top of page
  • samrodriques

Academia is an educational institution

If you are like most people, you probably think of academia as the engine of scientific discovery. You probably think of MIT, for example, as one of the centers of the world for new inventions and discoveries. And yet, if you have spent a lot of time in academic environments, you know the story is a little bit more complicated. Academia has a lot of weird flaws and hangups which I and others have written about before. I have now spent just under 5 years as a PhD student, and just over 2 years as a principal investigator in academia, and have thought about how to articulate the reality of academia to outsiders.

I think the key thing to understand is that academia in the US and Europe is really primarily an educational institution first, and a research institution second. This shouldn’t actually be surprising: after all, academia lives primarily within universities, and universities are explicitly institutions of higher learning. And yet, the educational mission of academia clashes with its research mission in very significant ways, and I think that the educational mission of academia accounts for a lot of what I and others in the metascience community think of as the shortcomings of academia. To name just three of those ways:

1. Academia has an inverted talent distribution. In most companies, you try to retain your most productive talent while pushing your least productive talent out. But because academia is educational in nature, you actually take the opposite approach: the most productive people are the people who leave the fastest, while the least productive people stay around for a while because they can’t produce the outputs they need to move on. Firing is either impossible (for PhD students) or taboo (for postdocs), because of the institution’s obligation to them as trainees. When PIs try to fight this, it leads to perverse results: everyone has heard stories of PIs who prevent their PhD students from graduating in an effort to squeeze additional productive years out of them.

This also accounts for the weirdness around academic productivity. As an example, many large labs have 50-100 people and yet publish only a handful of big papers a year, with each paper being led by only 1-2 people. Over 5 years, a lab of 50 people produces what often looks like about 20 people worth of work. What do the rest of the people do? The answer is usually “nothing.” Sometimes they don’t even show up. But they are students: they are supposed to be learning, and you don’t expel students except in extraordinary circumstances, no matter how badly they are doing.

2. Academia doesn’t do teamwork, for essentially the same reason as why we don’t let high school students take tests in groups. If you are a PhD student or a postdoc, you need to prove that you yourself are good at research. The way to do that is to produce your own independent research outputs. It doesn’t help to contribute to everyone else’s outputs.

I think that the same fundamental reasoning underlies many of the other issues that people commonly cite with regard to academia. Professors have to spend a lot of their time teaching, because, well, the core business of the universities is teaching. Universities are bad at tech transfer because, well, it’s really not their core business.

3. PIs get most of the credit for work they don't do. Of course, because PIs are teachers, it is assumed that most of the ideas come from them, and that students mostly just do what their PIs suggest. In many cases, however, this couldn't be further from the truth. PIs usually spend most of their time recruiting and fundraising, and only have vague knowledge of what is actually happening in the lab. The real technical insights come from the trainees, who get little if any recognition.

Considering this, I think the most important question to ask is: why do we rely on an educational institution to do all of the early-stage invention and discovery research in society? Shouldn’t we also have professional research institutions, staffed with professional researchers, to do early stage invention and discovery research as well? When the public imagines MIT, they probably imagine fully-fledged scientists walking around doing experiments and making discoveries; in reality, MIT is filled with people who have never done science before messing up experiments because they forget to include the right controls… again.

Now, to be clear, I’m not saying that educational institutions aren’t important — on the contrary, they are absolutely essential to build a talent base. And I’m not saying that PhD students aren’t productive — they can be spectacularly brilliant, as anyone who has worked in academia knows. Indeed, PhD students are often the ones who come up with the biggest and most important innovations, because they don't know yet what is possible and what is not possible. However, in bioengineering and biology research today, there are few if any environments that employ scientists to do early-stage invention and discovery research professionally. The Allen Institute is one; the Broad Institute is another. But, for the most part, when students graduate today, or when they finish their postdoc, their overwhelming options are either to become teachers themselves (as PIs) or to drop out of early-stage research entirely and work for a company.

This is crazy! If we think scientific discoveries are important, we should be supporting scientists to pursue them professionally. Universities should not be the only organizations doing early stage biology research in the public interest. The FROs are one new experiment in this space, and I expect great things. But we need more experiments here. Watch this space...

1,042 views2 comments

Recent Posts

See All

I often hear proposals to build new foundation models for biology. Here is the list of questions I ask. I rarely get past question 1. What is the core task that the model performs/the core thing the m

bottom of page