Jonathan F. Pingle

The Recent Decline in the Labor Force Participation Rate and Its Implications for Potential Labor Supply

Figure 13 shows the participation rates, estimated trends, and fitted values of the model for teenagers (ages 16-19), young adults (20-24), prime-age adults (25-61), and older adults (62 and over), built up from the more detailed demographic groups included in the model. These aggregates highlight the relatively good fit of the model for the broad range of middle age groups and illustrate our concerns about the deviation of actual observations from the model predictions for the youngest and the oldest age groups. [FIGURE 13 OMITTED] For the 16-19 age group, the model appears to have captured the general trends and turning points in participation rates, although for teenage males there have been long stretches where the model prediction deviated from the data. Most recently, the model has expected participation rates to recover back toward their trends. In fact, actual rates have remained well below trend; although this finding represents a failure of the model, it does mitigate the concern that endpoint bias may be dragging down the estimated trend. Note, however, that for ages 20-24 the model exhibits much smaller errors, which suggests that the model residuals for the teenage groups have not tended to be carried through to older age categories as these cohorts age, and thus that the changes affecting recent cohorts of teens are age specific rather than cohort specific. For ages 25-61 the model fits well overall and is not surprised by the developments of the past few years. Notably, the model effectively captures the dramatic change in slope in the participation rate of prime-age women and the persistent downtrend in the participation rate of prime-age men. In contrast, the actual participation rates for both older men and older women have exceeded the model predictions in recent years. The model also missed actual outcomes fairly uniformly across the older age groups in some earlier periods (for example, 1985-86), suggesting that we may have omitted some salient influence on retirement decisions from the model. For example, the errors in the most recent few years could be related to sizable movements in asset valuations, although, as noted above, we did not find variables representing wealth to be significant in the model. Nevertheless, the large and growing size of this group suggests that these errors represent a substantial risk to our estimated trend. Model Projections and Alternative Simulations We can also use the model to project how the trend in labor force participation will develop in coming years. To do this we employ the following procedure. For birth-year cohorts age 16 or above in 2005, we hold the cohort effects constant at their last values and essentially age these cohorts along the last observed age profile. (47) For cohorts who had not yet entered the labor market by 2005 (so that we have no model estimate of their cohort effect), we assume that the cohort effect is constant at the average value of the last few cohorts and then age them along the last observed age profile. As figure 12 shows, the model projects that the trend in the aggregate labor force participation rate will fall further over the next ten years; indeed, the projected decline from 2005 to 2015 is more than 3 percentage points, which is comparable to the increase over the first ten years of our estimation period, when female participation was rising so rapidly. About 2 percentage points of this decline reflects the projected changes in the age distribution of the population associated with the aging of the baby-boom cohort, and the remainder is due to the models estimates of the trends in the age and cohort profiles over the next ten years. In constructing this projection of the trend, we assumed that the sizable recent model errors for teenagers and for the oldest age groups were not a manifestation of changes in the trend. However, an alternative approach would be to interpret the errors as suggestive of a recent change in the age profiles at those ages. …

Social Security's Delayed Retirement Credit and the Labor Supply of Older Men

This paper presents estimates of the impact of Social Securitys Delayed Retirement Credit on the employment rates of older men. The credit raises lifetime social security benefit payments for recipients who delay receiving benefits after age 65 and offers a rare and important test of whether labor supply incentives built in to the program can promote work at older ages. The results suggest that the increased incentives raised employment among workers over age 65. In addition, the recent increases in social securitys Normal Retirement Age also appear to be pushing up labor supply.