Comparing Across My Five Career Categories

Continuation of “I Now Have Approximately Five Career Categories”.

Earlier, I outlined approximately five things I’d consider as my post-college career: grad school, law school, market research, computer programming, and working directly for an effective altruist organization.

But which one should I pick? I don’t actually know yet. In the end, I can only pick one of the five to start with (though I probably could switch easily if I needed to.) So I thought I would attempt to compare these careers across a few different categories, see what happens, and then outline some questions I still have and my next steps.

My Dimensions of Career Comparison

After talking it over with a few people, here are the dimensions I’ve developed to compare careers. (Note that here I’m using “career” very generically to refer to any post-collegiate opportunity.)

Skills: This speaks to the quantity and quality of the personal, transferrable skills one develops by doing the career. A career high in “skills” should prepare you to be better at doing other careers down the road.

Network: This speaks to the personal contacts you would develop through the career. Could you have the opportunity to network with people who could help you a lot down the road?

“CV Points”: This speaks to how well the career would look on your CV or résumé to future employers. It’s essentially how “impressive” the career is. It can also refer to the credentials you earn through the career.

Direct Impact: This speaks to the amount you can improve the world specifically and directly by doing your career. (Technically, all these other measures are supposed to be measures of indirect impact.)

Earnings: This speaks to how much you would earn through the career. This includes salary, bonuses, and other relevant perks.

Information Value: This speaks to how much the career would teach you about the general opportunities for you in the world and improve the quality of your future career choices.

Option Value: This speaks to how well the career “keeps your options open”. A career with low option value might keep you on a track to one particular career and make switching difficult or costly.

Costs: This speaks to how monetarily expensive it would be to get a start in this career.

Risk: This speaks to how much risk you would take on in the career. Is there a chance of failure? If so, would failure be disastrous to you?

Enjoyment: This speaks to how much you personally would enjoy performing the career. Would it be a fun opportunity, an annoying one, or somewhere in between?

Replaceability: This speaks to how good you think you would be at the career, relative to other people in that same role. Are you just doing something which someone else could easily take your place and do the same?

Law School, Compared

General Idea: Go to law school for three years, then get a job in corporate law, and continue to earn lots of money and donate as much of it as I can.

Skills: 2/5. I would develop a great knowledge of the law. I also imagine I’d develop general skills relating to analysis and research. But not much else, I think.

Network: 4/5. Presumably top law schools have really good connections, but those connections might be narrow to just law.

CV Points: 4/5. Presumably a JD looks pretty good on a CV.

Direct Impact: 0/5. I don’t think my job would directly make people’s lives worse, but it’s difficult to imagine corporate law really making things go better.

Earnings: 5/5. $160K per year at minimum, several million per year at maximum.

Information Value: 2/5. I’d learn if I want to do law…

Option Value: 0/5 (lower means more restrictive). Law school basically is only profitable if you want to be a lawyer for a long time.

Costs: 0/5 (lower means more expensive). Law school is going to be my most personally expensive option by far.

Risk: 0/5 (lower means more risk). Law school is also going to be my riskiest option by far. If I don’t do well in law school, don’t get hired, or get fired early, I will have lost a lot of time and money.

Enjoyment: 3/5. I do have a genuine passion for the law and have enjoyed things like Mock Trial and Moot Court. Whether this will extend into the law school and corporate law environments remains to be seen.

Replaceability: 5/5 (lower means more replaceable). I think quite a lot of people could be just as good of a lawyer as me, or better. But very few people who would take my spot would donate as much of their income. So I’m very irreplaceable as a potential large scale donor.

Notes: I wouldn’t do law unless I got into a Top 10 law school.

Graduate School, Compared

General Idea: Go to graduate school in political science, psychology, or statistics. Presumably get a Ph.D., but maybe a masters. Do valuable research.

Skills: 5/5. I’d develop exceptionally stronger analysis and research skills. I’d also develop a better grounding in statistics and methods for studying behavioral science. I’d potentially be able to leverage this to also improve the research of existing EA orgs.

Network: 5/5. Presumably I could connect with strong faculty and maybe alumni who could help me with my career.

CV Points: 5/5. A Ph.D. is one of the more impressive things you could get.

Direct Impact: 2/5. It’s hard to say. This depends a lot on what kind of research I could do, how good the research would be, and whether it would affect anything.

Earnings: 0/5. I’d basically live off of $18K a year for a long time.

Information Value: 3/5. I’d learn more about opportunities in studying behavior and potentially opportunities to apply my field in for-profit and non-profit settings. I’d also learn if I like academia. Presumably I’d also learn of other opportunities to do research as well.

Option Value: 1/5 (lower means more restrictive). I could leave in two years with a masters. But, for the most part, there’s a lot of pressure to stay on the track.

Costs: 3/5 (lower means more expensive). It at least wouldn’t cost me anything, but it’s expensive in terms of opportunity cost.

Risk: 2/5 (lower means more risky). Doing poorly could mean wasting a lot of time and money and not getting anywhere.

Enjoyment: 5/5. I expect I would like this a lot. If I weren’t worried about impact at all, I would definitely consider this job highly.

Replaceability: 5/5 (lower means more replaceable). I think my focus on wanting research to be relevant to impact is, unfortunately, quite rare. If I were replaced, the person who replaces me would probably have less of this focus, care less about achieving meaningful results, and care less about making sure those results are put in practice.

Other: I think I might be able to get a fair amount of these benefits pertaining to research without needing to go to an academic institution, and I don’t think this fact is reflected in the above considerations.

Computer Programming, Compared

General Idea: Get a job in computer programming, earn a fair amount of money (somewhere between $40K and $100K or more) and donate as much of it as I can.

Skills: 3/5. I’d learn a lot about computer programming, which is pretty useful as a general skill. I’d also learn about general management things done in businesses.

Network: 3/5. I don’t expect networking options to be out of the ordinary for any career, but I would get the benefit of potentially being in the Bay Area, where a lot of cool EAs / rationalists live, especially ones I don’t have much contact with.

CV Points: 2/5. I don’t expect this to be out of the ordinary for any career.

Direct Impact: 1/5. I suppose software does help people in little ways and makes them more productive. Depending on which company I work for, this could be larger. But I don’t expect this to add up to nearly as much as the indirect benefits I could achieve through donations.

Earnings: 4/5. Solid earnings, but not as good as law.

Information Value: 4/5. I’d learn if I like programming. I’d also learn if I like jobs in a corporate, for-profit setting.

Option Value: 5/5 (lower means more restrictive). It would seem pretty easy to keep options open by taking this route, as I could presumably leave and do something else at any time. I could also build up savings for more flexibility.

Costs: 4/5 (lower means more expensive). It’s possible I might need to make a down payment at some sort of programming Bootcamp, but this seems relatively inexpensive both money and time-wise compared to other options.

Risk: 4/5 (lower means more risky). Given my relatively weak background in programming, I run a small risk of not making it very far in this field and wasting some time spent trying to get in. But I expect that kind of time will be useful skill development (programming training) anyway, so it won’t be much loss.

Enjoyment: 4/5. It’s hard to say. I like programming. Programming is fun. But would I like coding other people’s projects as part of a team for long periods of time? I don’t know.

Replaceability: 5/5 (lower means more replaceable). Again, I’d be donating a large portion of my income and that impact is pretty difficult to replace.

Direct EA, Compared

General Idea: Work directly for an EA organization, like Giving What We Can, 80K Hours, GiveWell, etc.

Skills: 4/5. When I did my internship, the organization focused on a lot of explicit skills development and was a good environment for expanding as a person.

Network: 4/5. EAs seem more easy to network with than usual, and there would be a fair amount of well-connected ones in any direct work opportunity.

CV Points: 1/5. I think it would be difficult to explain the value of direct work in a way that looks good on a résumé.

Direct Impact: 5/5. The whole point of direct work is to get direct impact.

Earnings: 1/5. Like programming, it could benefit the economy and benefit people who get better products, but I don’t expect this to be a large effect.

Information Value: 1/5. I already know a lot about options within EA work, but I suppose there’s opportunities to learn about more.

Option Value: 3/5 (lower means more restrictive). I could presumably leave EA work whenever I wanted, but I think it’s a bit more difficult to transition from EA work to other things because of lower salary and worse CV building.

Costs: 3/5 (lower means more costs). The relevant consideration here is having to take a pretty low salary.

Risk: 5/5 (lower means more risk). I don’t see any risk here.

Enjoyment: 5/5. I enjoyed my internship a lot and I expect actual work to be the same.

Replaceability: 3/5. I’ve had a lot of trouble assessing my replaceability in this area. I think I do write and research better than the average person they’d hire, but I don’t know if that would be true of my hypothetical replacement and, to date, there are some very talented people in the EA movement and applying to EA work.

Market Research, Compared

General Idea: Go to work at an organization doing statistics for a market research firm.

Skills: 5/5. I’d learn about market research, potentially taking some information that is relevant to helping EA organizations improve. A place like Nielsen has very specific on-boarding processes for getting college graduates into the business world and seeing lots of the company. I’d also learn more research skills.

Network: 2/5. I don’t expect networking options to be out of the ordinary for any career.

CV Points: 2/5. I don’t expect this to be out of the ordinary for any career.

Direct Impact: 0/5. Again, potentially improving the economy is the relevant factor here.

Earnings: 3/5. Entering at about $55K is not high, but also nothing to sneeze at.

Information Value: 4/5. I’d learn more about careers in statistics. I’d also learn if I like the for-profit 9-to-5 environment. Also, a place like Nielsen is very good at rotating you around the company.

Option Value: 5/5 (lower means more restrictive). I could leave at any time with little cost.

Costs: 5/5 (lower means higher costs). There aren’t any relevant costs.

Risk: 5/5 (lower means more risk). There aren’t any relevant risks.

Enjoyment: 4/5. I spent a day at Nielsen and it went pretty well. I have friends in market research who are similar to me in relevant ways and like their work. I expect to like the work.

Replaceability: 5/5. Again, because my impact will come from donating salary, I don’t expect this impact to be easily replaced.


I suppose you’re all interested to know which opportunity “scores” highest, so here it goes: Market Research (40), Programming (39), Grad School (36), Direct EA (35), Law (25).

Does this mean it’s time to just go into Market Research? Does this mean that law is a terrible idea?

Well, no. I think it’s clearly naïve to tally up the points and declare the winner to be the career with the most points. Firstly, the top three choices are close enough that they’d be pretty sensitive to small changes in my assessments, and my assessments aren’t exactly solid. Second, I’ve done nothing to assign relative weights to the criteria – some criteria clearly should count for more. And I may have even missed relevant criteria. Third, my criteria might even be wrong in what I’m emphasizing – for example, I think the current set of criteria encourages a bit too much risk aversion.

What I Still Need to Know

  • What is my comparative advantage? How I can assess that reliably?

  • How much I would I enjoy corporate law? What is the environment like?

  • What is my chance of success (i.e., making partner) in law? What is my chance of failure (i.e., getting fired)?

  • How many hours of work is law school?

  • How much would getting a Ph.D. improve my career opportunities? What about a masters? What subject should I get it in?

  • How much should I be penalizing “delayed” income (i.e., income that only comes after a few years of schooling) according to things like the haste consideration?

  • How much would I enjoy computer programming?

  • What is my chance of success in programming? What should I expect my personal income trajectory to look like?

  • Is it actual useful to volunteer for EA organizations in my spare time? If so, how much useful would it be compared to (1) working longer in a job for more pay or (2) working for EA full-time as a staff member.

Next Steps

  • Get feedback on this (and other) drafts.

  • Continue to publish material on careers.

  • Continue to reach out to relevant people and ask questions.

  • Start reaching out to populations outside the EA movement, like Denison alumni.

  • Take the LSAT and GRE.

  • Try to find relevant experiences that can help me answer some of these questions (like job shadowing, etc.)

  • Apply to graduate programs, careers in market research, careers in programming, and EA work.

  • Continue to volunteer for EA orgs.


This essay was followed up in My Conversation with Satvik Beri”, “My Careers Conversation with Holden Karnofsky”, and “My Case Study: I (Mostly) Finished Choosing Between Careers”.

_25 Sep Update: I made the following changes:

PhD Information Value 4/5 -> 3/5 Programming Networking 2/5 -> 3/5 EA Org Replaceability 2/5 -> 3/5 Market Research Skills 4/5 -> 5/5_