JRav

In my own opinion, it's not necessary to get a book on optimization. I never took a course on optimization theory. Almost all of the relevant techniques (unconstrained and constrained optimization in R, Kuhn-Tucker, convex analysis, discrete-time dynamic programming) will be covered in math camp and or macro. The other stuff can probably be picked up as necessary. Knowing Lagrange multipliers is extremely helpful but that's calc 3/multivariable calculus material.

Peter DeMarzo actually has a paper in which he uses algebraic geometry to establish some equilibrium result: Computing equilibria of GEI by relocalization on a Grassmann manifold

If you're interested at all in Ramanujan, read "The Man Who Knew Infinity;" great book. And yes, he was a genius and a freak of nature.

Unless I missed it, I'm surprised no one has mentioned one potential pitfall of staying in the same place: by staying in the same department and with the same faculty for 9+ years will likely lead to some degree of "inbreeding." It's good to go to another department and be exposed to the ideas and work of other people. As an example, what I was taught in th first-year grad macro sequence at my undergrad and what was taught in the first year at Princeton were very different; the macro papers coming out of the two departments are also very different, both in terms of content and model-building. I was advised not to apply to my undergrad's PhD program (top 25ish, top 15 undergrad). I didn't. Had I stayed there, I wouldn't have learned nearly as much. The above, and you get to meet more economists.

You might also want to consider the books by Hogg and Craig or Casella and Berger. I don't know how much they're used in Math Stat classes but they're fairly common in the prob/stat half of many first-year econometrics sequences.

Here's a thread from last year with the funding packages from various econ departments and business schools (most probably went up a small amount).

How advanced is the probability class? Most first-year graduate econometrics classes review the relevant mathematical statistics, and it isn't that hard. If that's a measure-theoretic probability course (which it doesn't look to be), then take that. Otherwise, I don't think it matters that much (though, I might lean towards math stat).

There are several people in my cohort who never even took real analysis, let alone got an A.

To the extent that it's covered in Sargent's book. Here's the model, here's the statement, here's the proof. I covered it a bit as an undergrad, but as part of the first-year graduate macro course, not in intermediate macro.

For us, macro sort of depends on who's teaching. The last few years Sargent taught the first half so we did search models (McCall, Neal,...), equilibrium with complete markets, Kalman filter and some other time series stuff, self-insurance/equilibrium with incomplete markets (Aiyagari model), OLG, Ricardian equivalence, fiscal policy in a growth model, Ramsey problem, and a tiny bit of asset pricing--basically chapters 2-6,8-11, 14-15, 17-19 in Sargent's book. Second half also depends on instructor--this year we did RCE, Hopenyahn model, Aiyagari model, Lucas-Prescott, Mortensen-Pissarides, Williamson model, Bernanke-Gertler, Carlstrom-Fuerst model, Moll model, recursive contracts with one-sided commitment and asymmetric information. Metrics sort of depends. I think the first semester is standard (review of prob/stat, then first 3/4 chapters of Hayashi). Second half depends on the professor. This year we had a nice balance--half time series, half other stuff (extremum- and M-estimators, discrete choice models, censoring and truncation models, nonparametrics, treatment effects,nonlinear panel data, a tiny bit of semiparametrics). Some people also take or sit in on classes in probability theory and/or stochastic calculus.

Do the second option. Differential equations won't be all that useful: some differential equations may show up in macro but it's typically limited to constructing phase diagrams, which the professor will likely go over anyway. Between the two first-year macro sequences I've taken (one as an undergrad, one as a grad student), not once have I used any techniques from the course I took in differential equations.

Sit through tone week of a Ph.D. micro course and believe me, you'll see the importance of showing that Cauchy sequences in compact spaces converge. If you're interested in development, then no, having an extensive background in measure theory will not be particularly useful but having a solid grasp of probability is pretty importance for the econometrics sequence, and you'll need metrics for applied work. I don't know what good psychology and sociology courses will do you, particularly in the first year of a Ph.D. program. The first year is all theory and field courses should be largely self-contained. Take them if you want to, just don't expect them to help you much. (Disclosure: I have no interest in applied micro or applied anything so I don't have any experience in a development or labor course.)

Kehoe started the process of revamping the macro group and it's continuing. Speaking to a few upper-level students, they said that Rogerson and Aguiar have made the student-macro lunch much stronger. Also, Kiyotaki has placed quite well recently (NYU last year, and I think a Wharton placement a few years before) and Brunnermeier has consistently placed well. The program as a whole is being greatly improved as well.

University of Chicago Department of Economics | Frequently Asked Questions. See the middle of the FAQ.

Until you know where you're going and you've been told by the professors what book they'll use, don't bother. And it might be a good idea to relax a little bit since you'll be working like crazy soon enough, Different schools use different books, but Stokey/Lucas/Prescott should be invaluable wherever you end up, and is worth perusing.

Check the stickies. This has been asked and answer many times. There's an entire thread devoted to questions, which should be stickied.

As a Princeton student, I would of course vouch for Princeton (if you get in). The labor group here is amazing and macro just keeps getting better (and two of the three profs you mentioned in your original post are at Princeton, and Sargent has taught here the last two years). You also have finance people that touch on macro and metrics. I've never been to New Haven but Princeton is a wonderful place to go to school. Yes it's isolated but you're an hour away from NY and Philly by train and the town itself is very nice.

Last year Princeton and Chicago were both April 1 and Michigan was in mid-March (16th or 17th or whatever that Friday was).

For the OP, last year Princeton required fall grades to be uploaded.

Send them an e-mail. I e-mailed five or six departments and asked if I could e-mail them an unofficial transcript. I think all but one responded.

I made sure to mention at least two professors for every school that I applied to, just to show that the department would be a fit for me. In one case, it actually made for a more personalized recruiting push once I was admitted.

Last year, one of my recommenders didn't submit a letter for Chicago until February 2. I was ultimately waitlisted but I don't think it mattered that much. I was told by multiple faculty members that adcoms don't actually start reading letters and application files until the end of January anyway so it probably won't kill you if it's in a little late. Probably.

10-15k before taxes seems a bit low. Anyway, a lot of schools discourage students from working because it eats into time that could be spent on the dissertation and in the end, is costly to the student and the school.

In what ways is it mentally challenging? How about the material is flat-out hard?!? And there's a lot of it. If you've never spent 10+ hours on a problem set, or even 3+ hours on a single problem, then yeah, appreciating how hard this is will probably be a foreign idea.

Another first-year here. Yes, it's very hard. Yes, there's a lot of work and there have been times when I felt overwhelmed, and I'm sure there will be more such times. But I really like the people in my program and I enjoy doing the problem sets with them. The quality of teaching has not always been great, or even good but I have learned a lot. That said, I am not working every minute I am awake, I get plenty of sleep and I tyically don't work more than 40 or so hours a week (there have been exceptions, of course). And I've managed to find time to go out to dinner with people and go into NYC. So in sum, yes it's grueling but I'm enjoying my time.