Causal Interference (Part II): Synthetic Control Method in 1990 German Reunification

Sophy Ramírez  | 03 de octubre de 2018  | Vistas: 276

Kevin Grier is an economist and professor of political economy. In this forum, he explains this statistic method; a system that helps to evaluate the effect of an intervention in comparative cases. In the first part of the seminar, he analyses the 1990 German Reunification with the Synthetic Control Method.

He starts by quoting another expert in the field, the creator of synthetic control, Alberto Abadie, and makes a contrast between his belifs and what Abadie says. Sharing how synthetic control works better in giving information about the GDP per capita of the country that an average of information.

Compares regression-based with synthetic control methods, explaining why the difference between them makes synthetic control more accurate in the treatment period.

Regression allows extrapolation outside the support of the data, not necessarily always going to happen, but there is nothing stopping it from happening and in this sort of newer world of causal inference, extrapolation it’s not our friend”.

He answers some questions of the participants of the forum; explaining more about synthetic control and the importance of showing the hole information that this method can bring and regression doesn’t.

Grier points out the placebo test, and how Abadie makes use of them in the pretended treatments. Where a situation is considered but they took it into another year in history and calculate how it would be if the treatment would happen in the year they choose, or the situation they choose.

This kinds of falsification tests are becoming more and more demanded by referees if you are doing like difference and difference or if you’re doing synthetic control or if you’re doing even like a regression”.

He concludes mentioning the types of economic resources that can be used in synthetic control and how to use it.

Join us watching this video, and learn more about economics and statistics.

Watch the first part of this seminar:

• Causal Interference (Part I): Synthetic Control Method in California Proposition 99