To explore further the impact of these labor market variables, we use the
estimates for the full sample period (column 1) to produce counterfactual estimates
of what the poverty rate would have been in each year if our labor market
variables had been the only factors that had changed over time. Figure 5 shows this
prediction along with the actual poverty rate. The figure makes clear that we should
not be surprised that poverty rates failed to fall from 1967 through 2003. Rather, we
should be surprised that they did not increase by more!