NCPA - National Center for Policy Analysis

Debating the Results of a Welfare Reform Study

February 20, 2002

Beginning in 1996, the Manpower Demonstration Research Corporation conducted a study to compare how a random sample of Connecticut welfare recipients responded to welfare reform. Some 2,400 families were selected as a control group -- continuing to receive aid under the old welfare rules. Their record of attaining jobs was then compared to another group who entered the state's "Jobs First" program.

The results showed that those in the control group were nearly as likely to find jobs as those assigned to the "Jobs First" program. Analysts are still debating how to interpret the data, since the differences between the two groups were modest.

  • Five percent more Jobs First participants took jobs at some point -- and 9 percent fewer were on welfare rolls after four years.
  • Among the control group -- which were faced with none of the work requirements, time limits or extra subsidies used to push or lure their counterparts off public aid -- 81 percent left welfare for jobs.
  • At the end of the research -- just before the economy began to sour -- 72 percent of the control group were still off the rolls.
  • The cost of subsidies and services for those in Jobs First was $4,100 per person more than aid to their counterparts.

One small but statistically significant difference was noted. Some 2.6 percent of Jobs First participants became homeless in the study's third year. That's a percentage point more than in the control group.

State officials claim the study results confirmed that Connecticut's welfare reform program succeeded in replacing a culture of welfare with a culture of employment.

But some of those who conducted the study contend that people who moved off welfare would have done so without the reforms.

Source: Nina Bernstein, "In Control Group, a Majority Quit Welfare Without Reform," New York Times, February 20, 2002.


Browse more articles on Tax and Spending Issues