Friday, July 22, 2011

Questioning Assumptions

Via HP's delicious links, an interesting NY Times article from 2007 on the plight of heterodox economists, and the reliance of orthodoxy on mathematical models and plausible sounding theories as opposed to the more scientific approach of confirmation with test data. Here's an excerpt:

The experience of Mr. Card’s graduate students suggests how the process can work. Mr. Card is by no means on the fringe, but he said his research on the minimum wage in New Jersey “caused a huge amount of trouble.” He and Alan B. Krueger, an economist at Princeton, found that contrary to what free-market theory predicts, employment actually rose after an increase in the minimum wage.

When Mr. Card’s graduate students went on job interviews, he said other economists would ask questions like “What’s wrong with your adviser? Has he started drinking?”

This is why Mr. Blinder said he advises graduate students “not to do what I do” when it comes to challenging the standard model.

Criticizing the approach that currently dominates the field, Mr. Blinder said economists must look more closely at the real world instead of modeling it in the lab. “Economics is insufficiently scientific,” he said. “Mathematics may be useful, but mathematics is not scientific. It doesn’t generate refutable hypotheses.”


HispanicPundit said...

Every academic group, environmental science included, has boundaries...issues that have been settled for so long with so much evidence, for example, that questioning them or overturning them requires SOLID, STRONG evidence. And if you don't provide that level of evidence, you get peer pressure push back.

Think about it this way: what would happen if Freeman Dyson published an article questioning the "humans are causing global warming" theory? Lets say his evidence was "good" but not "great". Do you think environmental scientists would come to him and say, "your paper sounds good, tell me more about it". Of course not. They would generally shun him. Unless of course, his paper was GREAT - with something really novel - he would be dealt with the same cultural push.

Especially when the issue is political. One environmental scientist questions global warming, and he gets ALOT of media attention. Nevermind the fact that the majority of environmental scientists are settled on the matter. So when its political like this, and you know the media is going to run with it, you have to be even more careful. All academic groups act like this, economics is no different.

It makes sense too. It's not like the public is going to take the Card study, the David Neumark study, the back and forth discussions and decide on the merits who is right. No way. The public already has a preconceived idea - the minimum wage is good, global warming is not caused by humans - and they are just looking for confirmation. If atleast one economist believes this, if atleast one environmental science believes this, the science is not good enough and I'm probably right - that will be their train of thought. This is why it annoys the academic group as a whole.

Did you read Greg Mankiw's rebuttal to the NY Times claims here? He is making roughly the same charge (especially point #4). It makes sense.

So - was Card's study even 'good'? I would say no. Have you dug into the study Jon? I really suggest you do. First of all, it wasn't like he got some government data and manipulated, most of his data was simply from surveys. I'm not sure how familiar you are with the social sciences, but its near universal that surveys carry the highest margin of error. And to use surveys to make conclusions about minor changes in the survey set, now that's really bad.

Okay - I am a rightwinger and I have a predisposition to dislike the Card study. Fine. How about I give you what I think is my BEST argument against the Card study? Card himself.

David Card later published another study that contradicted his earlier study on the minimum-wage. My favorite economist, along with others, called him out on it. He has as of yet refused to respond. But the contradiction is clear. Take a look for yourself. I'd be interested in your thoughts on it. My thoughts on the minimum wage can be found here.

HispanicPundit said...

Oh and, another delicious set of mine that you might be interested and, and is really a section just for you, are these sections, here, here and here.

Paul said...

Aha! I knew you were using some type of bookmarking scheme/tool to track the various threads. :-)

I asked about this a couple of weeks back.

Paul said...

HP -

On the "contradiction" if you let me play devil's advocate for a moment.

What if the issue is not there is a contradiction but rather that the model being/diagrams being used by Caplan are wrong.... Hmmm... "wrong" is not the right word. I don't know how to articulate it but maybe by some odd analogy. Physics - physics of big things are knowable using Relativistic (Einstein) phycis. Physics at the quantum level operates at (what "appears" to be) different rules.

Economic models may suggest X but if the empirical data show not-X (or only partially X... whatever) then unless the empirical data is wrong (very possible, but for sake of discussion let me assume not wrong) then it is the models being used that need to be refined/adjusted.

HispanicPundit said...


Did you? I never saw you ask me. Yeah, I don't know how people did it before delicious. It saves me a ton of work.

Regarding the models: keep in mind that these are not complicated mathematical models - this is supply and demand. Algebra basically. This is the same model that Card appeals to in both studies.

Also, contrary to models you see in higher levels of economics and environmental science, these 'models' assume no real variables - they are just mathematical ways of communicating.

Or to say it another way, Caplan could have just said what the graph show. Either way, Card has to atleast explain a way, through words, or graphs, how to reconcile the two studies. So far he hasn't. And I dont think he can. THey are mutually exclusive (though there is the possibility that BOTH studies are wrong).

Sheldon said...

Michael Perelman's the Invisible Handcuffs of Capitalism provides other examples confirming that testimony.