As part of my work for Cognito Mentoring reviewing biomedical research as a career option (not much at the link there right now), I came across an interview with biomedical researcher John Todd of Cambridge University published by 80,000 Hours.

The whole interview is interesting, but one part of it struck me as interesting and somewhat hard to believe:

John would prefer a good person in his lab to an extra £0.5mn in annual funding. Generally, there are enough grants, so finding good people is a bigger constraint than money.

Here's the full context:

Our candidate does data analysis in finance, earning over $100,000 per year. They have an Economics degree for Chicago, and an Masters in Financial Engineering from University of California, LA, and reasonable programming skills. They’re planning to do an MD then PhD.

“This guy looks great. I’d love to hire him.” (when he has his MD, or even before).

“The MD and programming/statistics combo is lethal. Top of the world. There’s major demand.”

He probably wouldn’t need to do a PhD, because of the programming. After his MD, he could just apply to a lab. He should go into genomic medicine, which is what I do. Tailored therapeutics or stratified medicine will be played out for major health and economic benefits over the next 30 years. Check out Atul Butte at Stanford. He’s the perfect profile for this guy. He could be the new Butte”

 

£0.5mn is about USD 830,000 according to current foreign exchange rates. In other words, John Todd, the interviewee, indicated that a sufficiently good researcher was worth that much. Now, the question was framed in terms of additional funding, rather than reallocation of existing funds. But assuming that the existing funding for the biomedical research lab is at least one order of magnitude greater than the amount (£0.5mn) under discussion, I don't think it matters whether we're talking of using additional funding or reallocating existing funds. Essentially, I read John Todd as saying that he'd be willing to pay £0.5mn to attract a "good person" to his lab (actually, as framed, it could be interpreted as even more: he's willing to pay an ordinary salary for the person, plus forgo £0.5mn in additional funds, to hire the person). Note: I clarified with Ben Todd, the interviewer, that the additional grants were per-year rather than one-time grants, so the relevant comparison is indeed between the grant amount and annual income.

I haven't surveyed the biomedical research community, so I'm not sure how representative John Todd's opinion here is. Andrew McMichael offers a more guarded response, suggesting that 200,000 pounds are not as good as a great researcher, but he's less sure at half a million pounds, and in any case, good researchers bring in their own grant money, so it's a false dichotomy. But I've heard that there are other people at biomedical research labs who place even higher value on hiring good people than John Todd does. So in the absence of more detailed information, I'll take John Todd's view as a representative median view of a segment of biomedical research labs.

So, question: why don't there exist high-paid positions of that sort in biomedical research for entry-level people? For comparison, one list of the top ten professors in the US lists the tenth highest paid professor as earning slightly under US$500,000. The list is probably far from complete (Douglas Knight points in the comments to Chicago having at least 5 salaries over $700K, one in the business school and four in the medical school). Glassdoor list salaries at the J. Craig Venter Institute, and the highest listed salary is for professors (about $200,000), with all other salaries near or below $100,000.

I asked a slightly more general version of the question in this blog post. I'll briefly list below the general explanations provided there, with some comments on the applicability of those to the context of biomedical research as I understand it.

1. Talent constraint because of cash constraint: I don't think this applies to biomedical research. It's not that I think they are adequately funded, but rather, they do have enough funds that there shouldn't be a great different between how they would use additional funds and how they would reallocate existing funds.
2. Genuine absence of talented people: I think that this does apply in the very short run -- it's hard for somebody to acquire a M.D. and experience with programming at short notice. But this raises a whole host of questions: why not advertise for such positions prominently, promising high pay, so that people can use the existence of such positions to make more long-term plans of what subjects to study while they're still in college?
3. Talented people would or should be willing to work for low pay: While this argument works well in the context of effective altruism (because of the altruistic orientation needed for top work), I'm not sure it works for biomedical research. I don't see biomedical research as qualitatively different from computer programming or finance in terms of how altruistic people need to be to work productively.
4. Workplace egalitarianism and morale: There may be friction in labs if some people get paid a lot more, particularly if other workers aren't convinced that the people getting paid more are really working harder. This is a problem everywhere, including in the programming world. One solution that the programming world has come up with is to offer different levels of stock compensation. Another solution is acquihires: rather than paying huge salaries to star programmers, companies buy startups that have collected a large number of star programmers under their roof, and the programmers cash in on the huge amount of money reaped through the sale. Neither of these specific solutions works in the context of nonprofit, university, or government research.
5. Irrationality of funders: Employers and their funders are reluctant to pay large amounts. Biomedical research labs are often affiliated with universities and need to use the payscales of the universities. Even those that rely on other donations may be afraid that their donors will balk if they pay huge salaries.

Of course, one possibility is that none of these explanations really matter and I'm overinterpreting offhand remarks that were not intended to be taken literally. But before jumping to that conclusion, I'd like to get a clearer sense of the dynamics at play.

The nature of the explanation could also affect the social value of going into biomedical research in the following sense: if (3), (4), or (5) are big issues, that could be an indicator that perhaps superstars aren't valued much by their peers and funders (relative to the need to make people conform to norms of taking low pay). This suggests (though it doesn't prove) that perhaps the workplace doesn't offer enough flexibility for the sort of ambitious changes that superstars may bring about, so the marginal value of superstars in practice isn't as high as it could be in principle. In other words, if your bosses don't value your work enough in practice to pay you what they say you're worth, maybe they won't give you the autonomy to actually achieve that. On a related note, this GiveWell blog post hints that many experts think that bureaucracy, paperwork, and a bias in favor of older, established scientists, all get in the way of accomplishment for young, talented researchers:

• The existing system favors researchers with strong track records, and is not good at supporting young investigators. This was the most commonly raised concern, and is mentioned in all three of our public interviews.
• The existing system favors a particular brand of research – generally incremental testing of particular hypotheses – and is less suited to supporting research that doesn’t fit into this mold. Research that doesn’t fit into this mold may include:
  • Very high-risk research representing a small chance of a big breakthrough.
  • Research that focuses on developing improved tools and techniques (for example, better microscopy or better genome sequencing), rather than on directly investigating particular hypotheses.
  • “Translational research” aiming to improve the transition between basic scientific discoveries and clinical applications, and not focused on traditionally “academic” topics (for example, research focusing on predicting drug toxicity).
• The existing system focuses on time-consuming, paperwork-heavy grant applications for individual investigators; more attention to differently structured grants and grant applications would be welcome. These could include mechanisms focused on providing small amounts of funding, along with feedback on ideas, quickly and with minimal paperwork, as well as mechanisms focused on supporting larger-scale projects that require collaboration between multiple investigators.