Going into industry

A deep dive analyzing the postdoc shortage

Welcome to The Century of Biology! Each week, I post a highlight of a cutting-edge bioRxiv preprint, an essay about biotech, or an analysis of an exciting new startup. You can subscribe for free to have the next post delivered to your inbox:

Enjoy! 🧬

---

Postdoctoral scholars (postdocs) are a fundamental segment of the current scientific workforce. These scientists work in temporary positions in academic labs after their PhD. The number of postdocs has been steadily rising over the years, and competition for positions in top labs has been fierce. This general trend has inverted over time, where the number of postdocs is now declining, and new positions are receiving substantially less applicants.

This has been a steady beat for journalists to write about, with the most recent notable addition being a Nature news article about the “paucity of postdocs.” After another cycle of discussion about the topic on Science Twitter, conversation rapidly switched to other debates, such as whether or not scientific software should require graphical interfaces.

I became curious: what is actually driving this major shift now? What is different?

As I’ve written about before, I think that terminology can be incredibly useful for clarifying and organizing ideas. This is why I analyzed the TechBio movement to unpack the combination of ideas the term has come to represent, and the meaningful contrast relative to traditional biotech. On the other hand, insufficiently granular terms can verge on uselessness. In the life sciences, anybody leaving academics is tossed into the bucket of “going into industry.”

This incredibly opaque term has led to a considerable amount of confusion. It has played a part in obfuscating the fact that what life sciences PhD graduates actually do is rapidly changing.

Here, I want to analyze what led to a massive surge in the number of postdocs in science, and why this trend is changing.

This means we’ll be looking at:

• The growth of postdocs over time

• A major shift in the postdoc landscape

Let’s jump in! 🧬

The growth of postdocs over time

Postdoc positions have intrinsic risk associated with them. They are temporary, don’t offer competitive wages, and in the majority of cases involve moving away from the institution where the PhD was granted. So why does anybody sign up for this? As the skilled scientist C.J. Battey recently framed it, “Postdocs are paid in money, experience, and hope.”

The hope is that the position is one stop in a transition to a permanent academic professorship. When postdocs first became common in 17th century Germany, the position was likely modeled after the guild system which had three tiers: journeymen (graduate students), apprentices (postdocs), and masters (professors).

As is the case with many aspects of modern life, this radically changed after World War II. The general history is laid out in detail in The Postdoc Landscape: The Invisible Scholars. The Manhattan Project changed the way that policies around scientific funding were discussed. It wasn’t about supporting intellectual pursuits—science funding had direct consequences for national security and economic growth. These ideas are best encapsulated in the report entitled “Science The Endless Frontier” by Vannevar Bush in 1945.

Expanding the scientific workforce became a major policy item. Grant money flowed into the system, and new institutions were formed in response. The number of U.S. medical schools has doubled since World War II. The number of faculty positions increased by 400% between 1945 and 1965.

Over time, a crucial imbalance emerged: to support the rapid growth, the number of new postdoc positions began to rapidly outpace the number of new faculty positions. In other words, the hope of attaining a faculty position that was used to pay postdocs became increasingly diluted. This problem is laid out clearly in The Postdoc Landscape:

What has changed is the number of nontenure track or “other” academic appointments. There were approximately 2500 nontenure track positions held in 1975, expanding to approximately 25,000 in 2006. This is a 10-fold increase during a time in which the tenured faculty population remained essentially unchanged. Whatever the reason for this dramatic shift in the composition of the more senior ranks of academic research, the effect was the creation of a bottleneck that made tenure-track positions more and more difficult to obtain. This occurred at the same time that total enrollment in biomedical graduate programs was increasing, moving from 30,000 in 1983 to 50,000 in 2008. With no increase in tenure-track positions at one end of the pipeline, and a dramatic increase in PhD production at the other, the postdoctoral position became the main pressure-relief valve in an expanding biomedical research workforce.

The plight of postdocs has become a major topic of concern. During this period of major growth, there wasn’t even a sufficient increase in “industry” positions at pharmaceutical or biotech companies to make sure that all trainees landed on their feet. The percentage of unemployed life science PhD graduates has been growing since the 1990s. In 2000, a policy report entitled “Enhancing the Postdoctoral Experience for Scientists and Engineers, by the Committee on Science, Engineering and Public Policy (COSEPUP)” was published. Science Magazine covered the report with a catchier headline: “Shadow People No More: New Report Puts Postdocs in the Light.”

Factors leading to a sudden change in the number of postdocs

Despite an increase in policy reports and the growth of grassroots unions advocating for postdoc benefits, the general structure of the pattern remained roughly the same. The number of postdoc positions continued to grow much faster than faculty positions. This trend has stopped.

Figure 1a from “Biomedical science postdocs: an end to the era of expansion”

This chart is pretty easy to interpret: after decades of sustained growth, the number of postdocs in the life sciences has started to decrease. The most recent Nature news article contains many anecdotes highlighting what this looks like in practice:

Madeline Lancaster, a neuroscientist at the University of Cambridge, UK, can relate to that. In July, she received a total of 36 applications for a postdoctoral position in her laboratory, many fewer than the couple of hundred that she originally expected. “I had been nervous that I wouldn’t be able to go through all of the applications,” she says. Those 36 didn’t lead to a single appointment. “I still have not filled the position,” she says. “There seems to be lots of competition for strong candidates.”

Alright, so how do we explain this massive shift? Where are all of the postdoc applicants going? I think the explanation contains both a stick and a carrot. The negative aspect driving this trend is the fact that this imbalance is unsustainable in the first place. Over time, if academic job prospects look worse and worse, fewer scientists will choose to uproot themselves for temporary positions with low pay.

This seems fairly gloomy for a newsletter entitled The Century of Biology, so let’s think about the carrot—because it is a fairly big one. A major driver of this trend is the fact that biotech is rapidly maturing as an industry. As I wrote about in my Techbio taxonomy, biotech is beginning to follow the general trend of tech, where young scientists have far more agency and economic freedom. The market for top scientific talent is very different than it used to be.

Vince Buffalo @vsbuffalo

The postdoc labor shortage isn't a labor shortage, it's industry out-competing academia for good talent by offering higher wages.

science.orgAs professors struggle to recruit postdocs, calls for structural change in academia intensifyWith the pandemic exacerbating long-standing workforce issues, fewer applicants are vying for open postdoc positions

The recent article starts with the following anecdote: “Peter Coveney, a chemist and computational scientist at University College London, is ready to hire a postdoctoral researcher with experience in high-level computing. The problem: he’s struggling to attract a single qualified applicant.” Let’s think about who this professor is now competing with: startups like Atomwise, Schrödinger, and dozens of others would love to hire a talented computational chemist. The same likely goes for top AI companies like DeepMind exploring problems in the natural sciences, and major pharmaceutical companies with their own internal R&D efforts.

As Vince Buffalo pointed out, with fierce competition for scientific talent, academic wages have been left in the dust. The hope of a faculty position has diminished, wages are now several multiples lower than other opportunities, and the positions are still temporary. To make this more concrete, the median postdoc salary is $47,500. It is now possible for scientists to earn up to three or four times as much outside of academics.

It isn’t just the wages, though. While the blanket term of “industry” has remained the same, the nature of these new positions is very different. These new jobs have often ported over the trappings of tech startup culture, offering perks like snacks and drinks in the office, and a more dynamic work environment. Beyond the free snacks, these companies often offer the promise of being a part of a bold scientific mission and the ability to move at a pace much faster than is possible within academic labs.

These positions are offering young scientists the opportunity to work more flexibly with other driven people, do science instead of writing grants, and sip kombucha and bring their dog to work while they’re at it. While being paid more. This is a very different experience than being a Scientist I in a lab of a massive pharma company for a 25% pay raise.

Now, it’s clearly not all roses. Many of these startups will fail, and not all scientists have the skillset that these companies are looking for. Still, an interesting indicator of this big change to “industry” is that established professors at top universities have also been jumping ship to work at new companies and institutes. This was a major point of conversation during the exodus of professors to join Altos Labs, and another good example is the amount of talent that Arcadia Science has attracted.¹

Celine Halioua @celinehalioua

when your academic collaborators ask to join Loyal full time... :)

Even at Stanford, we often hear lectures about “alternative career paths to academia” that are full of tons of low-resolution pie charts talking about where PhD graduates end up:

Source: benchfly

The reality is that most of us will end up doing jobs that didn’t even exist when these pie charts were created. We are talking about becoming computational biologists, genetic designers, automation engineers, cell-line product managers, genomics software product owners, TechBio VCs, and a ton of other extremely interesting and cutting-edge roles.

So where are things likely to go from here?

I think that we will continue to see a decrease in postdocs over time, and still see some professors move into new roles as well. Again, this is probably a good thing. It means that the nature of “industry” in the life sciences is rapidly changing and becoming much more attractive. If we can continue to accelerate biotech company creation, this trend should only become more pronounced.

This will force academia to adapt and become a more competitive option. This could potentially come in the form of better wages, but also more permanent staff positions, and an increase of PhD graduates directly landing professorships, which is already the case in fields like computer science. Lots of interesting ideas have been thrown out recently about what this could look like:

Sri Kosuri @srikosuri

@BarakRotblat @ClementYChow @ChrisPickett5 I'm no longer in academia, but from afar, I'd generally do the following in bio. 1. Cap PhD to 5 years 2. No postdocs; Most faculty hired out of PhD 3. Staff positions exist but rare; payed at ~75% prof salaries 4. No Tenure. 5. Clinical research staffed professionally

Are we going to see a graduate student exodus? Probably not. This hasn’t been the case for other fields like computer science that also have competitive job markets. As Noah Smith has argued for economics, a PhD in a field where you learn in-demand skills for biotech is a pretty great deal. You get paid to learn new skills and work on independent research projects, all while increasing your career prospects.²

One big shift that I think few people are really anticipating is that biotech undergraduates are likely to see dramatically improved prospects. There should be an increased demand for skilled scientists with bachelor’s degrees in bioengineering, computational biology, and molecular biology. Some of these scientists will also get the opportunity to found and lead their own companies. In other words, look out for groups like the Nucleate Dojo to make a big impact! If we want to truly scale the impact of biotech in The Century of Biology, we can’t remain entirely bottlenecked by PhD programs.

So let’s think about the full arc of this trend. Postdoc positions began as fairly stable apprenticeships leading to academic professorships. As science rapidly scaled after the second World War, these temporary positions got stretched and expanded, becoming an unstable pressure valve for the life sciences. As industry shifts in ways similar to tech, scientists are earning better wages and gaining more agency. This is causing the pressure valve to burst. This is probably a good thing. This flood of talented scientists into new companies has the potential to generate massive value for society, and to force academia to adapt and change.

---

Thanks for reading this analysis of the postdoc shortage. Each week, I post a highlight of a cutting-edge bioRxiv preprint, an essay about biotech, or an analysis of an exciting new startup. You can subscribe for free to have the next post delivered to your inbox:

Until next time! 🧬

1. There are likely many motivating reasons for professors to make this kind of switch. These include frustrations with the publishing industry, and the increased difficulty of attaining funding (more on this soon). Some of these new institutes are experimenting across many of these pain points.

2. It’s worth pointing out that the skills learned are often translatable into other positions outside biotech, such as data science.