The Life Cycle Model, Replacement Rates, and Retirement Income Adequacy

The key insight of the life cycle model in economics is that a household’s consumption at any given time is determined not so much by its current income as by the total income available to the household over its lifetime. A replacement rate can be a useful tool in approximating the life cycle model’s predictions for how households wish to prepare for retirement. The Social Security Administration’s Office of the Chief Actuary (SSA OACT) publishes two different calculations of retirement income replacement rates, each of which find that Social Security benefits replace about 40 percent of a typical retiree’s pre-retirement earnings. Some interpret these figures as indicating that Social Security benefits are insufficiently generous and that U.S. households’ total retirement saving is inadequate. But SSA OACT’s two methods for calculating replacement rates each violate the life cycle model in a meaningful way. SSA OACT’s career-average earnings replacement rates, in which lifetime earnings are first indexed upward by the rate of economy wide wage growth, exaggerates by roughly one-fifth the real value of earnings available to a household for consumption over its lifetime. This overstatement lowers household’s measured ability to replace their pre-retirement earnings. SSA OACT’s final-earnings replacement rates effectively compare Social Security retirement benefits pre-retirement earnings only in the years in which the individual worked, ignoring the life cycle model’s prediction that household consumption is a function of long-term average earnings including years in which a household member was not employed. A variation of that method, calculated by the Congressional Budget Office based upon recommendations from the Social Security Advisory Board’s 2015 Technical Panel on Assumptions and Methods, take the SSA OACT approach further by comparing Social Security benefits to an average of above-average earnings years in the period immediately preceding retirement. A replacement rate calculation more consistent with the life cycle model would compare retirement income to an average of real earnings calculated over a significant number of years. Such an approach would find substantially higher replacement rates for the typical retiree. It is important both for Social Security policy and the analysis of overall retirement saving adequacy that replacement rate calculations build on the insights of the life cycle model that guides most economic analysis of retirement saving.