gramophone

I think you don't need to study every detail of dynamic programming of LS, if this is your first study of graduate macro. In my opinion, you can just omit chapter 1-6, if you have no problem to transform sequential problem into recursive problem (i.e., bellman equation), and if you already know how to solve bellman equation. i.e. (i) how to solve bellman equation analytically by first order condition and envelope theorem (ii) how to implement value function iteration and get numerical (approximated) solution of bellman equation. Computational issues introduced in chapter 4 LS can be omitted, but chapter 5 and 6 might be useful because these two chapters are basically showing how to solve bellman equation by examples. Have ever read Per Krusell's lecture note? I think it is more readable and actually better than LS for self-study.

In my opinion, first 15 chapters are essential and what every graduate level macro course must cover. The rest consists of higher level, or sometimes less important topics, so you may omit in your first study. In my program, one year graduate macro course is equivalent to chapter 6~chpater 15 of LS+some selected chapters among the rest+topics which LS does not cover.

It is a really nice book! I recommend it, too.

I remember Real Analysis by Stein and Shakarchi is readable.

Have you ever read "Theory of Value" by Debreu? Compact, rigorous...but not readable as Bewley and omits interesting topics you might be interested in, simply because most of them didn't exist at all when he published this book in 1959! Anyway, I also recommend Bewley's book. In my opinion, it's easy to read, and much more clear than MWG, but maintains a high level of rigor in terms of mathematics.

Very impressive. Actually, some of my friends had pretty similar profiles (of course, if they didn't lie. :-)), and all of them got two or three admissions from top 5 schools. I think your chance at top 5 is pretty high.

As others already pointed out, decision theory comes from statistics theory. Basically most papers in decision theory deal with 'individual decision making under uncertainty.' It covers questions like (1) which decision criterion is reasonable (expected utility theory, maxmin criterion and so on) under uncertainty, (2) how we should deal with data to update our belief on the world (Bayesian, and non-Bayesian updating) and so on, which apparently look like statistics problem. The topics in decision theory seem to be similar to behavioral economics theory, but there are some subtle differences between two. First of all, decision theory comes from statistics, and behavioral economics comes from psychology. Hence, their view points are a little different. For example, for ambiguity aversion problem, most decision theorists consider this as a 'rational reaction to lack of information' and try to rationalize it by various models, e.g. second order belief, multiple prior and so on. However, most behavioral economists consider ambiguity aversion as 'irrational bias' and they think incorporating this into economics model would make economics model more realistic. However, of course, two fields have been becoming more and more similar to each other, recently. I think the best introduction to today's decision theory (in economics) is Gilboa's book (title: decision theory under uncertainty).

Tsyvinski and Golosov (both are at Yale) are two most impressive researchers in public finance. But their work is basically in the context of macroeconomics, which I think is a new trend of public finance. But their tool is micro-based, of course. Yale and Stanford are definitely the best schools in empirical IO and only NWU can be comparable with them in terms of faculty, I think.

This post is interesting to me because the question is basically same as what I had. Among your choice set, it is hard to say one school dominates other schools in terms of theory faculty, even though Stanford is best in terms of "size of faculty" if you include Stanford GSB. You might already know that NYU theory faculty is not weaker than other three schools in even though its general ranking is slightly lower than others. (But I am not at NYU. I am at one of other 3 schools.) You know, there are a hundreds of considerations you should take into account. My first consideration was the atmosphere of department. Fortunately, I had friends to talk with about this issue in every places I considered, and finally chose the current program because faculty in this place seemed to be most cooperative and accessible. Stanford, Berkely, Yale, NYU...you never gonna make a wrong decision. If there is any problem, it would be due to yourself, not due to your choice.

I can't understand why some people are really concerned with RePEC ranking. This ranking is totally based on its own database, and unfortunately, I fail to find some big names in the database, with high frequency. Actually, I just randomly picked one historical paper written by Abreu, Pearce, and Stacchetti (Econometrica 1988), and found that two of them are not listed.

I agree with others. Ranking matters only in that it is easy to find a good advisor at top schools than lower ranked programs. If you are not perfectly sure about your research interest, I think it is wiser to take upper ranked program's offer. But if I am pretty sure about my research interest, I can accept much lower ranked program's offer for better advisor and comfortable environment to do research. Even in the job market, I think "who your advisor is" is more important factor than "which program you were".

"behavioral/psychological economics" can cover a lot of things. If you only mean theoretical stuff like professor Rabin's research, only Caltech, Berkeley, and other handful schools have sufficiently man senior faculties in this field, as far I know. If you also consider applied researches with behavioral argument (e.g. behavioral development economics, and so on), there can be pretty many programs in the list.

You should check kellog. It is the best place.

under the semester sysmem, 13 economics courses (including 3 graduate courses) 16 math or stat courses (including 2 graduate courses) I think my case is not ordinary.

Yale placed four in top 20. UPenn, UCSD, UChicago, Michigan. (And they placed a lot in good biz schools) Yale also had an amazing year. But I am not to say that Harvard's placement is not so impressive one. At glace, there is no other school whose placement was better than Harvard in this year.

That is exactly how my SOP was. I cited professors' papers to state my interest specifically, and to stress that my interest is well fitted to the program's strong points. But I tried not to mention only one specific field or one specific professor. I tried to mention several fields and professors which are related my interest. That's all that I wrote. If you have similar strategy, I think it will help you a lot.

I mentioned professors by name in 3 SOPs. One of them admitted me, one of them rejected me and one of them just waitlisted me. All of them are top 5 schools. In my case, it seems that this really helped me a lot, I think. But it is beyond doubt that I mentioned only active professors in my SOPs.

Melitz is going to Princeton, according to Mankiw's blog. I remember words saying Kehoe seems to go to Princeton. I don't know anything about Randy Wright.

In my school, "Real Analysis" is a course for measure theory, Lesbesgue integration and some preliminary functional analysis results. For this course, Royden is a standard text. The textbook wriiten by Stein (with another coauthor...I forget his name) is also a good start. If "Real Analysis" means a course for introduction to metric space, definition of differentiation and Riemann integration and so on, I think every textbook mentioned above are ok, but I prefer Rudin's PMA. Marsden's textbook is also good.

I think Harvard and MIT are absolutely the best programs. The faculty's experience to have studied a lot of marvelous students also should be considered. If you are interested in theory, Stanford GSB can be a tier. Traditionally, Princeton, Stanford and Berkeley are considered as the second tiers. It's still true, I think. But the gap between them and Yale or Northwestern is now little, if exists. I don't have any idea about Chicago. But as I know, some good students hesitate to go to Chicago recently. It's not good signal. It's just my opinion. :-) I think that you should make a dinstinction between good program for researcher and a good progam for PhD student.

I also have 3.5, but I have been admitted top schools. I also know another guy who has 3.0 but was admitted to Yale and Princeton. I think it does not matter.

In fact, I heard exactly same comments. Let's exchange ideas at the visiting weekend in Yale. (I can't go to Stanford this spring. I'm not in USA)

I have been admitted by both Yale and Stanford, but I don't know which program should I choose. My prime interest is micro theory (for example, repeated game, contract theory, etc.), and have a little interest on metrics, but I also know that in many cases, students would change their topics during their course works... I know that traditionally Stanford has been stronger in the field of micro theory, but I heard that the gap between two become little after good and active researchers have moved to Yale, for example, Samuelson, Bergemann, Horner, and so on. I'm also waitlisted by some schools which have even stronger micro theory faculties, but at the present, only Yale and Stanford are my options. I need your advice! Thanks.

I am also notified via email that I have been admitted. I am relieved now! I have been only waitlised by my dream schools (except Yale) inculding so-called Top 5 schools.