Starting in 1994, soon before the 1996 welfare reform bill (PRWORA) the state of Florida explored what was roughtly a pilot version the “Family Transition Program” (FTP). I recall reading at the time that this was the only reformed welfare program with a hard time limit on benefits. Importantly, Florida officials decided to conduct a randomized controlled trial
Bill Gardner reports the research of Peter Muennig, Zohn Rosen and Elizabeth Ty Wilde
From 1994 to 1999, Florida randomly selected a group of welfare recipients into either the Family Transition Program (FTP), a program that included many features of the later PRWORA, or the then-standard Aid for Families with Dependent Children welfare program (AFDC). Muennig and his colleagues looked at Social Security Records for 3,224 participants in the experiment to establish whether participants had died in the 18 years following the randomization.
What they found was that
“participants in the experimental group had a 16 percent higher mortality rate than members of the control group (hazard ratio: 1.16; 95% confidence interval: 1.14, 1.19; p < 0.01). This amounts to nine months of life expectancy lost between the ages of thirty and seventy for people in FTP.”This was a surprise: Muennig et al. had expected that the increased employment associated with FTP would lead to improved health. Instead, FTP harmed the participants.
This is analysis of a randomized controlled trial.
The p value of “<0.01″ is putting it very mildly. It is conventional and wise to refrain from calculating p-values for statistics way out in the tails — they are shockingly low and might mislead those who don’t fully understand that they are calculated using assumptions (everyone). The confidence interval indicates a z-score on the order of 6. The p-value if foolishly calculated out would be on the order of 2 in a billion. The result is not due to random sampling error.
Note that the authors were surprised by their result. This adds to its credibility, since it is unlikely that they cherry picked to get a result which they found strange and puzzling.
I have an objection to Gardner’s post is the question mark in the title ”
Now one never knows anything for sure, but a question mark is not usually used when discussing the results of a randomized controlled trial with an overwhelmingly statistically significant result.
Also, oddly, he doesn’t conclude that it might be a good idea to reverse welfare reform and stop killing people. Instead he concludes ” Every social policy should be evaluated with an open mind using the strongest possible empirical design.” OK do it. How about a policy proposal ? TANF (reformed welfare) costs less than AFDC (old welfare). With a point estimate of 9 months of life lost per participant due to replacing AFDC with TANF one can calculate the lives lost per dollar saved. This would not be the sort of number one is used to (we are talking thousands of dollars per life) . Or ok go to the low end of the 95% iterval so the months lost would be on the order of 6 per participant. If you think a life is worth $50,000 or more, you should have no problem at all decided that we should reverse welfare reform now.
I think the scientifically solid result has been almost completely ignored, because it is politically incorrect. Criticism of welfare reform is barely allowed. Noting the overwhelming evidence that it kills would count as a 4 pinocchio pants on fire fact. Few serious people would dare admitting that they find statistical analysis of randomized controlled trials convincing.
I googled for other discussion of the study
I googled (welfare reform kills florida) and get two links to Gardner then a bunch of links about how wonderful welfare reform is and a few claiming that Obama wants to kill it.
Same for (welfare reform kill florida)
I think this is the google pattern for an academic study which did not influence the general debate. I think the problem is that the result is politically unacceptable. Re-reforming welfare is not on the agenda.