I just hope they run this as a proper research study. The Millennium Villages weren't set up with controls and so there's very little research value to be gotten from all that effort.

Things to note:

1. Who is in and who is out is fluid. There are no well-defined communities and some people will get into the BI cohort purposefully (i.e. staying with a friend who is in the BI district or claiming to be family).

2. People move ... much more than one would think.

3. People are individuals but really, families are the core unit of society. Don't just measure individual outcomes, measure family outcomes.