PhD in Economics Advice for Research Heavy Undergrad

[Huzker]

Bayes, I don't want to get drawn in an unnecessary fistfight over this but I know several predocs at top 10 who take Real Analysis for the first time as predocs. Top 10 PhDs are far far harder to get into than predocs, partly because until recently predocs were largely restricted to domestic students (MIT SEII still might be restricted to domestic student). The inelastic supply of international students, along with their vastly superior preparation in terms of coursework (they often have taken the entire PhD core), means that domestic students are forced to exploit their comparative advantage in applied research. Often, this means less focus on math and more on actually doing RA work since early in undergrad. In fact, at my predoc, most predocs are not sufficiently well prepared to take PhD level first year courses (and in fact mostly opt to take 2nd year courses). There are a large number of incentive issues too here, because often not taking PhD micro is better than taking it and not getting an A.

See here the people taking MATH 301 are predocs taking analysis for the first time. One person is taking grad micro for credit out of 30.
Attachment 7276

Also, yes, I don't believe it is worth it to go to schools outside a certain range as they just become a way to exploit cheap labour from grad students via teaching assistantships. You want to have fun teaching intro micro while taking a full load of grad courses just so you can get your meager 15,000 stipend go for it. But don't sell these idealistic pipe dreams to people, where they slave away for 5 years at a program only to get jobs they could have had with a bachelor's degree.

Frankly, you sound like a bitter and jealous person by insinuating than I am not a predoc; I don't have anything to prove to you or anyone else and I am just offering a pathway for people who didn't go to Ivies as undergrads to get into predoc positions. I didn't go to an Ivy or a top 15 phd granting school (I went to a top 15 LAC) and I got into a predoc just fine. If any curious person reading this really wants to verify my credentials they can do so by messaging me and I'd happily tell them who I am.

Went to the same school as where alphatrunk works (from info of the screenshot), I can confirm what he/she described regarding coursework preparation for predocs and domestic students to be accurate. However, acceptance of predocs are more about fit than coursework preparations, or other factors typically considered for grad application. It is entire possible for one to get rejected anywhere from predocs but get into top 5 PhD programs, or the other way around

[guessandverify]

I'm not sure why this was downvoted, to be honest. As a T30 predoc, a lot of this is good advice. My predoc let me become "indispensable" precisely by mastering computational methods, which ended up leading to coauthorship on the project. But even if you don't end up becoming a coauthor, the prof can then sincerely say "the project would not have happened without Predoc X."

As for advice on applying to predoctoral programs, the fact is that you will end up applying to pretty much all of them in sight. However, you can "optimize" your placement outcomes with a few heuristics: try to pick faculty who are established scholars (with top 5s), even if they are at slightly lower ranked schools over APs at top schools. Try to pick *projects* where you have an absolute advantage in some aspect of the project, not just a comparative advantage. Such projects often involve heavy computational component, interacting with proprietary server backends (such as projects that require use of Facebook data; see the work of Michael Bailey, Johannes Stroebel and Theresa Kuchler), and use of machine learning methods where RAs quickly become indispensable. Of course, pick projects that you are actually interested in for the most part; you want to be able to write credible cover letters for these positions.

The tasks you will get if you hear back would depend on the type of position: reduced form/applied micro type positions usually have a Stata/R exercise where they make you clean some data, figure out some specifications, estimate these, and report back with tables and figures. I haven't seen what data tasks for structural work looks like (even though my current position is in structural); some people in my cohort were asked to submit SQL code samples and MATLAB samples from existing projects. These could include code for dynamic programming (so like value function iteration), or discrete choice demand models in the BLP style (some IO people at Yale look for this). For theory, you usually get a problem set that depends on the field you are applying to work in. Theory positions are rare.