liamc

Topkis's book Supermodularity and Complementarity is very well done, and I found it helpful for learning what questions to ask when trying to understand a model.

(i) If I knew I wanted to do empirical micro, I would consider most places in the top 15 to be plausible alternatives to Caltech. I don't know enough to say any more than that. (ii) I started in the Behavioral & Social Neuroscience program, but then switched to Social Science. That is a separate issue which I won't get into publicly. I have only ever attended here, but my impression is that Caltech has several significant advantages over other grad programs. There are many, very good faculty, relatively few grad students, very few undergrads, and lots of resources. This translates to nice offices, light TA duties, accessible faculty etc. This is obviously colored by my own experience, but I think people tend to overweight subfield fit in these decisions.

I do empirical micro at Caltech. I have been mostly on my own for subject-specific stuff, but I get a lot of helpful advice and questions about causality, behavior, econometrics, and political economy from the faculty here. Feel free to PM if you have more specific questions.

What was your math SAT score? In lieu of university math grades that may be the best predictor of your potential success in grad economics at this point.

Follow Larry Summers on twitter.

This is entirely speculation, but you might have better luck if you apply to places that do neuroeconomics. Even if that's not what you want to do, there will be economists at those places who have some knowledge of neuroscience and respect for your neuroscience professors. These include Caltech, Duke, George Mason, NYU, possibly UT Austin and CMU, probably others. Did you post a while ago? I seem to remember reading your story and perhaps corresponding with you.

OP, here's a suggestion: 1. Go to the best program you can get into, and work your *** off learning game theory and econometrics. 2. Work on applied theory or applied micro. Your background in critical theory will give you plenty of interesting research topics.

Do these exist?

Books I own which I believe many economics students could benefit from, which are relatively hard to get cheaply or electronically include (in no particular order): Ok, Real Analysis with Economic Applications Samuelson, Foundations of Economic Analysis Downs, An Economic Theory of Democracy Camerer, Behavioral Game Theory Savage, Foundations of Statistics Topkis, Supermodularity and Complementarity Roth and Sotomayor, Two-Sided Matching Piketty, Capital in the Twenty-First Century Posner, Economic Analysis of Law

I agree with moneyandcredit, except for this: It's not that hard to learn the literature on some topic on your own or by corresponding with experts at other places. What you really need are professors who can help you ask interesting questions and evaluate the ability of some theoretical or empirical method to answer them, and there are people like this at all good programs. Many people's interests change substantially once they start graduate school anyway. So, figuring out what topics you're interested in seems like a poor reason to delay graduation a year.

You should apply to Caltech! We have lots of good micro theorists and a few very good structural I/O people. Also take a look at BC and UCSD. Even if you end up someplace whence theorists don't place well, you should still give it a try for the first year or two. Trying and failing to be a theorist is a good way to learn as much theory as you need to do applied work (in my experience). EDIT: Just re-read the OP. I am not suggesting that Caltech (or any other school I mentioned) is outside the top 20 ;-)

Definitely transfer if you can. If you do transfer, the BA/MA might not be your best option: you might be better off double-majoring in math, or getting a bachelor's with one major and taking some classes in philosophy, psychology, history, etc, or graduating in 3 years and going straight to a PhD. As soon as you get accepted for transfer, you should find yourself an advisor at the state school and go talk with him/her about your plans.

If your college sponsors teams in the Mathematical Contest in Modeling, do that. It was the closest thing to theory research I did in undergrad, and also a lot of fun.

One of the best economics bloggers is a psychiatrist: The Last Psychiatrist

Caltech, definitely, for experimental and behavioral. Many economists (and a few researchers in other fields) now use reference dependent preferences, non-exponential discounting, probability weighting, and other ideas from behavioral economics in their work. Even outside the hotspots others have mentioned, there are likely many places where you will be able to learn enough behavioral economics to use it and contribute to it. That said, people who understand behavioral economics really deeply can teach it much better, so there is a real benefit to learning from somebody like Colin Camerer.