My guest on this episode is Kevin B. Grier of the University of Oklahoma.
Our topic for today is a paper Kevin wrote on the economic consequences of Hugo Chavez along with coauthor Norman Maynard.
I had Francisco Toro on the show last year to discuss Venezuela's economic history, so you can listen to that episode if you want a refresher on Chavez. For this episode, our main topic is the empirical method Kevin used to quantify Chavez' effect on Venezuela: synthetic control.
Synthetic control is a relatively new empirical technique. It grew out of an older technique called difference in differences (or diff-in-diff). Diff-in-diff is simple and intuitive: Given two statistics with parallel trends, we can compare their changes before and after some intervention affecting only one of them to see the effect of the intervention. So for instance, if you wanted to know the effect of Seattle's minimum wage increase, you could compare the employment trend among low-skilled workers in Seattle to the same trend in Portland. Then assuming Seattle and Portland would have had similar trends if not for the minimum wage hike, we say the difference between the employment growth in the two cities is attributable to the minimum wage hike.
But what if Seattle and Portland don't have similar trends? What if there's no labour market similar enough to Seattle's to provide a valid comparison? That's where synthetic control comes in. Seattle might not be like Portland, but it might be like a weighted average of Portland, San Francisco, and several counties just outside Seattle. We could construct this weighted average and call it a synthetic Seattle; it is designed to mimic the dynamics of Seattle's labour market before the minimum wage hike. Then if the synthetic Seattle deviates from the real Seattle after the wage hike, we can attribute that difference to the hike.
This is what Kevin has done to study the impact of Hugo Chavez on Venezuela. Listen to the episode to find out his results!