Integral

Econphd.net's field rankings were an excellent resource, but yes -- they are a decade old and departments change. I don't know if there are updated in-field rankings aside from hunches, guesswork, and "informed knowledge" lurking around. You'll need to hunt and peck at individual department websites to get up to date information.

Note that Conley's paper is about research outcomes, not initial AP placement, so it doesn't tell us much (if anything) about the benefits of being a middling T5 vs "the best in your year" at a T20 for placement purposes.

Transferring up is very difficult, and you may be required to retake coursework. There's no reason why you couldn't spend your third or fourth year studying at the T5 you desire. If you're really that good, your advisors could arrange for you to take a year off visiting that other school. Note also that it may be better to be the #1 student at a top-20 than a middling student at at top-5. You'll have more resources devoted to you at the top-20, extremely strong letters, and the placement coordinator/your advisors will pull for you. (I do not have data to back this intuition up though...someone could presumably go through the placements of all schools in the past few years and check middling T7 candidates against top-class T20 candidates).

The marginal value of the verbal (above some middling threshold) is precisely zero. I don't even remember what I got on the verbal part three years ago. Send in the 170Q. Isn't 170 the upper bound nowadays? So yes, send in that one.

Math: I'd do the following. Complete Calc I-III, LA, Diff Eq, intro stats, and WUSTL's required intro to proofs course by second semester sophomore year. That gives you 4 semesters to take advanced courses. Junior year, take the 4111-4121 analysis sequence. Consider taking the 493-494 statistics sequence. If you like macro, take 495. Senior year, take additional courses in topology (417-418) and graduate economics. Look up the math major's Track E Economics: Talk to professors to get a feel for their "grad-track" courses - the subset of courses that are harder and more academically oriented. Take Econ 413 as early as you can, probably sophomore or junior year.

An undergraduate degree in economics is solid preparation for graduate work in business (MBA), law (JD), and public policy (MPP). If you have enough econometrics, it can be excellent preparation for a statistics MS program or graduate work in survey methodology.

Echoing msokrates here. There is a book. "The Workflow of Data Analysis using Stata," by J Scott Long. It's going to set you back about $55. Buy it. It will change your life, or at least how you go about doing empirical research. It's an A-to-Z set of guidelines for documenting your empirical work. Don't worry about the "Stata" in the title, 90% of the book applies to any empirical project in any programming language. He goes over all of the steps, from setting up your directory structure, to inputting raw data, to cleaning data, to analysis, to presentation, and describes useful tips on leaving a paper trail so that you can reproduce the analysis months later. I cannot recommend the book enough. Documentation, readable code, plentiful comments, and reproducibility are the most important technical skills you can have. For an idea of "best practices," pitched at an undergraduate level, see here: http://www.haverford.edu/economics/faculty/rball/Instructions.pdf. The book I mention above is a much longer treatment and goes into far more detail. It is well worth your time to develop a workflow.

Another vote for about 600-ish words, no more than 2 single-spaced pages. Use schools' requirements as a guideline. UC-Davis was as low as 500; some other schools limited the SOP to 800. So 600ish is good for a general draft; make some longer and some shorter as you need.

Oh, that gets hard. Am I under oath? :P Here's my feeling. 1-11: As above. Harvard, MIT, Chicago, Princeton, Yale, Berkeley, Stanford, Columbia, NYU, Northwestern, and Penn (the "top 10"). 12-25...fudge a bit and let me have 15 schools: Ten that probably deserve it: Wisconsin, Michigan, Minnesota, Caltech, UCLA, UCSD, Cornell, Duke, Carnegie-Mellon, and Brown. Then, I don't know...pick a couple from: WUSTL, Maryland (?), Boston U, Rochester, UVA, UNC, Michigan State, and Boston C (?) (my bias here being obvious). How you rank inside that group depends on your interests. For macro, Minnesota is higher. For applied micro, Michigan State is higher. For econometrics, UCSD is very high, indeed top-10 or top-3 material. For experimental, Caltech is strong. At the end of the day it's less about strict ranking, and more about "bundles of schools you should apply to if you want to do X." I know less about international schools and will refrain from commenting on them.

One more vote for Top10 = {MIT, Harvard, Chicago, Princeton, Yale, Berkeley, Stanford, Penn, NYU, Columbia, Northwestern} Yes, that's 11 schools. Maybe toss in Minnesota (especially for macro, and for their legendary first year).

1. It's certainly worth asking, though it shouldn't affect your application decision. 2. I have a buddy at BU and will ask him. 3. Marginal cost is low, apply to BU regardless of answer.

I wouldn't say it's an uncommon combination, but perhaps I'm biased by my own experience. MWG + Varian was the textbook combo in my undergrad's PhD program (ranked roughly 40-60 in econ). The first semester covered the first 14 chapters of Varian in lecture, with problem sets coming out of the parallel chapters in MWG; the second semester was all out of MWG. The idea was to ease the students into MWG - some of whom had weaker mathematics backgrounds. An MWG+Varian course can be taught in two ways: either with Varian as the main text and MWG as a supplement for more advanced students, or with MWG as the main text and Varian as a supplement for weaker students. OP should certainly figure out which is the case before enrolling in the course.

Easily the most true post in this thread.

I use a Nexus 7. Ranks high up on the list of "best $200 I've ever spent." I do think the 7" form factor dominates the 10" for leisure reading, assuming your eyes are good enough (mine are right on the margin). 10" is too large to cart around everywhere; 7" fits in a coat pocket. I had a 10" tablet in my first year and never used it. I use my Nexus7 every day.

I"m treating this as "only" two books, so I'm shooting to maximize average value, as opposed to marginal value given my current bookshelf. Academic: Dejong and Dave, Structural Macroeconometrics Leisure: Kierkegaard, Either/Or Why: Dejong and Dave contains a two-year or even three-year graduate macro sequence within its pages. There is very little in applied macroeconomics you need that isn't in that book. Kierkegaard's book is a treasure-trove. Part A is a collection of aesthetic essays that bring richness and variety; Part B is a tightly-focused ethical essay. I never get tired of it.

I can only bring in a tiny bit of anecdotal evidence, but it seems to me that it's easier to go from mathematics to economics, or from economics to political science, than the reverse. Even then, it's not widespread. Look up, say, Chris Blattman. He's someone on the edges between econ PhD, polsci PhD, and MPAID, which sounds like the field the OP is shooting for. One can get a Political Science PhD and do work in game theory (in political science jargon, "formal modeling") - but expect it to show up in the APSR, not the AER or JET. One can do empirical work in political science, but it's going to end up in the AJPS or Journal of Politics, not in the Journal of Econometrics or REStat. It is very difficult, if not impossible, to get an Assistant Professorship in an Economics department with a Political Science PhD.

You are in the right sequence. 320 is the correct "first course in analysis." You do not have to take any more analysis, but 421 would likely be useful for you. 421 dominates 425 for admissions and for surviving the program once you're in.

applicant12: Come on, you're an economist. (And I mean that in the least arrogant way possible) optimal number of work hours is the smallest you can get away with subject to the GPA constraint for funding, the comps constraint for staying in the program, and the burnout constraint that beyond some point, more "work hours" bring you negative output. That number is usually somewhere around 50/week. But if you love what you do, it doesn't feel like "work." I work 40-60 hours per week on macro-related stuff, but I count most of that as leisure.

On the bright side, refereeing is guaranteed to be double- or even triple-blind!

Background: I'm a longtime Windows/Linux user, but every single computer in our department is a Mac. I've had good experiences with both. In my view, the three most important considerations are: 1. Software needs 2. Your personal workflow 3. Cross-computer compatibility Software: The big three software requirements of an economics PhD - LaTeX, Stata, and Matlab - are available for both Windows and Mac machines. As of version 12, Stata runs a nice GUI in both Windows and Mac environments. Matlab runs identically on both platforms. LaTeX has excellent plugins on both. There is really no difference between the two on those fronts. However, Mac has the excellent Papers software for handling your growing PDF collection; Windows has a suite of utility programs that I could not live without (notepad++ in particular). Given that the "big 3" work well cross-platform, you can make the choice based on software you personally cannot live without. Which brings me to... Personal workflow: Some people just work better on Macs. Some work better on Windows machines. Each OS has its own quirks, and I don't think either one is compelling enough that you should switch from the other and pay the cost of learning a new OS's foibles. My workflow is heavily dependent on Notepad++ and Q-dir, both of which are Windows-only. One of my professors is wedded to Mac-specific utility software. It comes down to taste: your productivity is paramount. Compatibility: If the entire department is running Macs, it helps if you have a Mac as well. Support will be better and there is no risk of compatibility issues. Similarly for the Windows case. Give this less weight than the others, but it is there. Note that ThinkPads retain a VGA port, which is immediately compatible with the presentation tools used in most universities; Macs require an adapter. Other issues: You'll probably never open Excel, Word, or PowerPoint during your first year. Word documents become LaTeX markup; Excel spreadsheets turn into Stata data and do-files; PowerPoint presentations become LaTeX/Beamer presentations. (Edit: "never" is too strong, but grant me some artistic license.) It is not immediately obvious to me which platform has better Box/DropBox support - perhaps others can weigh in? Some on this forum swear by their cloud backups. Note that Macbooks will cost more than similarly-specced ThinkPads, HPs or Dells. Bottom line: Go with what you're familiar with, as the core software you'll be using will run smoothly on either platform. I personally do not know of any compelling reason to switch from one platform in favor of the other. About half of my classmates use Macs, the other half use Windows.

Everything synth said is true. To sort of reiterate, 1. You can always send your fall grades in, even if the school doesn't request them 2. I don't know of any school that wouldn't add your fall grades to your file 3. Very few people look at the applicant files until after job market season has calmed down. I would say "no-one" but I don't like absolutes.

I had no idea you could purchase Matlab for $100. Is that a perpetual-use license? I'd be more than happy to fork over $100 for a personal, perpetual copy. Anyway. I recommend both Stata and Matlab. Stata has deep roots in the applied microeconomics literature. Matlab has deep roots in the macroeconomics literature. Both of these are great for securing RA jobs. I like Stata. I've been using Stata for 6+ years. It's my home language, and so obviously I'm biased in its favor. In general, R is a fantastic tool, and has the advantage of being free. I don't discourage it, but I encourage Stata/Matlab for their compatability with other researchers' code. It depends on which "bubble" you are in. Governments are in the SAS bubble. Macroeconomists are in the Fortran/Matlab bubble. There's an R bubble. There's an extensive Stata bubble, and so on.

Software: The Core: Stata, Matlab, and LaTeX (via MikTeX and the associated TeXworks editor) are the essential core of "productivity" software for...me, at least, if not most grad students. Spend a little time setting up estout and linking your Stata -> TeX workflow together. Text editors: A good text editor will double your productivity. I like Vim on Linux and have grown fond of Notepad++ on Windows. You're going to spend most of the next two years looking at a text editor - for writing Stata scripts, Matlab scripts, or LaTeX documents - so spend some time to find one you're comfortable with. Office software: OpenOffice and LibreOffice are great, free alternatives to Microsoft Office and have full- or nearly-full Word, Excel, and PowerPoint compatibility. Not that you're going to spend much time in Word or Excel. Backups: Dropbox and Box are great web services for backing up your work. I'm on Box because it offers more free storage (10 GB instead of 2 GB) and has better sync to my phone. If you have a Windows computer and want to sync to a flash drive, you can use the Microsoft SyncToy (it's free). On Linux, there's Unison. Odds and ends: A good windows explorer, one with tabs, will make your life easier. I recommend Q-Dir. Similarly, you will find a tabbed PDF reader to be worth your while; I like Foxit. I've never used EverNote/OneNote/etc, but some find them useful. Organizing your PDFs is (or will soon become) a nightmare. Pick up Mendeley or some other useful PDF tagging software...what is really needed is an iTunes for PDFs. I don't think a solid solution exists yet. -- Other general tips: Since you've just reformatted your hard drive and are starting fresh, think carefully about your filesystem and how you use it. I have one folder per research project, with subdirectories for code, the TeX draft, associated literature, raw data files, etc. Set up a system that you like: consistent workflow pays great dividends in the beginning/organizational stages of your research, and has a high multiplier. Pick up the book, "The Workflow of Data Analysis using Stata." Much of the advice is useful in general, and is not Stata-specific. Make sure you're backing things up, either to a flash drive or to the internet via Box/Dropbox.

Redacted by moderator.

Sure! I like to highlight the more focused and less well-known bloggers, but your point is well-taken. I learn more from Rowe and Kimball than I do from Krugman and Mankiw - but your mileage may vary. I keep Cowen in my daily readings precisely because he dabbles in a variety of topics. Money-macro is blessed by a relative abundance of quality bloggers. Ah, I forgot Mark Thoma, whose daily link lists are indispensable. Surely worth putting into your Google Reader replacement of choice.