Hide table of contents

This is actually about two distinct roles at international organizations. If there's one thing you take away from this, it's that Research Assistant roles at policy organizations can vary a lot!

I'll abbreviate Research Assistant as RA throughout.

My Current Role at World Bank DIME

One is a job I currently hold as a RA at World Bank DIME, an impact evaluation and research unit. I assist on a research project whose ideal goal is publication in a top journal. This includes data cleaning, analysis, scripting, checking data quality, running regressions, offering suggestions in analysis calls, figuring out what the Principal Investigators want, and so forth. It’s very close to an academic “predoc” Research Assistant role that students do between undergrad / Master’s programs and PhDs.

The project revolves around development economics and causal inference, with a focus on infrastructure and structural transformation. It’s a blend of policy, research, development, impact evaluation and growth-adjacent topics.

My Previous Role at International Monetary Fund (IMF)

The other is a job I formerly held as a Research Assistant at the International Monetary Fund (IMF). I pilot-tested a software tool for better forecasting and data management. This included quality assurance testing, data migration, data entry, and scripting. My RA role was about as opposite from research as you could get and the tasks I was given was quite unconventional. The day-to-day was closer to that of a Quality Assurance Engineer or Data Engineer.

The project revolved around macroeconomics and international finance, with a focus on how to best organize data for scenario planning and technical assistance. It’s a blend of public finance, debt sustainability, fiscal policies, natural resource policies, and international macro.

Background

I currently work at World Bank DIME, an impact evaluation and research unit. I've only been here a few months, starting from July 2023. Before that, I worked at the IMF for about a year. Prior to both roles, I did a Master’s degree in International Economics and Finance at Johns Hopkins SAIS, a policy school in Washington DC. Before that, I was a Software Engineer for 4 years and before that I was teaching myself to code after a very unsuccessful post-college job search. I am strongly considering an academic career in economics and may apply to Econ PhDs next year. But I am also considering non-academic roles in development, and also PhDs in other fields like Public Policy, Statistics, and Political Economy.

I went into the Master’s program after many unsuccessful attempts to switch into development work. I had no exact plan coming in but I chose this program in particular because of:

  • It was a 1-year program, which meant less tuition and less foregone earnings.
  • International Economics sounded close enough to Development Economics that I thought I’d be learning similar stuff. (It’s very different! International Economics is more macro and finance. Development Economics is more applied micro and impact evaluation).
  • I saw my program had high placement rates in the IMF
  • I wanted to explore the "working on growth is better than global health" argument a bit more and thought, “What better way than by working on macroeconomics?”

At the time, I thought Econ PhDs didn’t influence policy much, that they were beyond my abilities, and that I wouldn’t really like it. All three turned out to be false once I started taking classes. While I was still interested in macro-finance policy, I found myself being more interested in the development research focus so I pivoted my focus towards that. In the Spring, I applied for a mix of academic and policy predocs. I landed the RA role at the IMF, found it wasn’t the best fit for me, and then got the RA role at World Bank DIME where I’m working now.

Getting the Job

That’s a quick and stylized summary of the journey. Excluded here, mostly for my own sanity, is the hundreds of job rejections I faced over the past decade. There is a tremendous amount of luck, competition, and prestige-bias in getting a role like mine. I don’t know if I’d recommend my own path to many others and if I knew how much time and energy it would take, I’m unsure if I would have embarked in the first place.

There's a lot of information online about how to get an econ RA job (such as this one) so I won't talk too much about this. But one thing I will emphasize is to apply very early. Start checking deadlines now if you're considering this for 2024 summer or fall. Even if you don’t land a job in the Fall cycle, you’ll have a better idea of what to do for the Spring hiring cycle. Interviewers evaluate you on some outside-the-classroom knowledge and judgment calls with their data tasks and interviews. Getting practice with these is useful.

Large agencies to tend have standardized procedures with early deadlines. (I missed the hiring cycles for the Federal Reserve Board and World Bank DIME deadlines in my first RA search.) 

For the IMF, you submit a single job application that then gets viewed by multiple departments. If you already reside in the DC area, you can apply year-round. If you reside elsewhere, there’s a deadline which I’m less familiar with. If a department likes your application, they’ll choose you for an interview. From there, it’s up to the department how they conduct the process. My Master’s degree is highly valued at the IMF, so I ended up with 5 different interviews in a single month. My classmates had similar stories.

The World Bank is decentralized in its hiring processes, so I don’t know anything about how hiring works outside my unit. I know of two RAs who got their jobs through connections, one who was highly recommended by a SAIS professor and another who was highly recommended by a Fed Economist. My unit in particular does have a standardized hiring process and should be putting up applications soon.

What is a (multilateral) Research Assistant anyways?

In academia, a RA is someone who works under a professor for 1-3 years before doing a PhD. In economics, this includes data cleaning and exploratory analyses in service of an academic publication. 

At policy organizations like think tanks, multilaterals, and government agencies, the role is still a time-limited role done before a PhD. But the responsibilities are broader and overlaps with “policy”. I put that in quotes because “policy” itself is a broad term but loosely speaking, you’ll usually be assisting economists on data-oriented tasks that are more “operationally relevant” rather than “publication focused”. (This wording is shamelessly stolen by a bureaucrat who prefaced they were trying to be delicate when asking me about my job.)

The upside is exposure to technocratic policy-making. Some lessons I’ve learned:

  • Success in policy is a function of how good you are at understanding a proposal’s relation to dozens of crucial considerations (both in terms of people and the connotations of a choice). I was routinely awe-struck at how some middle managers could succinctly summarize how a choice would affect all of these in a few sentences. Having broad situational awareness is very helpful for coming up with solutions that handle tradeoffs well.
    • Relationships matter a lot, even when the job is supposed to be quantitative.
    • Technical discussions are mostly about intuition and communication. Diving into details doesn’t really happen unless you’re in one-on-ones. And when it does happen, it’s usually a fun dance between intuition and precision.
    • One example of connotations is saying you think development practitioners should focus more on economic growth. This tends to be an implicit critique of the global health and metrics focus. Now it’s entirely consistent to think global health is the best way to spur economic growth! But that position is rarely expressed so you’d have to pay extra attention to how you’re communicating if this is your proposal.
  • Things move slowly but once they stick, they stick around for a long time.
  • Publications don’t have to be academically prestigious or rigorous to be influential. One example is the ad hoc reports that are tailored to individual countries. These doesn’t make for great classroom discussion but it does help a government think about, say, a budgeting decision they need to make next month based on their messy incomplete data. Another example may be broad descriptive reports of a new monitoring tool to warn people of disasters. These are also faster to write. While there still may be a peer review process of some kind, it's not the slog that journal peer review is.
    • Rachel Glennerster has a great blog post here on the difference between policy research and academic research.
  • Harmonious communication is crucial. Your career depends on gaining buy-in from a medium-sized pond that you swim in for the rest of your life. You only get to die on a hill a few times in your life. Depending on the issue, you might only get one hill to die on.
  • While I am still a “randomnista” and want more RCTs in the world, these roles did hammer home how many decisions need to be made “without evidence”. Clean statistics are hard to get, so not every question has a nice question. This is well-known in macroeconomics but it's also true in more "applied micro" questions. It's even true in easily testable interventions like cash transfers.

The downside is that the analytical challenge and skill development of your job can vary a lot. This came as an enormous surprise in my IMF job. I figured that ~25% of my time would be on a policy research project of some kind. I ended up doing little of this, though partly this was because I never took any initiative to partner with an economist which was partly because I ended up not developing a strong interest in macro-finance. The project I was hired for did have an interesting mission, but ultimately didn't develop the skills I wanted. My current role at DIME has been the opposite of that.

Overall, I would frame policy RA roles as trading off between tasks that have “operational relevance” versus a “publications focus”. If you care a lot about potentially having an ambitious academic career, you’ll probably want to get a RA role with a “publications focus”, as the PhD admissions process cares about letters from active researchers. These are much less common, as research departments are a small part of policy organizations and may not even exist in the first place. If you’re feeling undecided on academic research or leaning more towards stuff with immediate relevance, you'll probably get better career development with the faster pace and quicker deadlines in roles leaning towards “operational relevance”.

That said, switching departments within an organization after a year or two or experience is doable, as you’ll be a known quantity with less risk than an outside hire. However, this probably varies by organization and the role you want to switch into. 

Pros

  • Autonomy Besides a few scheduled meetings and the expectation that I respond to Slack messages within a few hours, I have enormous flexibility in how I approach my work. I do try to adhere to a 30-60 hour work week but I take full advantage of shuffling these hours through the week.
  • Flexibility in how much effort I put into the job Some RA jobs expect more of you than others but generally speaking, you have the choice between making it a basic 9-5 job or one where you dedicate a ton of time to it. I even know someone who intentionally puts in less time because they realized they wanted a career outside economics and policy and use that time for online classes. But understandably, they get more of the monotonous tasks at work.
  • Academically stimulating work, coworkers and internal seminars. Even at the RA level, there’s a lot of opportunities for networking, discussion, and a few occasional interesting problems. It's a nice intersection of policy-focused and research-focused topics.
  • Working on important problems. It’s weird finally being the one talking about my cool job. I’ve been hearing about other people’s jobs for years and now I get to be the one with a captive audience once in a while.
  • Exposure to what a policy economist lifestyle would be like. Knowing this is an option if I strive for the tenure-track academic job and “fail” is psychologically useful.
  • A wide range of solid exit options. This results from (1) a widely recognized name in the resume, and (2) training. Large organizations can offer standardized trainings that smaller organizations and one-on-one RA roles cannot.

Cons

  • Loss of structure. The flip side to autonomy is needing to self-organize.This creates a lot of day-to-day uncertainty on if I’m spending my time correctly and if I’m prioritizing correctly. My managers at World Bank DIME and IMF have been supportive and responsive to emails, but it remains a role with very little oversight.
    • This does serve a useful purpose. PhDs and academic careers require a lot of self-direction, so early exposure is good for skill development and discovery of whether you’d like this lifestyle.
  • Slow bureaucracy and arcane rules. Stuff moves very slow and I will occasionally get sidetracked by an infuriating process that nobody understands. And from what my colleagues told me, I've already been exposed to the top end of efficiency in international organizations at the IMF.
  • Location inflexibility if I strive for academia. Budding academics go to where they get admitted for the PhD and then where they get their job offer (assuming they get either in the first place). I have an insanely good support network in DC so relocating would be an enormous sacrifice.
  • Lower wages and benefits for a while. I make okay-to-good money right now (~$50,000 in World Bank, ~$80,000 in IMF), but it’s definitely a lot lower than what I could be making. And if / when I start my PhD, I’ll be getting a smaller stipend. When I finish the PhD, I’ll have a very good salary. But there’s definitely a period of reduced financial earnings. I've been keeping intentionally not paying off some student loans so I have some extra money to use during the low periods.
  • Impact uncertainty. People don’t get good indicators of how good they are at research until maybe the end of the PhD. Even if I turn out to be amazing at research, it will still be the case that most of my findings are useless. There’s a large iteration process to finding something novel and impactful. And even after that, there’s still replication, generalizability, scalability, and implementation. This makes research very disconnected from the end result of social impact.
  • Elitism and Lack of Diversity. I have some complicated feelings about this that I'm still figuring out. So I won't go into detail. But I'll say quickly this has overall been a negative for me ever since I started the Master's program.

Skills Developed

  • Time Management I have a manager on paper but it’s a very self-directed role with many competing priorities and many stakeholders. It is rarely the case that I have a hard deadline. Motivating myself, creating mini-deadlines, and managing the emotional swings is something I am constantly working on.
  • Managing Up Since the time of Principal Investigators is scarce, I’ve paid increasingly close attention to what people say in the few meetings I have with them. It’s also made me better at initiating 1-on-1s and raising issues in a concise way. Additionally, I'm learning to manage the trade-off between more frequent low-quality discussion versus less frequent high-quality discussion. I haven’t always gotten these decisions right and I still have a lot of room for improvement. But I’ve definitely improved.
  • Prioritization: In data work, there’s potentially hundreds of choices of what I can do at any given moment. The choice of where you should be laser-focused and where you should just get the gist of it is tricky. I can spend a ton of time verifying the data quality today. Or I can clean this other data set and get quick descriptive statistics there. This is another trade-off I have to decide on regularly.
  • Statistical Writing and Communication: I do formal presentations on analytical results to Principal Investigators on a somewhat regular basis (1-3x a month) and informal discussions on blockers more frequently. It’s been both a skill and challenge to pick out the right regression table formatting and data visualization configuration.
  • Learning to Express Disagreements: Economics has an unusual culture in how people can disagree with each other. This is strongly encouraged of juniors as well. I play-acted this style for a while over small disagreements at first while feeling like it wasn't that big of a deal. Then I actually felt like something was wrong and struggled to articulate it. Then I finally got to a point where I could sense discrepancies and express them.

Overall, I love my job. I’ve found the pros to outweigh the cons. While my job is not yet leading research, my experience feels very similar to what Kevin Kuruc described in his blog post, "Writing about my job: Economics Professor". (I also reused a lot of wording and formatting from that post.) Being able to shift around my work hours as needed is a big perk and I don’t have much of a taste for things that require money. I also find I’m developing more transferable career capital (that is, skills that would be good in other roles) than I initially thought. Mostly, that's liking the job a lot, wanting to be very good at it, and thus trying to excel at the soft skills that every job benefits from.

What about the impact?

My current philosophy to careers is building skills and experience since most people have their biggest social impact in their mid-to-late career. High personal fit means this is plausibly where I’d have the most impact even though I'm highly uncertain on the value of research (and academic research in particular). 

At the same time, I’m also realizing the skills that help me succeed in empirical research also would help me elsewhere. These projects are increasingly done in large teams where partnerships, initiative, and a general entrepreneurial spirit is useful. The world in which I'd good at this kind of research is probably also a world where I'd be good at working / founding a startup. There's also returning to software engineering doing direct work and/or earning-to-give.

But for now, I’m spending the next 1-2 years being the best at my job as I possibly can and doing some informational interviews to see what other options there are for me.

Contact Me

I am more than happy to give advice. My one request is to skim the following before reaching out:

37

0
0

Reactions

0
0

More posts like this

Comments1
Sorted by Click to highlight new comments since: Today at 9:56 AM

This is an excellent post! 

I really like seeing profiles of jobs that are closer to being "entry-level" for classic EA-flavored career tracks, to give people a better sense of what they'll be doing early on (it's common for other things, like the 80K podcast or EAG talks, to be focused on work from more senior people).

Curated and popular this week
Relevant opportunities