What works? In particular, which social programs work? One way to find out might be to copy some of the procedures for clinical trials for drugs. For example, your city might be considering funding a program for homeless that places them in apartments and gives them rent and subsidies for 6 months while they get job training. One thing to do would be to sign up two groups to track, those that get the help and those that do not.
There are similar trials going on in the US and in developing countries. For example, one might wonder whether having a reliable source for vaccinations with small non-monetary rewards would help with the low rate of compliance in a country. One thing might be to divide villages into those that get the resource (and/or the rewards) and those that do not, and then track the vaccination of children. Here is a web site with descriptions of some of the trials in developing countries. The lead investigator on that site, Ester Duflo of MIT, was profiled recently in the New Yorker; the full article is ‘gated’ for subscribers only, but the lengthy abstract will given you an idea if you want to pursue it through libraries.
In fact, in NY City such a trial is going on to test a program called Homebase,which was begun in 2004, and offers job training, counseling services and emergency money to help people stay in their homes. Now when you apply to Homebase you have to agree to be part of a lottery that will determine whether or not you are part of the program.
The testing of Homebase has become very controversial. Those against it maintain that one should not treat people, particularly very disadvantaged people, like “lab rats.” The pro advocates say that money is limited and good data is needed to make the decisions about which programs to implement.
Who is right? What do you think?
Let me make two observations:
1. The trial in NY City is unlike the drug trial in one significant respect: you know from the beginning whether or not you are getting the real thing. Further, those who aren’t know that others are, and that it is just random bad luck that is preventing them from getting it. In fact, both groups have to accommodate the fact that random luck has made the difference in their getting the chance. Duflo mentions a comparable problem with a project in India; once people know they are being tested, the motivations for behavior can increase.
2. Proponents of the program in NYC argue that not everyone could get the resources anyway and they are just tracking those who don’t. But this may not be true, since the grounds for deciding who does and who does not get into the program appear to be different. It didn’t used to be done by lottery. In fact, the story of one women going through the lottery and losing is heart-breaking. Does that signify a moral difference that may not be present in all the cases?
Well, I don’t have all the answers. What do you think?