This morning I had the chance to present preliminary findings from my Total Sanitation Campaign project at ISI Delhi. It was much fun and I appreciate the questions and comments I received. Thank you all!
For those of you know don’t know the project, I seem to have found that the TSC has, on average, importantly reduced post-neonatal mortality and increased height among rural Indian children, and that it has done so quite inexpensively as health programs go. For much of the paper I use the program’s administrative records on latrine construction. At one point in the argument, I consider how much it would change my conclusions if these administrative records overstate the number of latrines built and in use, and argue that any such inflation is almost surely not driving my results.
There was a gentleman in a tie who sat in the back of the room who – as near as I could tell – took serious issue with my entertaining the possibility that the government records might not be perfectly accurate. He seemed to be an official of some sort, judging from his vigorous defense of the official data and from the fact that he was wearing a tie. But I would think that anybody associated with the program would be more concerned with my bottom line that the program is helping than with this robustness check. And, to its great credit, the TSC itself is quite open on their webpage about potential data problems, as are Kumar Alok’s memoirs of setting up the program (for example, the story of entering data by hand into an Excel spreadsheet before it made it online).
Are you out there, friend? I’m baffled! I was hoping to try to puzzle the objection out after the seminar, and I assumed that you would also want to chat, after talking about it so much during the presentation. But I guess you had to leave immediately? If this was you, please send an email!