The purpose of this paper is to shed new light on the welfare debate i
n two ways: (a) to provide new evidence on the effectiveness of long-t
erm workfare programs; and (b) to examine in what contexts the program
s were most effective. The study examines welfare trends between 1981
and 1990 for counties in Ohio. A quasiexperimental design is used to c
ompare 26 counties that experimented with workfare with Ohio's remaini
ng counties. The findings indicate that workfare lowered caseloads sig
nificantly, particularly for the two-parent (AFDC-U)population. The fi
ndings also indicate that the effectiveness of the program increased i
n counties with lower unemployment rates and smaller urban populations
. The programs appeared to be more effective in more conservative coun
ties, but these findings are mixed.