I survived random assignment
I attended an amazing conference on employment and criminal records yesterday and could blog at length about several of the papers. My favorite moment, however, came in a discussion with Mindy Tarlow, the CEO of CEO -- New York's Center for Employment Opportunities.Ms. Tarlow has been providing employment services for recent prison releasees at CEO for about 15 years. As the Times reported last year, They do job readiness training, transitional employment, job placement, and retention. The transitional jobs are especially important for new releasees -- they can be on the job site just days after incarceration and they get a paycheck at the end of each day to address their pressing financial needs.
I've long appreciated the program's approach, but hadn't seen what I'd consider really solid evidence on its effectiveness until yesterday. CEO now has 2 years of data from an independent randomized evaluation by MDRC -- and they're showing significant treatment-control differences in new confictions and incarceration.
When I remarked that it takes real guts for an agency to voluntarily subject itself to rigorous assessment, Mindy said, "I know. MDRC gave me this button that says I survived random assignment."
posted by christopher uggen @ 2:00 PM
1 comments
1 Comments:
I'm impressed with agencies who volunteer also. I'm inclined to think that rigorous assessment should be built into any funding for a pilot and ongoing measures be built into contracts with community agencies. Money is so tight these days that we can't really afford to fund programs on a handshake anymore.
The trick is reaching agreement on the outcome measures and experimental design up front. Quality of life for the client (in this case a job) and community where they live (recidivism) seems to be the most common. Sometimes you get one but not the other. Or you get halfway through a project and a partner becomes reluctant to stick with the plan because they have a client in the control group that has needs.
Interesting.
Post a Comment
<< Home