WARNING

Examples below (in parentheses) might be outdated since positions and applications change all the time!

I found the process of applying for predocs quite painful (actually more than graduate programs), so I hope this blog helps! Don’t be too anxious!

What Are Predocs?

A predoc (short for pre-doctoral fellow) is someone engaged in academic research before starting a PhD program. The term is commonly used for individuals working as research assistants or fellows at universities, research institutions, or labs. Predocs typically gain research experience ~~good recommendation letters~~ to strengthen their PhD applications.

Depending on the institution, predocs may hold titles such as research fellow, research scientist, research associate, research professional, research technician, or research assistant.

Where to Find Opening Positions?

The most important source of open positions is your connections. For example, you can directly ask your supervisor if they know of any openings through their private network. Using personal connections maximizes your chances of securing an ideal position—one that aligns with your research interests—while also saving time and effort.

Beyond personal connections, key online resources for predoc openings include:

During the main application season (July to the following August), these sources may list over 150 positions. Employers include:

Research institutions (e.g., Children’s National Hospital)
Universities (e.g., Northwestern University)
Private labs (e.g., GovAI)
Companies (e.g., Integra FEC)
Federal/governmental institutes (e.g., The Federal Reserve System)
NGOs (e.g., the IMF and World Bank)

Most positions are based in the U.S., but opportunities exist worldwide. Generally, economics-focused positions are posted earlier (July to December), while business school positions are posted later, beginning in January. However, some business schools—such as Stanford GSB, Northwestern Kellogg, and Columbia Business School—may post openings earlier. So if you don’t get an offer (or even see an interesting position) before February, don’t panic, more are to come!

Considerations for International Candidates

If you’re applying from outside the employer’s country, it’s crucial to check visa sponsorship requirements and citizenship/residency conditions or so. For example, some U.S.–based positions may require candidates to have lived in the U.S. for at least three of the past five years (generally due to involvment of sensitive data).

Applications

The first step in applying for a predoc position typically involves submitting application materials, which may include:

CV (usually two pages)
Cover letter (typically one page)
Unofficial transcripts (for all degree programs and sometimes individual courses)
Writing/coding samples or GitHub portfolio (if applicable)

Institutions have different application submission methods:

Public application systems (e.g., Empirical Fellow at Northwestern Kellogg, J-PAL)
Google Forms (e.g., UChicago Booth)
Email submission (e.g., Vancouver School of Economics at UBC, where applicants compile materials into a single PDF)

It is crucial to carefully read and follow the specific application instructions for each position.

Hiring Process

The hiring process usually consists of two to three rounds:

Initial application submission
Screening process for selected candidates:
- An initial interview (not so common, often at HBS, Kellogg, Wharton)
- A data assignment
Final-round interviews:
- These typically follow the data assignment, though in some cases, the assignment may follow the interview.
- Some positions may offer fly-out opportunities for selected candidates (e.g., Real Estate Center at The Wharton School).

Important Notes for International Candidates

If you do not regularly use email, take extra precautions:

Check your inbox and spam folder frequently—important updates (such as a data assignment or interview invitation) arrive via email.
- Check your email at least once per day to avoid missing deadlines.
Ensure your email account can receive external (potentially foreign) messages.
- Gmail is generally recommended.

Application Materials

Application materials generally include:

CV (typically two pages)
Cover letter (typically one page)
Unofficial transcripts (for all degree programs and sometimes individual courses)
Coding/GitHub samples (optional but recommended)
Writing samples (optional but recommended)
Recommendation letters (often required later in the process)

CV

Your CV is a crucial part of your application package. Prioritize research experience, but also include relevant experience such as internships and volunteer work.

Tips for a strong CV:

If you are from an English-speaking university, ask your career services office for a CV template.
Proofread your CV with career services staff or peers.
Use concise, clear sentences that are easy to scan quickly.
Add an overview of skills (e.g., programming languages, software proficiency) at the beginning.
Perform a spell check!

Follow specific instructions if provided. Some positions require additional details, such as GPA or referees—always adhere to these guidelines.

Cover Letter

If you are from an English-speaking university, ask your career services office for a cover letter template.

Key considerations for cover letters:

Most positions require a cover letter.
Many positions specify content requirements—follow these instructions carefully! Otherwise, your application may be overlooked or dismissed.
Common requirements include:
- Programming language proficiency
- Research area preferences or preferred supervisors
- Referees
- Relevant experience

Tailoring is essential!
The minimum tailoring effort should involve:

Uploading the job post, the supervisor’s personal website/Google Scholar page, and your cover letter prototype to ChatGPT for refinement.
For your top-choice positions, invest additional effort in customization!

(Optional) Coding/GitHub Samples

You don’t have to submit coding samples if you don’t have something that meets the job’s requirements. However, be cautious—if the job requires proficiency in Stata and you lack relevant samples, consider how you will demonstrate competency elsewhere. Some candidates have received offers without submitting a coding sample, as data assignments often serve as a coding assessment.

If you submit a sample, ensure:

The code is well-formatted and has sufficient comments for readability.
Inputs and outputs (or an example dataset) are included.
If submitting a large Python project, include:
- A requirements file listing necessary packages.
- Separate scripts for functions and execution commands.

(Optional) Writing Samples

A writing sample (e.g., a term paper, honors thesis, or publication) is highly recommended.

Key points:

The main goal is not to assess technical complexity but clarity of communication in English.
Poorly written samples—especially for non-native English speakers—can lower the chance of being hired.
Ensure proper proofreading and polishing before submission.

Recommendation Letters

When are recs required?

Only 20–30% of positions** require recommendation letters at the beginning of the hiring process.
For most positions, recs are requested at the final stage as an administrative or confirmatory step.

Data Assignments

“Data assignments” can cover a wide range of tasks, including:

Data preprocessing
Econometric applications
Unstructured data processing
Machine learning applications
Mathematical proofs and theory questions
Summarizing a recent research paper

Most assignments allow completion in any programming language, but some may require or recommend using Stata or Python

Python is often used for unstructured data processing or machine learning applications (especially using large language models).
Stata is common in applied economics research. If you plan to apply for Stata-heavy roles, ensure you have access to a licensed version before applying.

Submission Guidelines:

Assignments typically require both a report and the code used for analysis.
The report should be clearly written in LaTeX with proper typesetting (e.g., no equations in the margins).
Code should be:
- Well-commented and structured for readability.
- Optimized (e.g., avoid 10 nested loops)
- If submitting a large Python project, include:
  - A requirements file listing necessary packages.
  - Separate scripts for functions/passcodes.

One goal of these assignments is to evaluate coding habits—clean, efficient, and readable code is essential!

Interviews

Initial Interview

The initial interview typically aims to:

Assess fit with the research team.
Recommend candidates to appropriate faculty or research groups.

Final-Round Interview

The last-round interview is the most critical stage of the hiring process. This interview is usually conducted:

by the supervisor you may directly work with.
over Zoom or a phone call—ensure a quiet, private space with no background noise.
on schedule—be punctual!

Preparation Tips:

Review the interviewer’s CV and research papers to understand their research themes.
Look at their “Work in Progress” section in CV to anticipate potential tasks.
Prepare to discuss your research/work experience and data assignment smoothly especially if English is not your first language.
Consider drafting a self-introduction to practice fluency and confidence.

What Determines Success?

A successful interview hinges on meeting the supervisor’s expectations for their projects. In some cases:

If a candidate is a strong match, they may interview just one person before sending an offer.
In other cases, supervisors may interview 10+ candidates before making a decision.

Follow-Up Emails?

Follow-up emails can demonstrate initiative and commitment. I personally only sent follow-up emails after the final round and for my favorite positions.

If you send a follow-up email:

Make it concise—long emails are unnecessary.
Tailor it to the position—mention specifics.
Proofread for grammar and spelling errors.

A poorly written email won’t help, so if in doubt, it’s better not to send one!

What and How Many Positions Should I Apply For?

There is a bias-variance trade-off involved in the application process:

If you only apply to positions that are a strong fit, you save effort. However, during the hiring process (data assignments or interviews), you may struggle compared to candidates who have applied to many positions and gained experience.
- For example, you might get nervous in interviews or mismanage time during a two-hour data assignment.
- This results in high variance in your performance due to limited exposure (small sample size).
If you apply to many positions, you gain experience (by making and learning from mistakes) and reduce variance in your performance.
- However, your fundamental fit for positions (bias) does not necessarily improve.
- You may also waste time on applications and end up in a role that isn’t an ideal match, often due to risk aversion.

However, if you consider yourself a non-parametric model (for example, if you don’t know exactly what you like, aren’t sure what you’re especially good at, or simply want a job), then a larger sample size reduces bias as well. Applying to more positions is recommended! Increasing the number of applications helps you explore possibilities, gain experience, and ultimately refine your preferences.

Should I Do It?

At some point, you may ask yourself: Should I really apply for a predoc?

The best approach is often to apply for a combination of predocs and PhD programs with weights of your choice.

If your goal is not a top PhD program (as per your own definition), a predoc may not be necessary.
However, applying to a range of predocs while also applying to:
- Highly ranked PhD programs (for strong candidates),
- Lower-tier PhD programs or Master’s programs (if budget allows),
- Or some combination of these
…is a strategy that lowers the risk of getting no offers at all (recall microeconomic theories of risk).

From an individual perspective, the optimal approach is:

Maximize your choice set by applying to as many programs as possible.
Evaluate your options wisely once offers come in.

The downside?

This strategy isn’t great for market equilibrium—it makes the process more competitive and exhausting for everyone.
Unfortunately, this is a prisoners’ dilemma—everyone applies for more programs to maximize their utility (myself included!).

In the end, balancing effort, risk, and personal career goals is key.

Other useful resources

Guide2EconRA: Links to many lectures, tutorials, advice/tips.
Econ RA Guide: Overview, applications, positions.
A collection of links by Quan Le: Blogs, advice, opportunities, overview of grad school/research in econ.
FDR Pre-Doctoral Training Curriculum: Resources for predocs at the Princeton Empirical Studies of Conflict Lab, including programming, statistics, GIS, causal inference, typesetting, data pre-processing and visualization, and text methods.