Jump to content
Urch Forums

On PhD Applications


Recommended Posts

Dear TMers,


I'm back! With my promised thread on PhD applications.


As some of you may know, I was rejected by all top 10 schools for fall 2010 entry, but I've had much success in the 2012 app cycle. I think I have learned a lot about the application process, and I'll try my best to articulate what I've learned in this thread. Just a head-up: what I'm going to say here will mostly be about the top PhD programs, and since all the information here represent only my personal opinion, you should take them with a grain of salt. Here's my profile and results: http://www.www.urch.com/forums/phd-economics/139568-profiles-results-2012-a-2.html#post906782


Top PhD programs want to admit people who will (1) perform well in coursework and (2) excel in independent research. The primary way of evaluating (1) is to look at the history of courses the candidate has taken. The primary way of evaluating (2) is through letters of recommendation.


On courses - there are a lot of discussions about which courses would help PhD application. I'd have to say that I concur with most people that while I don't think there's any hard requirement beyond basic calculus and linear algebra, it would help a lot if you have challenging courses under your belt with good grades. By "challenging", one should think from the adcom perspective: they want to see signals on your transcript that show you are really smart and academically outstanding. Which courses fit into this category largely depend on your institution and your other constraints: it could be challenging undergraduate math classes (for LAC applicants), PhD level economics classes at a good econ department (for those who have access to such classes), or it could simply be undergraduate econ courses at a well-known school (if you go to a top undergrad). One simple rule-of-thumb, as Mankiw mentioned on his blog, is that more math cannot hurt (on this "coursework" dimension of your profile, ceteris paribus). I added the parenthesis to Mankiw's statement because there's usually a trade-off of your time: more math could mean less PhD econ class, less research, etc. That's a trade-off that only you can make because it depends on the opportunities available to you and your comparative advantages. I've personally been following this "math rule" until I landed my RA job at a top 5 school, after which I took a PhD econometrics course.


On research: although everyone knows that research experience is important for PhD admission, I want to emphasize that this is usually the dimension that's lacking for most "frustrated" applicants who think they underperform in the applications. In this great post, one of my fellow MIT classmates uses the word "crucial" on RA work: For the ambitious, prospective PhD student: A Guide : Core Economics

I personally cannot think of any single person in my MIT class who has not worked as a "serious" RA before. I know that "serious" is a vague word, but I'm still throwing it out here because during my application in 2009 I thought I had done "serious" RA work when I in fact had not. You could consider your RA work "serious" if it fits one of the following categories: 1) you've done RA work full-time for at least one year; 2) you've done RA work over the summer under direct supervision (e.g. meet 3 times a week); 3) you've done RA work part-time during the school year and your name has been acknowledged in your Professor's paper / draft. This is not and should not be a hard criteria, but the rule-of-thumb for judging whether RA experience under your belt is "serious" enough is to ask yourself the following question: would your professor know you any better if you work for him/her full-time for an additional month? If your answer is no, then I think you should be confident that you've done some serious RA work and that your professor has already known you enough to make a good judgement of your research ability.


Be reminded that grad schools want to admit candidates who will excel in independent research, and while doing RA work for someone will almost certainly show some aspects of your research skills, it will not necessarily show your ability to work "independently" in coming up with research ideas and executing it well, hence it might not be sufficient, and it is definitely not the only way, to show that you will excel in independent research. Another great way to do that is conducting independent research, or writing a good thesis.


On letters: the ultimate goal of adcoms reading your recommendation letters is to obtain unbiased information about you that cannot be obtained from the rest of the application package. Therefore, from their perspective, reading letters from economists they know is almost always preferred over reading unknown letters. Therefore you should almost always pick recommenders who are established. Beyond this criteria, I personally do not think letters from big names (e.g. Nobel Laureates, famous economists who are 50+) have any real advantage over letters from those established, but not necessarily renown, economists (e.g. young, recently tenured economists at a top 10 school).


The noisiest part of your profile is usually your research potential, and that's precisely why RA work has been emphasized - your professor will be able to write letters about your research potential. Once you've chosen your recommenders, however, you should not let the content of the letter just be in their hands, but instead you should actively communicate with them about what part of your profile is not as visible on paper, and see if they can help you by reducing the noise of the signal in their letters. For example, if you have a piece of excellent independent research, or if you took a particularly difficult class, you could always communicate with your recommenders on these signals. Repeated signals here might even be a good idea - if your well-known recommender A knows your class performance in [DIFFICULT PHD CLASS], and your not-as-known recommender B supervised your excellent thesis, you could always let A read your thesis (and ask for feedback!), and ask if he could include a comment about it in his letters (most experienced recommenders will not need such reminders!) As a rule-of-thumb here, I think research-letters are almost always preferred over coursework-letters because information contained in the latter usually repeats in your transcript and if not, it could always be explained in the research-letters by asking your coursework-professor to write a short email/ note to research-professors regarding your class performance.


SOP: many people say this cannot help you if you write well, but it will hurt if you write poorly. I think empirically this is not true: I wrote a great statement this time with 50% of the space about an independent research project, and I repeated get asked / hear positive comments about my project during my school visits. I believe SOP not only could help at the margin, but it often serves as the "glue" that keeps your whole profile together.


Final remarks: what has changed between 2010 and 2012 in my application: I have 1) performed very well in a PhD class at a top school and got a letter out of it; 2) worked hard at an RA job and obtained three research letters from established professors.


Personally, I think recommendation letters is the single most important aspect of your profile that sets you apart from the rest of the applicants. I'm not saying that your transcript is not important, but I think most of the "underperformance" frustration sources from this factor: many excellent students at LACs or universities outside of the top 10-15 have great (and similar) transcripts, but many do not have access to good research opportunities or established recommenders. This is where full-time RAs can be especially helpful: I come from a top LAC with 3.96 GPA and a math major, but I didn't have access to any established researchers that could write my letters in the 2010 cycle.


Some remarks on Plan B: there's always the question of whether one should take a good offer this year, or wait for a year or two and work as an RA and aim for an even better offer. That is largely situation-dependent, and I cannot comment too much on this without knowing the candidate really well. However, I should say that if you are confident about yourself and really think you are underperforming in one cycle, waiting one/two years is actually neither bad nor uncommon. Over the past two years, I've learned so much at the job from my professors that not only helped me on paper for the applications, but it has also helped me grow tremendously as a researcher. In fact, I think more than 50% of MIT's incoming class who did undergrad in the US have some full-time RA work. The proportion is similar at the other top 5 which I RAed at. However, even though I would recommend my RA job without any hesitation at all, one should be cautious in choosing a full-time research position, be it at a top school, think-tank, Fed, or other research institute. You have to be confident that the professor /researcher you will work for is really interested in helping you grow as a researcher, aside from getting you to do work for him. This is not always the case, even with well-established professors.


I hope you have gained some useful information in this ramble, and please leave any questions in the thread and I'll try to respond periodically.




Edited by OutOfGame
  • Like 2
Link to comment
Share on other sites

Thanks a lot, OutOfGame, I remember seeing your profile when browsing the Profiles & Results threads, and I must say I was impressed. Rejected by all top 10 schools, despite a seemingly strong background, and going away to strengthen your profile, then re-apply and get accepted everywhere was very impressive. Congratulation to you!
Link to comment
Share on other sites

Thanks for your post OutOfGame. How much will having a Master's thesis(from a very well known non US masters and with a letter from the thesis supervisor) as one's only research experience hurt? In between my MSc and starting a PhD in the US, I have to return to my home country for a year. I can RA there for a whole year, however the professors there are not even close to as well known as those where I'm doing my MSc. How would a strong course work based letter from a famous economist compare to an equally strong research based letter from an unknown economist? If I stick to the better known writers from my master's institution while RA-ing back in my country, will the RA experience without a letter count for anything? I know it's very hard to come up with specific answers to these questions, but I'd appreciate any thoughts.
Link to comment
Share on other sites

Hi, thank you for this excellent post I found it immensely useful.


I saw your profile you have taken advanced math courses like Func Anal, Measure, etc., what is your opinion about the difficulty of the grad level economics & metrics courses as compared to the advanced courses that you have taken?


I know there is no straight forward answer, but, I just want to have an idea about it.


Thanks again and good luck.

Link to comment
Share on other sites

badasseconomist: short answer to your question is that Func Anal / Measure was way harder in content, but PhD metrics had a huge variance on grading even after conditioning on your background, so it's a touch call. In my metrics class, the source of variance comes from: 1) prof put down a question worth 40% on the midterm on a topic that she said she won't test us; 2) the TA was quite irresponsible and was grading exams and Psets in a very ad-hoc fashion (so I got 8% on that 40% problem, a few friends also didn't prepare at all and scribbled something down randomly but got 35%+; also there were quite a few Psets in which the grades were written when the Psets were held upside-down).
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...