Professor Abadie invents new methods for economic research
Alberto Abadie refines the tools of economics — and gets some interesting results along the way.

"Teaching is complementary with research. When you need to explain material that you believe you understand very well, that is when the real test is. Teaching enriches my understanding of my own research and the research of others, and new research questions often originate in the process of explaining what we know already."

— Alberto Abadie, Professor of Economics and associate director of IDSS

Back in 1988, voters in California passed Proposition 99, a measure that raised cigarette taxes and placed other restrictions on cigarette sales. Has the law worked? To answer that, it would be good to examine an identical state where no such law was enacted, and compare it with California.

But as Californians know, there is no other place exactly like it. For social scientists, this is a problem. Rigorous studies often employ a test group and a control group, to see the effects of a change among otherwise similar sets of people. In this case, the control group is missing.

Alberto Abadie has a solution to this problem. It’s on display as he opens his laptop and shows some data to a visitor in his office in MIT’s Department of Economics. Abadie’s fix is what he calls the “synthetic control method,” which involves constructing an aggregation of demographically similar people from elsewhere, to serve as a kind of statistical control group — a composite non-California.

“You would like to observe what would have happened in California in the absence of the intervention,” Abadie says. “And of course this is something that you cannot observe. But other states in the U.S. have similar dynamics and factors. So you can use a combination of these other states to approximate what would have been tobacco consumption in California, in the absence of the intervention.”

Read full story at MIT News

Browse more SHASS stories about Social Innovation


Suggested links

Economics at MIT

Institute for Data, Systems, and Society

Alberto Abadie's webpage