Tuesday, April 20, 2010

Econometrics and Economics

I went to two presentations at ANU recently by well-known American economists" Joshua Angrist and Frank Lichtenberg. Angrist presented work on the charter school in Lynn, MA. The presentation was clear and I thought quite impressive and convincing in showing a major impact of attending the school on students' performance, at least in mathematics. Lichtenberg showed the effect of diagnostic imaging and chemotherapy on cancer survival. I thought the results were fairly convincing that these procedures had large effects on reducing mortality.

I thought both these pieces of research were impressive but I wondered if they were economics. Angrist's work showed that charter schools increase education incomes but there was no policy evaluation or modeling of the decision to choose to apply to the charter school lottery etc. I asked Lichtenberg about what the results said about the cost effectiveness of research and development in the two areas. He hadn't done work on that. He did provide evidence of the monetary equivalent gain of the improved outcomes. I don't think it is a problem that this research isn't economics really. But the question then is: "If this is economics, is applying econometrics to climate data economics?" I think more economists would say no here and say that that kind of research doesn't belong in economics journals. Does the Angrist/Lichtenberg type of research belong in economics journals because it is about people (and uses econometrics)? So what is economics exactly?


  1. We have a whole department of health management and economics over in the school of public health, much of which isn't economics in the sense you mean. That said, health people face the same valuation issues environmental folks do when communicating research to policy makers - a lot of the policy analysis texts suggest pulling health and environmental stuff out of the aggregate to help communicate to decision makers who are not super comfortable with valuing either. Thus 'This policy will save 40 million dollars in direct expenditures and will additionally keep 20 people a year from getting a cancer that is hard to treat and has a 25% mortality rate. If we value avoiding a fatal cancer at 5 million each and a treatable cancer at two million, this is 40 plus 25 plus 30 million dollars.' So it sounds like that's what L has done, and then someone making a policy decision about funding could drop those numbers into a straight policy question with costs as well as benefits.

    For me, as long as the paper can stand up to the 'And why do we care, from a policy perspective?' question, I let it in the big tent.

    Now the theorists with empirically indistinguishable models but different policy implications are going to come after me.

  2. I don't care whether something is economics or not. As you know, I'm interdisciplinary. But I just wonder why some of the stuff I do isn't considered economics if this is. Lichtenberg had no cost effectiveness or anything in his paper. That's what I asked him about and he hadn't looked at that.

  3. The acadamic establishment in Australia is slow to recognize the importance of Environmental/Ecological Economics. I think in the coming years, it's importance will be increasingly felt in the world including Australia. Then, there will be more recoginition for people who actually do economics that is useful(as opposed to research in obscure economic theories) to the society.