ÿÈÕ³Ô¹Ï

Labor force participation (LFP) rates are changing - and, at least for some groups, changing dramatically. These trends have important societal implications. For the most part, they indicate longer stays in the labor force and later retirement. Such trends may allow for the accumulation of greater wealth by the time an individual reaches retirement age, allowing for higher levels of income replacement and less pressure for increases in Social Security benefits. These trends may also reduce the financial stress borne by the working generation in supporting the retiring generation. Given the importance of these and other potential effects of changing LFP rates, the ability to predict the future of such rates could be helpful to decision-makers. While LFP trends to date have been well documented, there have been few projections of trend lines that have instilled confidence. The purpose of this paper is to report on new LFP projections based on novel but rigorous methods taking advantage of panel data from the Health and Retirement Study. We begin with LFP rates based on the Current Population Survey, then move to two-year labor force retention rates and retirement hazards from HRS data. We then introduce the prediction tool, the HRS respondent's subjective probability of retention in the labor force at age 62 (or 65). We show the trends in this measure over the course of the past two decades and demonstrate its favorable comparability with actual retention behavior. We present individual - and population - level predictions of LFP rates based on subjective probabilities. We conclude with some data about possible cause of the increased LFP rates.