- Linda Gorman
