I participated in a social program funded by the gov in the 90's as the data person. It was a combined early childhood education / social service program. I won't discuss what I thought of the outcomes or the cost / benefits, but I think there is a piece a lot of people miss.

There are consultants in DC that take all the data from a program and summarize it. They have a contract and aren't always the folks who have the assistance contract on the grant. People are sadly, people, and sometimes the summarizing folks have some pretty powerful incentives to put a spin on the numbers, drop certain grants, and well, other things. Life is incentive based after all.

When I look at the outcomes, I often wish there was a way to get some anonymized raw take because I am a bit mistrustful of third-party analysis particularly when they aren't up on simple things like geographical and economic differences. Plus the tendency to roll all the results in one set of numbers ignores how big the USA is and that San Francisco is a bit different than Salt Lake City or Rosebud.

PS: one thing, it is amazing how many children start life with hearing problems - fix that early and save the rest of us a lot of money. Plus, fatalogic has it right in my book, life is complicated, one axis of action doesn't really cut it.