Syndicate content

Add new comment

Submitted by Berk Ozler on
Hi Dan,

Thanks. It is a puzzle, isn't it? I first thought it was thatchd vs. ALL, but that's not it. It cannot be baseline controls, because pure controls don't have them. So, while I have not figured it out yet, I can offer a tidbit from the paper. "In addition, the magnitudes and significance levels obtained in this within-village analysis are broadly similar to those found when comparing treatment to pure control households (Online Appendix Table 38), with three exceptions. First, the treatment effect on assets is larger when estimated across villages than within villages." So, it looks like they prefer the 104.56 to the zero...

I'll let you know if I figure it out. I'll also forward your question to Johannes to see what his answer is...