Since 1981 (the earliest year of data), there is a clear and persistent increase in both the observed and selectivity-corrected wage differentials between disabled and nondisabled workers. Both differentials show that over this whole time period, nondisabled workers earned, on average, wages that were 23 percent higher than those of disabled workers. In addition, the corrected wage differential increased from 13 percent to 30 percent, indicating a deterioration of earnings of disabled workers relative to nondisabled workers over this time period. The decline in relative earnings is consistent with a downward trend identified by Haveman and Wolfe (1990) beginning in 1974. However, the growth in the wage differential appears to have been mitigated since 1992; the selectivity-corrected wage differential grew from 13 to 29 percent between 1981 and 1992, and has hovered around a mean of 28 percent since 1992, which was the first year of implementation of the ADA.
In addition, and particularly since 1996, individual selection into the labor market, or differences in selection between disabled and non-disabled workers, does not seem to be biasing the observed wage differentials between the two groups. The one exception to this might be the period 1992-1995. During this four-year span, the observed wage differential underrepresented the wage differential corrected for selection. The implication of this is that disabled workers were positively selecting into the labor market to a greater extent than nondisabled workers, driving the observed wage of disabled workers as a whole upward (and the wage differential downward). This is consistent with the labor force participation decline observed in Chapter 2, if the disabled labor force nonparticipants beginning around 1992 had systematically lower earnings potential than the disabled persons who stayed in the labor market. This would likely be the case as a result of the flow of Aid to Families with Dependent Children (AFDC) recipients to SSI (Lewin Group 1999). This result is not consistent, however, with the conjecture that the disabled workers entering the labor force post-ADA were those with the most limiting disabilities (Chirikos 1991). Whatever might have been making the observed and selectivity-corrected wage differentials diverge in the mid 1990s seems to have disappeared, since the two series have basically followed identical paths since 1996. The implication of this is that since 1996, selection into the labor market has had essentially the same impact on wages for disabled and nondisabled workers; self-selection explains none of the remaining wage differentials since that time.
Was this article helpful?