Case study: choosing between working at effective altruist organisations, earning to give, and graduate school

See an update on how Peter’s getting on here


Back in May 2013, I realized I would be graduating in a year and wondered a lot about what I should pick for my first career. The questions I had at the time were:

  1. Should I aim to work in an effective altruist organisation, go to graduate school, or should I earn to give?

  2. Where should I look for employment if I want to earn to give — law, market research, or programming?

I spent a little time considering other options (finance and consulting careers), but the bulk of my time was spent comparing EA org employment, grad school, and the three earning to give careers.

Lessons learned

  • Direct work in EA is promising, but there are limited employment opportunities and a generally strong base of talent to draw from that makes replaceability an issue.

  • Graduate school also seems promising, but programs with high direct impact seem limited in employment opportunities.

  • It’s important to consider factors about the career other than salary when doing earning to give. Law was my best earning to give opportunity at first glance, given that it had the highest salary of the options I was willing to consider. But when I looked more deeply at non-salary factors, it became my worst option.

  • Market research and computer programming are my most promising paths and I should consider both further. They allow good salary potential while offering many other benefits.

  • Publishing my ongoing thought process was valuable in ways I couldn’t even imagine at the time, creating the opportunity to meet people I couldn’t have met otherwise.

  • Spending time directly in Oxford was also incredibly valuable in meeting with people that could help me think through my decision process.

  • An analytically-minded person can train in programming quickly enough to seriously consider programming as a career path. While I started with intermediate computer programming knowledge in irrelevant computer languages, it took me about 150 hours of training over 20 weeks to know enough to interview competently. I don’t know if this is a unique case, though.

The research

My methodology

First, I compiled a list of every high impact career I could conceive of. I then narrowed the list down to the careers I thought were worth pursuing further, based on personal factors, and published that information online for comments.

After a bit of time following up on the comments I received from publishing, I then deliberately set out, crafted a set of questions I wanted to know about various opportunities, and interviewed people about them. I ended up having informational interviews with 31 different people, including five people from 80,000 Hours.

At the same time, I did an internship at Giving What We Can in Oxford, England. This taught me a lot about the opportunities for direct work and put me in a position to meet with 80,000 Hours and a lot of other effective altruists.

Based on the information I gathered from my interviews, I compared my five options on the basis of 11 different criteria, as suggested by 80,000 Hours. I also pursued all five options to the best of my ability until I had the confidence to eliminate options. Eventually, I came to a final decision I could be confident in.

Why these five categories, and not others?

There were many careers I didn’t consider much — such as finance, consulting, and medicine — which seem to be popular careers of those who use 80,000 Hours’s research. I didn’t choose any of these three careers mainly because I felt underqualified and felt that they would be enough of a poor fit for me enjoyment-wise that I wouldn’t do well enough in the career. I’m currently open to changing my mind on finance due to the extremely high earning potential, however.

Some opportunities for direct work, like fundraising (along the lines of Effective Fundraising’s model), I didn’t investigate further because the options were too limited. Other opportunities, like working in nonprofit organisations doing work on the ground, like Innovations for Poverty Action, didn’t seem to match my skills or interests very well and didn’t seem enjoyable to me.

I also didn’t feel comfortable and confident enough to be an entrepreneur. While I’m open to potentially creating a start-up someday, I wanted a lot more knowledge of how organisations work and what the market is like.

What was my first guess?

My first guess, made on May 2013, was that the best thing for me to do would be to do direct work (seek employment in the Centre for Effective Altruism) and enter law school and do earning to give if the first plan didn’t work.

What factors did I use to compare my careers?

After talking with Jess Whittlestone and Nick Beckstead from 80,000 Hours, I adopted the following 9 criteria for evaluating my career:

  • Direct Impact: The actions of the career itself directly improve the world. (Technically, all these other measures are supposed to be measures of indirect impact.)

  • Earnings: The career has a large compensation package.

  • Skills: The career will develop me in valuable, personal, transferable skills that are useful for furthering my career or otherwise being better at effective altruism.

  • Connections: The career will connect me with people that can help me a lot down the road.

  • Credentials: The career will look good on my resume, improve my impressiveness for employers, and qualify me for better jobs.

  • Information Value: The career teaches me more about what career opportunities there are and improves the quality of my future career choices.

  • Flexibility: The career keeps my options open in terms of allowing me to switch to another career easily.

  • Enjoyment: I would enjoy the career.

  • Replaceability: I’m not doing something that someone else easily would do just as well.

Should I do direct work?

Direct work seemed good on direct impact, skills, connections, flexibility, and enjoyment, but weak on earnings, credentials, information value, and replaceability. I spent a lot of time trying to assess my own replaceability in direct work before a conversation with Satvik Beri outlined an incredibly simple procedure — just ask EA organisations whether they’d rather have you or money. I was in a good position to do this, having just interned for an EA organisation.

While I interned, I felt like I produced above average work, but I didn’t think my work was significantly better and I hadn’t adjusted for the fact that people tend to think more highly of themselves than is justified. When I take these two aspects into account, replaceability becomes a concern.

However, a larger concern for me that emerged after talking to people was that employment in an EA organisation is actually rather difficult to come by, given a limited number of positions and for the time-being has high standards in recruiting due to being a start-up and needing lots of particular skills. This was especially a concern for me given that I am an American citizen but most desire employment in the UK-based Centre for Effective Altruism, as CEA faces some difficulty in hiring Americans full-time.

A final concern of mine was the possibility of coming to a major disagreement with CEA about what the best approach is for making the world a better place. While this was not a dominant consideration, the flexibility and independence provided by an earning to give career seems like a benefit to me. This seemed especially true if I picked a career that still affords me training in skills relevant to EA direct work and/or gives me a significant amount of time to volunteer for EA causes, as I could transition to direct work if I felt like my skills were better served there than earning. Generally, I felt like it is easier to transition from earning-to-give to direct work than it is to transition from direct work to earning-to-give.

Therefore, I decided to no longer seek full-time direct work for the time-being. However, I still believe that some kinds of direct work (personal EA side-projects) are higher value to me than spending the time earning a high salary, so I want to choose a career that allows me the flexibility to also have a significant side-project / volunteering component.

Should I go to graduate school?

If I chose this option, I would pursue a Ph.D. in either political science or psychology. I had considered a Ph.D. in economics, but felt underqualified and felt like I wouldn’t enjoy it enough to make it worthwhile. I was searching for programs that would allow me to study the psychology of civic engagement, volunteerism, and/or donating. For example, I became very attracted to the work of Elizabeth Dunn at the University of British Columbia and considered trying to apply there (see “Flipping the Philanthropy Switch” over at The Life You Can Save for more on this kind of work).

This option seemed to be strong on connections, credentials, enjoyment, replaceability, and skills (though skills is debated as the skills learned may only be useful for academic work). However, it seemed weak on flexibility/risk and earnings. It also would be weak on direct impact depending on what program I chose.

Ultimately, I decided against graduate school and while I did take the GRE (the US admissions test for grad schools), I chose not to move forward on any applications. While I found research to be high-value, my intuition was that I would have more influence in making key research happen by funding research rather than carrying it out. Also, I felt like I could do a large majority of the research I want to do (currently, mostly stuff related to studying vegetarianism) without taking on the costs of graduate school. Lastly, I felt like there were enough people already working in the graduate programs I had considered that I wouldn’t add much (one could think of this as having no “room for more people”, as an analogy to “room for more funding” in GiveWell top charities).

While I think it’s unlikely, I wouldn’t rule out trying to apply to graduate schools in late 2014 or late 2015, however.

How good is law for earning to give?

With graduate school and direct work out of the way, I focused up more on earning to give. The first I focused on was law as I saw that it had the highest salary potential of the options I was considering. If I pursued law, I would aim to enter corporate law, which pays at $160K per year and then rises by ~$10K/year for ten years before you have an opportunity to make partner and earn $500K/year to millions a year depending on the firm. Law seemed strong on salary, connections, replaceability, and credentials, but had drawbacks in skills, direct impact, information value, and enjoyment. It is also utterly terrible in flexibility, considering I would have to pay money into three years of law school before I would earn anything.

I was pretty keen on law for several months as a top choice, but I ultimately stopped valuing it as much as I started to place more emphasis on personal enjoyment, skills development, and flexibility. Eventually, having to commit three years and $50-$150K to law school with no promise that I would end up in a situation that worked well for me was a pretty clear deal-breaker.

Another aspect of law that I came to really not like was the expected long hours of the job, which would prevent me from being able to spend significant amounts of time on EA side projects. Since I expect that I will be able to do side projects that produce more EA value than earning to give, this also made me not like law as much relative to the more 40-60 hour a week careers.

I still took the LSAT (the US entrance exam for law schools) anyway, because I had prepared for it a lot, and thought it would be nice to have the score. I ended up doing well on it, but not well enough that my admittance to the top 10 law schools would be likely. (Common wisdom suggests that one needs to be in a top 10 law school in order to have a good shot at a corporate law job.) So, unless I retake the LSAT, this option was pretty much over whether I wanted it to be or not. Given that I had already decided before seeing my LSAT score that I didn’t want to do law, I wasn’t that upset.

How good is market research for earning to give?

The general idea is that I’d seek employment doing statistics for a market research firm. This seemed strong on skills, flexibility, enjoyment, replaceability, and information value and not particularly weak on much, though salary, credentials, and connections seem weaker than other options, relatively.

Salaries are lower in market research, with starting salaries in the $40K-$60K range and less indication of upward movement in salaries with promotion. But one potential advantage I hadn’t considered much is that few EAs, if any, had gone into market research, so it could be a ripe new field to explore for EA potential, especially if marketing principles could then be applied directly to the EA movement. Therefore, I’m still seeking employment in market research.

How good is programming for earning to give?

I also considered computer programming as an earning to give career, which is quite popular in the effective altruism movement. Like market research, this also seemed strong on skills, flexibility, enjoyment, replaceability, and information value and not particularly weak on much. Salaries are good, starting in the $50-$80K range and rising upward to $120-$200K, I think.

The interesting thing about computer programming for me is that, while I was a long-time hobbyist programmer, I had never really pursued programming seriously, I didn’t have a tremendous amount of experience, and I lacked a computer science degree. So I needed to spend a significant amount of time learning programming. Luckily, the market for programming jobs is good and a lot of companies are willing to focus on “on-the-job learning” if you can demonstrate your general aptitude and your ability to learn quickly.

This sounds good to me, so I’m still seeking employment in programming.

What was my “final” decision?

Ultimately, I decided to go with a model where I would mix both direct work and earning to give by seeking full-time employment for 40-60 hours a week, but then spending significant portion of my “free time” volunteering and doing EA side projects. I liked this model the most because I expect the side projects to be worth more than additional salary and it should teach me more about the comparison between earning to give and direct work.

Six months after starting my plan to learn about jobs (and one month after sending applications), I received an offer of an 8-week “trial hire” internship from AvantCredit, a web development start-up that offers online loans to people with near-prime credit ratings. I got this offer on the basis of experience in both web development and statistics. I accepted the offer and I’m excited to be employed there.

AvantCredit offers a mixture of interesting challenges in both programming and statistics, which are two skill-bases of mine I want to improve. The office culture sounds pretty awesome. Additionally, I’ll gain experience about what it’s like to be in a start-up and learn more about finance and financial risk, in case I want to look more in finance. I also think there’s a strong chance I might end up staying with AvantCredit long-term.

Longer-term, I’m still looking at careers in market research and computer programming. I’m also looking into jobs that offer particularly high donation matching, which is basically like a raise for effective altruists, though I haven’t explored this much in depth.

What did I learn?

I really liked the methodology I ended up using. Initially, I was worried that writing up information about my career choice publicly could lead to employers viewing my application more negatively and hurting my chances at receiving offers. I still can’t rule this out, but I do think the benefits of publicly publishing have more than made up for it. Publishing the work lead to people commenting on it who became great contacts that I would not have made any other way. One of these contacts ended up telling me about AvantCredit and giving me an internal referral. I don’t think I would have received the offer without his help, let alone known to apply in the first place.

I also think it was a really good idea to take an internship at Giving What We Can, which I actually considered not doing at one point (because I was worried about not seeing my family for most of the summer). There really is no substitute for joining an EA hub directly and meeting tons of EAs in person.

Lastly, I learned that you apparently can train yourself to learn enough programming to get a job offer at a moderately quick pace, though I may have got lucky with a variety of things. I spent 10 hours a week on programming over a period of about 10 weeks, and about 5 hours a week on programming for the 10 weeks before that, prior to receiving the AvantCredit offer. In doing so, I was able to significantly increase my knowledge of jQuery, Ruby on Rails, and R. I hope to up this commitment to make sure I’m even more ready when I get there and when looking at other companies. Notably, I received an offer from AvantCredit also based on knowledge of statistics and from having an internal referral, so I don’t know if I will end up having enough experience to interview well at other companies.

What would I have done differently, if I could start over?

I wish I was able to more quickly identify that grad school and law school wouldn’t pan out and therefore not “waste” the time studying for and doing the GRE and LSAT. There definitely were warning signs I could have seen respectively if I had spent more time comparing the careers to my other options. Also, I could have seen the lack of good graduate school programs and the problem with having unrealistic expectations about the kind of LSAT scores I could achieve.

On the other hand, I’d be worried that heuristics that would have saved me time via the GRE and the LSAT would also have cut out computer programming, however, as I was also concerned that training in programming quickly was too much of a dream and I wouldn’t know enough fast enough, which turned out to be wrong.

While much less significant of an issue, I also wish I had refined my decision criteria a bit more. The nine factor model I used in this essay actually was an 11-factor model that included “risk” and “cost”, but both of these considerations ended up being unnecessary. Additionally, my thinking in some of the other categories could be refined. I like the five-factor model offered in “A Comparison of Medical Research and Earning to Give” and probably would adopt that if making broad comparisons again in the future.

What impact did 80,000 Hours have in this decision process?

Overall, the existence of 80,000 Hours completely changed my career, but the choice to pursue additional coaching did not further assist in making my decision.

Before I looked into 80,000 Hours, I was pretty intent on pursuing a Ph.D. in political science for a long time, and thought that would be a great career path for me. (I still do think it would be very enjoyable; I just think it’s less impactful than my other options.) Despite engagement with effective altruism, I hadn’t considered that maybe this career path was wasn’t the best route until I had a conversation with William MacAskill. I think 80,000 Hours has done a lot to put the idea of career choice into the general EA zeitgeist. I am nearly certain I would not have spent any time considering the ethics of career choice if 80,000 Hours did not exist.

The actual career advice session was not useful to me, but I suspect this is because I was an unusual case — (a) I had already come in to the advice session having done a significant amount of research on my own that I would have done otherwise; (b) I was already very familiar with 80,000 Hours’s content having read all of their research material, completed a workshop with them, and worked in their office; and (c) my particular decision most hinged on very particular details about salaries and career capital that no one had done comprehensive research on yet.

On the other hand, the workshop I had completed for 80,000 Hours was impactful because it moved me away from thinking solely about the impact of my career and more about thinking about the skills and career capital I would develop in my career, which I agree is a better model for thinking about careers. Additionally, the conversations I had with people working for 80,000 Hours — Jess Whittlestone, Roman Duda, William Macaskill, Benjamin Todd, Nick Beckstead, and Pablo Stafforini — as a part of my personal research were all helpful in thinking more about my career choice. Therefore, I strongly suspect that a careers advising session with 80,000 Hours would be beneficial to many people thinking about their careers.

Ultimately, I did heavily rely on research done by 80,000 Hours to inform my decision — both in assisting my initial selection of high-impact options and in comparing them. The existence of 80,000 Hours moved me away from pursuing a Ph.D. program in political science toward earning to give in web development. 80,000 Hours very much impacted my decision.

See an update on how Peter’s getting on here