Columbia PhD and post-doc advising

Prospective students or post-docs: Applicants often email faculty to ask if they will be their advisor. If you work in my field of expertise and are admitted to a Columbia PhD program political science, SIPA, or economics, or the Earth Institute post-doc, then I consider it my job to be your advisor, so long as you are dedicated and the guidelines below appeal to you. Given the overwhelming number of applicants to these programs, it is uncommon to speak with faculty in advance. Merely explaining in your application letter which faculty you would like to work with and why is sufficient. For more details, see my advice on applying to PhDs.

Current Columbia PhD students: I am always happy to talk to Columbia PhDs student about their work, regardless of field. Please do contact me or sign up for office hours online to discuss. New students should feel free to come by just to chat generally, especially first year PhDs. The advice below applies mainly to regular visitors and (most of all) people who will be main students. Before coming to see me, however, do look at the advice posts on the side bar first, to see if they apply. That way we can skip generalities when we meet and get to the meat of the discussion.

Macartan Humphreys and I have been thinking about ways to make our PhD advising more effective. We think if we communicate our expectation and work habits more clearly, and put some structure and commitment around working together, then our students are going to succeed all the more.

To that end, here are some thought on how to work together and some more specific guidelines for our core students.

General tips on working together

Most advisers are flawed people set in their peculiar ways, and so busy they feel like they are losing their minds. I am no exception. In light of this, I can’t underestimate the importance of “upward management” in your work–with me or any time in your career.

Based on some of the feedback I’ve gotten from my past grad students, and some information they said was helpful, here are some ways that I think I advise most effectively:

  • It’s always good to send concise written updates (a couple of paragraphs by email) in advance of a meeting and, for specific questions, to try to formulate them beforehand
  • If you are working with data or theory, I recommend bringing figures/tables to our meetings to discuss–it’s much easier to give concrete feedback
  • I tend to be blunt, and see it as my job to tell you to drop things that you haven’t sold me on. The idea is to have you build a good argument for keeping going, or downgrade it and work on other things.
    • If the dissertation is the goal at the end of a soccer field, think of your ideas as balls you kick down the field: You need to keep several going in the early stages, but drop the ones that don’t move
  • Like other faculty (especially ones with small kids) I usually need advance time to review something
    • minimum 24 hours for a 1-2 paragraph update before a meeting
    • 1-2 weeks for a longer paper or memo
  • When I say I will do something by a certain date, I will usually do it, and if I don’t, I ask you to poke and remind me because sometimes things get missed.
    • I expect the same of you: commit to deliver things by specific dates and stick to them
  • I will make extensive notes on what you send me, often electronically, so if this is something you want:
    • Please email as a PDF
    • Please have 1.5 inch margins, 1.5 to double spacing, and 12-point font
    • These instructions sound anal (and I am!) but in general I think people give more and clearer edits and comments when they have the space to do so
  • One weakness: I only remember something if it’s written down, so please send me an email if you need me to do something
    • If you don’t have a response in 48 hours for small things or 1-2 weeks for big things (such as “please read this paper”) then either I’ve been hit by a truck or I just forgot–in either case send me a reminder email
  • Another weakness: I am quite forgetful and will usually contradict my past self’s advice all the time–apologies in advance
    • On the plus side, it’s like getting two opposite-minded advisors in one
  • In general, faculty will start to forget your work if you only see them every four months, so visit regularly
    • It comes down to what you want to spend an hour on: catching us up to the frontier or pushing the frontier together?
    • I would normally expect my core students to be seeing me, showing me tables/figures, and discussing progress at least twice a month

For those who have Macartan as an advisor, he likes all of the above too.

Finally, I can focus on your specific work in more depth if we jump past all the generic advice. Please take a look at the advice posts that apply to you:

Guidelines for core advisees

Macartan and I encourage students to seek us out regularly, whether we’re in your field or not, or a main advisor or not.

Our thought is that for our core students it would be useful to have a small, well-defined group that we interact with regularly and that interact with each other regularly in more or less structured ways. There are several reasons for this: to try to standardize and raise the level of training; to economize on certain advice and advising time; to increase communication and feedback; to give students structure and commitment devices; and to raise the value of but also the expectations for working with us.

At the end of the day, we want you to have the absolute best possible dissertation and job, and we think this will help. If it doesn’t, we plan to change the model.

Participation will be required for people who have one of us as their main adviser. That is, this is the main forum in which we will advise our main students. It’s largely a substitute for for office hours, though we will still see you one-on-one a lot. The reason we are doing this is because, in our experience, our students tend to disappear for long periods of time during which our aging brains forget what they are doing (and we worry that their younger brains are not making dissertation progress). This is a commitment device for us all.

Key features would be:

  • A focus on a peer group of 2nd, 3rd, and 4th year students
  • We expect the peer group to consist of about 6 to 8 students all working on themes broadly related to the political economy of development, violence, or experimentation
  • All members would be expected to share a set of skills and in particular to have taken the formal theory and statistics sequences, and ideally some strongly recommended courses (below)
  • We will have at least 5 or 6 meetings together per term
    • In every meeting, everyone should give a quick update on work done and progress since the last meeting
    • 1 or 2 students will also present something more extensive, having provided a short memo (and ideally figures/tables) 24-48 hours in advance
    • Everyone will be expected to read these in advance and give feedback
  • Since so many of our students’ work is field work in developing countries, there will also be a focus on funding graduate study and field work, helping you identify project partners and funding sources, and advising you on proposal writing and project management
  • Students of Macartan are required to attend the weekly SSDS seminar and present there once a year, and this is strongly recommended for Chris’s students
    • I’m not making this a requirement for students who are only mine, so long as I cannot attend regularly myself
  • We discourage being away during the semester or living in other cities, except for field work or urgent family reasons of course
    • We may experiment with video conferencing in these exceptions, but not until we’ve fined tuned this more general advising approach
    • Why do we discourage this? To be honest, from experience I’ve seen it is detrimental for students to live elsewhere just to be with a partner and (absent children) generally see grad school–like a job–demanding a physical presence to be really effective. You’ll have to balance competing commitments in life, and make your own choices, but in the meantime our job is to tell you how to get the best dissertation and job.

We don’t inflict this overzealous approach on all the students we talk to, but rather our smaller group of our main advisees. You may find this merciful or not. Please let us know if you want to join the meetings. We keep the following in mind:

  • We are hoping to keep the group small (under 8 or 10) since we think that will be more effective, and that’s probably the natural number of main advisees
    • Our hope is everyone who is interested can participate without getting to an unwieldy size, since our goal isn’t to create an exclusive club or clique (and if that happens we’ll probably consider this experiment failed)
  • We will generally focus on students where one of us is a main adviser
    • Note that this is “a” main adviser and not “the” main adviser, since we know some of you will have one of our colleagues as your dissertation chair
    • Indeed, this should in no way stop you from seeking advice more broadly with the faculty (and in fact we will insist on it)
  • Basically if your subject is one where you would be coming to see us regularly for advice, then you belong in the group because the whole point is for us to play this role more effectively
    • Ergo, if you think this would be a good group for you, come to see us regularly in office hours and it will undoubtedly happen
  • The composition of the group may change from year to year as some people drift to new topics or interests, and away from areas where we can be of help

Taking part in this is not without commitments, but we think it will lead to a richer graduate education and better outcomes.

Methodological training expected of core advisees

Over time we will may develop more formal advice and requirements on courses. At the moment, if you want one of us to chair your committee or otherwise be a main advisor, we typically expect you to have the following:

  • At least a 2-course sequence in game theory
  • At least a 2-course sequence in statistics or econometrics (encompassing linear and non-linear regression methods)
  • At least one course in more applied caual inference, such as:
    • Green’s experiments course
    • Hirano/Wawro/others on applied causal inference
    • Others as they are offered

Note these are requirements for people for whom we are main advisers, not necessarily all the students who will interact with us or have us sit on their committee.

We make these requirements because we see these skills as essential to basic literacy in the profession, especially in the political economy of development. So are many other skills, but stats and game theory are among the hardest to learn on your own. And your life after graduation is a slow decline into technical obsolescence, so beef up now.

The following, meanwhile, are strongly recommended for our core students:

  • One of us (and failing that, someone else in the department) will usually be teaching a Political Economy of Development class and we generally expect our students to take it
  • If the department’s course is not available, Suresh Naidu’s course is an excellent alternative. You should consider both if PE of development is your main field of study
  • Nunnari and Casella teach a 2-course sequence in formal political economy theory and lab experiments. You should consider one at least and both if you specialize in lab experiments or political economy theory

Other coursework worth considering if you are specializing in PE of development:

  • We assume you have done the comparative core (international relations for some of you)
  • Microeconomic theory
  • Auditing or taking the development economics courses by Kaur, Pop-Eleches, Urquiola, or Verhoogen
  • Courses in the statistics department can be good ideas but should not substitute for the required or strongly recommended courses above
  • I think languages are better learned in the summer in the place where people actually speak them, and advise you not to use up a grad credit on them when you only have about 16 in your life
  • We are encouraging of other methods (comparative cases, ethnography, archival research, etc) and have advised many such students in the past, but given that we specialize methodologically as well as thematically we expect our core students to at least have minimum expertise in quantitativeskills
    • We don’t expect this to be an issue, since students with a very different methodological approach might have us on their committee, but we would be unlikely main advisors
    • Do strongly consider getting training in formal qualitative methods as part fo your graduate work since this is a rare opportunity to learn them