Predocs

Predocs

What is a “predoc”?

A predoc essentially implies working as a research assistant for 1-2 years, before going on to do a PhD. This comes mainly from the US, where most people do not do a MSc and therefore often lack the experience and skills to go straight into a PhD from undergrad.

Predocs are controversial due to the fact that they have become a soft requirement for PhD studies despite providing ambiguous value for prospective researchers, and by now constitute a much-diluted signal of quality for grad school applications, while entrenching elitism in economics academia.

Personally, I think there is value to be had in doing a predoc, but you need to be very clear with yourself about what you want to get out of it.

Should you do a predoc?

Predocs sometimes get a bad reputation. Before deciding to do one, be clear on what you want to get out of it, what you are missing out on, and how it fits into your plan.

Pros

• Time to build a stronger application and boost your chances of getting into a good PhD.
• Make some contacts in academia, hopefully within your area of interest.
• Get solid applied research experience, hopefully on something you find interesting.
• Get a strong letter of recommendation from your (hopefully well-known) supervisors.
• Pay ranges from decent to solid.

Cons

• You are delaying doing a PhD.
• Opportunity cost: if you can get a good predoc you can almost definitely get a far higher paying job elsewhere.
• If you end up not doing a PhD, you could have gotten better experience elsewhere.
• Work hours: Depends a lot on the project and place, ranging from normal to horrible. Reach out to current/past predocs to get a feel.

To Summarise

• If you do not see yourself doing a PhD / economic research: Probably don’t do a predoc.
• If you are considering doing a PhD / economic research: Consider doing a predoc, take it as an opportunity to explore economic research, your own interests, and to see whether you actually do want to do this..
• If you definitely want to do a PhD / economic research: If you have a good reason not to apply for a PhD straight away, then a predoc is probably a good option.

Application

Note: I only applied to predocs at LSE, but I think the experience is relatively representative.

Places to Apply

Nearly all top US unis hire a dozen or more predocs each year, keep an eye out on their job postings - sometimes this is done through specific research centres. There are a good amount of other openings scattered throughout Europe, and LSE in particular hires quite a few (about 10-20 each year). Most people will tweet about it as well, so become a part of EconTwitter if you aren’t already. @econ_ra centralises some postings, predoc.org is another good place for resources.

Cover Letter

Here is a general template that I successfully used, most of it can be reused between letters.

1. Interest in the project: Don’t underestimate this, mention some specific reasons why you like this project/group in particular. Link it to some of your own research (context, topics, methods, data… Lots to pick from). Do some research, this is a place to show you know the area of research and why this project matters.
2. Research & Coding Experience: Probably the most important section. Be explicit about the projects you worked on, what specifically you were doing, the languages or methods you were using, what you produced and contributed, and what you learned from it.
3. Future Plans: Usually, that you plan to go on to do a PhD and then do research somewhere. But how does the predoc fit into this, why aren’t you already doing a PhD?

ONCE YOU HAVE WRITTEN YOUR FIRST DRAFT: For some reference, here are the cover letters for the 3 successful applications I had at LSE: 1, 2, 3. I encourage you not to look at them before having at least written a first draft.

Coding task

Most applications will give you a coding task, usually after a first round of screening. You will usually have a choice of coding language (between e.g., Stata, R, Python for applied micro; MatLab for Macro).

In general, you will not be asked to do anything particularly complex, as the main goal is to check whether you have the basic skills you will be using on a day-to-day basis: data cleaning/processing, producing descriptive statistics / graphs, some simple regressions and analysis. Make sure your code is clear, structured, well-commented, and don’t try to do anything unnecessarily fancy, stick to basics. You can find some practice coding tasks floating around the internet, and replicating existing papers is a fantastic way to practice. There is little downside to intelligently using GenAI to assist you, as you will use it on a daily basis on the job. Emphasis on the intelligently: use it to assist you, don’t blindly copy paste, make sure you understand what you are doing and whenever possible do it yourself (if you need truth-telling incentives, you may be asked about what you did during the interview; also, from colleagues who assessed tasks: “If it’s obviously ChatGPT I give it a 0”).

Interview

Interviews will tend to be a mix of questions assessing your interest in the project and self-reflecting on your own experience, and some technical questions testing both your coding/research practices, and your economics / econometrics knowledge. Some will give you a paper to read in advance to discuss, others will ask you questions within the field of the project, give you a simple case study. Expect some questions on the coding task.

General advice for interviews: Take it easy, “Don’t Panic”. Take 10 seconds to think before you answer if you need to.