I'm hiring RAs to work on data analysis for projects on education, infrastructure, and political economy in Latin America. Spanish/Portuguese language helpful; remote work possible. Please share, and/or apply!

Links in next tweet. #econra #econ_ra #EconSky

Good job econ twitter exodus getting us all in on the ground floor over here

We often urge on politicians the "courage" to champion reforms we favor. But they're smart & rational & want to keep their jobs. My paper shows what happens when a pol steps off the equilibrium path, and why so few do. 13/n

In sum! Voters (at least in 🇱🇷) ARE attuned to service quality: if you improve things enough, you can win them over. But improving service quality often means changing the status quo for lots of public servants, and they also have a voice. Alienate them at your peril. 12/n

The program also dramatically reduced teachers' political activity, including polling booth staffing & campaign participation. This was electorally consequential: negative incumbent vote effects were concentrated in places where the program disengaged teachers most. 11/n

(There was plenty for teachers to dislike. The prior paper showed that the risk of dismissal/transfer went way up at treated schools. Teachers also had to work much longer hours. Some likely opposed involving private mgmt in public ed as a matter of principle.) 10/n

There was a lot of variation in how well the policy worked. In places where it increased test scores a lot, there WAS a positive electoral effect. So voters' rewards (and punishments) were commensurate to the policy's effectiveness. 7/n

The result surprised me: voters near the treated schools were SIGNIFICANTLY LESS LIKELY to vote for the pres. candidate from the incumbent party which had implemented the policy.

So much for voters rewarding service quality, right??

Wrong. I'm an economist. There is nuance. 6/n

The program generated a lot of controversy and press coverage, and was implemented in an election year, giving it unusual public visibility. I matched polling-station-level electoral data to nearby treatment and control schools to test how the treatment affected vote totals. 5/n

But are leaders rewarded when services improve? Much of the (great!) accountability lit. focuses on corruption and other malfeasance. Precious little empirical evidence seeks to measure how voters respond to *improvements* in service quality. 3/n

There's a widespread belief that voters want better public services - including schools - and the persistence of bad ones in poor democracies simply shows that the electoral system is too compromised to deliver accountability. 2/n

I realize you're rolling your eyes, but it's a striking graph and I actually don't know the real answer -- is there a consensus on why Kenyan GDP took off in 2006 relative to the rest of East Africa? especially since GDP growth rates looked kind of comparable? maybe I'm missing something obvious