Fewer U.S. scientists are pursuing postdoc positions, new data show | Hacker News Reader

Fewer U.S. scientists are pursuing postdoc positions, new data show

science.org

47 pointsbookofjoe71 comments

I made $16,800 a year while getting my PhD. Sometimes that required me to teach, when we had funding, it was just research. Grueling hours, abusive PI, for 6 years of my life. I helped bring in over 4 million dollars in grants to my university. After about two years, I was highly skilled and founded a new core facility that I kept running for the next 4 years while finishing my PhD.

When I graduated, I had someone offer me $42,000/yr for a postdoc. At age 30. I politely declined and took an industry job for $125k instead.

We don't need H1Bs. We need better PhD student and postdoc wages.

> We don't need H1Bs. We need better PhD student and postdoc wages.

You need both.

Your PhD classmates from abroad (most likely Russia, China, Korea, and India) would have gotten an H1B after they graduated from the PhD program, and their (edit: post-) postdoc employment would have been on an H1B.

I'm assuming this was the late 1990s or early 2000s based on the numbers provided

One of my friends who did their PhD and postdoc at Stanford has been facing the J1 to H1B hellhole because employers increasingly make the applicant pay for the entire O-1 process now. They decided to screw the whole process and return to China.

I stuck my own narrative in here, perhaps I should have been clearer. One thing I see over and over again are people saying that there aren't enough highly educated Americans, hence why we have to import non-Americans. I'm trying to point out why that is - we've made it economic suicide to go into advanced science as an American.

At my university, the vast majority of the incoming PhD class in Chemistry and Biochemistry were American when I joined. When I left, it was closer to 50%, the other 50% being mostly Nigerian and Ghanan (no idea why those two countries specifically). The Americans had just been priced out.

Or maybe people who have dedicated half a decade to getting an education and/or working in the US in highly skilled, advanced knowledge positions should have a path to residency so they're not the mercy at their employer for the rest of their life. Just a thought.

Agreed, but that kind of reform ain't happening.

It's been stalled since the 2010s.

I’ve heard the path to getting a professorship in China after getting a PhD in the US was basically a sure thing, and could be used as a backup plan if the cushy US industry job path didn’t work out (or as a zero-risk plan if you wanted to move back home).

Now, it’s the opposite; people with hard-to-obtain visas get stranded in the US because the job market back home is too competitive for them.

This is great for China, but terrible for the US.

We need to bring the US education system back up to modern standards so that more Americans succeed in industry, and we also need to again make it a no-brainer for skilled workers to stay here.

Lots of Chinese (like my friends) who studied in the US would have preferred to stay in the US, but began facing hurdles around employment, visas, and background checks.

The headaches became to much so they "decided" to move back to China because at least they could earn a US level salary, or move to Chinese adjacent countries like Singapore.

My friend did research in condensed matter physics with semiconductor applications, but no lab in the US would hire him due to National Security reasons. Now he works for an organization "supposedly" part of the Chinese MIC.

This is compensated by immigrants from other countries choosing to stay in the US.

Untrue the vast majority of foreign postdocs are on J1

And until the late 2000s, if those postdocs wanted a private sector job in the US they needed an employer to sponsor their H1-B.

Also, the J-1 didn't become the primary route for postdocs until the early 2000s

I had a similar experience. My initial stipend in grad school was $13,500/yr. My first job after defending was a VAP position for $55,000, also in my early 30s. After a year of that, I jumped to an FFRDC for twice as much.

And with a BS in CS or Math what would you have made going to work for a FAANG?

How do universities have all these endowments and not pay more?

>> $16,800

This is like 5 weeks of work at minimum wage... You would have been better off with a side hustle at macdonalds.

>> We need better PhD student and postdoc wages.

Living wages would be a start, then better.

EDIT: I fat fingered the 2... that was supposed to be 25... Do note that I live in CA so macdonalds is 20 bucks an hour now here.

5 weeks of work at minimum wage is nowhere near $16,800...

"I mean, it's one banana, Michael. What could it cost? 10 dollars?"

Some people live in a bubble.

I fat fingered the 2... that was supposed to be 25...

Flip those numbers. That's a full year of minimum wage... 16,000 / 8 / 40.

You should still be able to edit your comment.

For postdocs (and often grad students), the universities aren't paying the wages, it's the research grants. In some cases, those are capped by the funding agencies. NIH in particular has rigid payscale guidelines for postdocs in the bio sciences, for example. You can find the list here [1]. A first-year postdoc makes $56k, and after seven (!) years they can make $68.5k.

Also, a general rule of thumb is that postdocs and grad students cost the same! That's because whoever is paying the grad student (either a research grant or a department teaching stipend), also has to pay their "tuition", which basically doubles the cost.

[1] https://www.niaid.nih.gov/grants-contracts/salary-cap-stipen...

In state tuition where I was rang up at about $3000/semester. It was waived by the school if you were a TA. If you were doing just research, it was paid out of the grant funds. Then I had to pay about $600 out of pocket on student fees per semester.

To make $16,800 in 5 weeks at 40 hours a week you'd need to be paid $84/hour

That's actually just about a year at Federal minimum wage.

In my case, I on-paper only worked 19 hours/week so they didn't have to pay me health insurance.

> And with a BS in CS or Math what would you have made going to work for a FAANG?

To address this bit - the problem is (and I think you're actually trying to make this point) - we need biologists. We need physicists. We need material scientists. SaaS is sucking all the oxygen out of the room because the alternatives pay so much less. By the time this thing equilibrates, there will insufficient American institutional knowledge in core sciences.

Not everybody can work at the same washed-up Silicon Valley panopticon providers, you know. Some people need to be, like, accountants and stuff.

Postdocs are such a scam role. These are PHD educated scientists who have spent 4+ years on PHD getting paid like $50k-$70k/year with basically dead end jobs where they must leave to another place to get a tenure track role since they're already considered 2nd class at the institution the postdoc is at.

> getting paid like $50k-$70k/year with basically dead end jobs

$50 - 70k would be nice, but we’ve seen lower in some states

This has been going up a bunch in the last few years thankfully, but not by enough to make it livable for people with children or in high cost of living areas.

Really

Search on YT (as an example) what's the outlook on postdocs for most people on the academic world (some notable Exceptions can be Found though)

I’m guessing one big driver is fewer professor jobs open

Yeah the competition for those professor jobs is pretty fierce. Don’t you basically have to do a postdoc to be considered for a tenure track position in the States these days (unless you are some prodigy)?

Yeah, a major reason that people do post docs is to pad their research credentials and network, but that’s pointless if there are no academic jobs open. The main problem is tenure model.

Unsurprising. The pay sucks, hours are long, and the largest industry to hire postdocs (Biomedical) is going through a fairly severe recession that has impacted lab developers in Boston and SF as well.

Edit: My analysis of the second order effect is wrong.

From the article - "Much of the shift can be attributed to employment numbers in the biological and biomedical sciences, which saw a 3% drop in the total number of postdocs from 2021 to 2022 and a 10% drop in the number who were U.S. citizens and permanent residents. It’s not clear from the recent data release what’s driving the trends. But the declines coincide with a boom in hiring at biotech companies. And data from a separate survey released by NSF last year indicated fewer U.S. Ph.D. graduates in the life sciences are pursuing postdocs than ever before, with an increasing number heading to industry"

Basically, it looks like you don't need to have a postdoc anymore to work as a researcher in the Biotech industry now.

That said, PE/VC funding has dried up in Biotech compared to a couple years ago

Paying postdocs private sector hourly salaries will probably resolve the issue, but that effort needs to come from the NSF, NIH, DHHS, and other grantmakers.

> Paying postdocs private sector hourly salaries will probably resolve the issue

What's the actual issue, though? Is there something bad that this effect is causing?

Salaries are basically limited to $50-70k a year with 50-60 hour workweeks.

If you did a PhD in Biochemistry, you could make $100-200k a year plus stock working 40-50 hours a week at most working for Pfizer in Cambridge MA instead of earning around $60k working 50-60 hours as a Postdoc at a regional Boston based program like BU or Northeastern - you have to pay the same amount for rent and groceries, but you earn half as much and work much longer hours.

At that point, the only reason to be a postdoc is to get your foot in the door if you went to a less recruited program or are from abroad.

For example, in the article - “My personal experience is that Americans were always less willing to put up with conditions that foreigners like myself had to put up with,” says McDowell, who moved from the United Kingdom to the United States for a postdoc in biomedicine.

>If you did a PhD in Biochemistry, you could make $100-200k a year plus stock working 40-50 hours a week at most working for Pfizer in Cambridge MA instead of earning around $60k working 50-60 hours as a Postdoc at a regional Boston based program like BU or Northeastern - you have to pay the same amount for rent and groceries, but you earn half as much and work much longer hours.

I was recruited pretty heavily relative to my peers, when I finished up my biochem PhD. Clearing $100k was actually a pretty tall order without a postdoc. There are jobs out there, and I managed to get one of them, but they were very competitive. With a postdoc, $150k was on the low end, but also still competitive. Basically no one is hiring at $200k in biotech, beyond director roles.

> I was recruited pretty heavily relative to my peers, when I finished up my biochem PhD

When? Your numbers don't make sense in the late 2010s to present.

Even my friends working as Sr Associate Scientists (BS only) for employers like Pfizer or Biogen are earning $100-140k base after barely 2 YoE, and younger friends and family of mine are getting $100-110k base for new grad roles now

Last year.

Take a quick look at Indeed:

https://www.indeed.com/jobs?q=structural+biologist&l=Boston%...

Amgen, Senior Scientist - $125-$150

UCB, Scientist I - $96-$125

Noblis, Molecular and Cellular Biology Specialist - $109,100 to $190,900 with 8+ years of relevant experience.

Actalent, Sr. Engineer of Internal Cell Culture Development and Manufacturing - $80-$115

$16k PhD stipend in 2024?!?

Which program did you do your PhD at!!!!

Even UCSC has a stipend of $24k, BU a stipend of $27k, and UIUC a stipend of $25k.

No offense, but it's sounding like you attended a non-target program. At that point it absolutely makes sense to switch to software.

Also, you may need some hiring mentorship. Indeed is not the best platform for roles.

Also, Assoc to Sr Assoc is just 1.5-2 years so you are absolutely going to break $100k in 2 years

I was at an R1, public land grant university. By the time I realized how insane it was, I was close enough to graduating that I just decided to stick it out.

I wasn't a great undergraduate and didn't have the background I'd have liked for the field I went into. I took a gamble on myself and it paid off, but yeah, there are some PhD programs out there that do pay a bit more. However, I am aware of a couple that actually paid less.

There are hundreds of land grant programs ranging from Tier 1 programs like UCB or UMich to regional state schools UAlaska Anchorage.

Just saying "R1, public land grant university" isn't relevant.

> I wasn't a great undergraduate and didn't have the background I'd have liked for the field I went into

That was your problem. Not "H1Bs" or "foreign" researchers.

UC Berkeley alone can graduate around 700-800 Bio and bio related undergrads (MCB, Chemistry, etc) a year. Let alone the dozens of other similar programs (UCs, UMich, UW, Big 10, Ivies, etc)

That said, with your background in structural bio you will do well in the bioinformatics and potentially even ML space, and will probably be able to realize some actual changes that you wouldn't be able to in the more traditional and hierarchical system that exists today in Life Science research.

>Just saying "R1, public land grant university" isn't relevant.

There are 146 R1 universities in the United States. There are about 35 in the western half of the United States. There are no R1 universities in Alaska. There are 105 public R1s. R1 universities are widely considered the top tier for PhDs. In my field, I had a once in a lifetime opportunity to get hands on training with a new, otherwise inaccessible technique. I did well, and like I said, was recruited heavily. This included turning down a job at the premier institute in Germany where this technique (cryo-electron microscopy) was basically developed. My PI was prestigious and well respected in the field. His previous students went on to ultimately take director positions at places like Pfizer.

Yes, I could have taken $30k/year to go a school on the California coast, but it'd have come out to about the same once you adjust for cost of living.

>That was your problem. Not "H1Bs" or "foreign" researchers. $16,800 is not significantly less than what other PhD programs I looked at were paying. Low 20s was pretty common.

From what I remember, Cornell was paying like $24k/yr at the time.

But what's the actual problem that fewer postdocs causes?

There really isn't one in my opinion. Academic postdoc positions in many fields are pitched as a stepping stone to professorship, but professor positions are very few so you end up with PHDs continuing the low-pay grind after graduation to prop up academia for a dream that doesn't come true. I think it's good more are recognizing this and just going to industry, maybe eventually this will cause problems and academia can be motivated to fix itself instead of viewing post docs as disposable. Or maybe there's no problem to fix and this is academia just getting correctly recognized as the grind option for people who are actually ok with that

The professor positions themselves aren't particularly cushy at the end of that road. Some postdocs take a look at the prospects of 60-70 hour weeks, very little time to focus on anything scientific or technical, and huge piles of administrative work, and relatively low pay and choose to industry options not out of a lack of the ability to land a professorship.

> academia can be motivated to fix itself

The fundamental issues with postdoc salaries and grantmaking lies with the US Government. It is the grantmakers like the CDC, NIH, CMS, DARPA, etc that essentially set postdoc salaries for grants.

And these agencies are completely averse to change.

Furthermore, us in industry are NOT going to fund fundamental or experimental research due to relatively low margins.

Postdocs are the single most productive members in a research group.

The PI is typically a tenured or tenure track professor, who got the job based on their research ability, but now they are relegated to an office, teaching 1-2 classes/semester, writing grants, reviewing papers for free for for-profit journals, reviewing grants for NIH/NSF, etc.

Grad students are not well motivated, not skilled, need lots of training, take classes, need to work on a thesis, and often times also teach. So they are low output and high time investment.

Postdocs require little-to-no time investment as they already have most prerequisite skills, have no other obligations beyond research, and sometimes even bring their own funding. They are basically professional researchers and are by far and away the most productive members of any given research group.

Much less experimental and fundamental research.

Companies aren't going to invest in experimental research in the biotech space as the costs are so high.

Most compounds end up getting validated in some way shape or form by researchers, then productionized by the private sector.

This has a pipeline effect on development in the industry.

For example, mRNA Vaccine research was purely experimental research done by postdocs and junior researchers at UPenn in the 1990s, before being commercialized by BioNtech and Moderna in the mid-2000s.

Lack of research. Most academic research is done by postdocs through grants from e.g. NSF. NSF and other grantors cap postdoc pay at below living wage in expensive cities like Boston. (The professor and university who hire the postdoc have very little flexibility to pay them more.)

Add that it's a temp job with no security and fewer people want to sign up. So a lot of cutting-edge research can't get done.

But is that because a lot of value can be found in these same people producing in industry at the moment? I wonder if there could be a cycle of sometimes discovery being more valuable, because biotech is fairly commoditised, and a cycle of implementation sometimes being more valuable, because we haven't made the most of discoveries.

Eric Weinstein has been complaining about this for a long time. [1]

Probably a result of the monopoly + monoculture of the higher education system. Universities love taking advantage of their postdoc and phd candidates. Thankfully, industry sometimes compensates the ones who justify higher market value.

[1] Listen to almost any bog podcast he has done.

I think it's particularly bad in the US. When I was looking to get a PhD here, I emailed a ton of professors asking if it was possible to do the PhD part time so I could keep my job, and most of them just didn't get back to me, and the ones that did acted like what I suggested was a completely ridiculous proposition.

On the other hand, when I looked at UK universities, every professor I emailed got back to me, and while not all of them were amenable to part time remote PhDs, all of them did know someone who was and was happy to forward my request over and they actually did!

I'm not 100% sure why the UK seems to be a little less horrible on this front, because eventually I was able to get into a a PhD program with the express understanding that it's part time, and I have a regular job.

It always gives me a bad taste when large institutions problematisize their recruitment efforts. the is equally the case for nurses, teaches and the military.

This is a very conscience decision by the democracy that has funnelled money elsewhere. if nothing else, you can be surprised that these effects have been so slow.

power to the people is, in this case, taking the better deal. work in the industry for double the salary (and still get to publish). People will move back to academia when the deal is on par (eg. more freedom to research the subject you find interesting?)

it's like individuals are choosing based on market incentives or something.

I used to want to be an astrophysicist or something similar in pursuit of understanding high energy/quantum/cosmic phenomenon and generally trying to do my small part in furthering our understanding of spacetime and the universe. However, The horror stories of getting a PhD and general complaints regarding the nature of academic research are a significant part of what lead me to change course and take an early exit from academia, and generally dissuaded me from that pursuit. The PhD's (Students and graduates) I talked to were all saying the same thing about terrible pay, terrible treatment, terrible grueling grinding hours, begging for funding that could fall through and sink your project no matter your merits or efforts or attempts to prevent it, the publish or perish culture, etc etc etc

At first I thought they were just jaded or cynical and I was just talking to the wrong people, or perhaps the local uni was just atypically bad in someway, but as I branched out and asked more people from more institutions I kept hearing the same thing pretty much universally. I heard it then (over a decade ago) and I hear it still now.

I'm sure there are exceptions, but of the many I've talked to, few feel it was worth it, and every single one agree it was a grueling grind regardless. And I fully understand that few things worth doing come easy, and that discipline and persistence can be valuable "rewards" in their own right (started my own business, more than once), but on the same token, that doesn't mean things need be so difficult or detrimental, nor does it mean that the process can't be improved or modified in some way so that it doesn't so mercilessly beat people down.

Guess that's a long winded way of me saying: yeah, no shit. It can and should be better, and am not the least bit surprised it's on the decline. How bad does it have to get before there's a serious, genuine effort to fix it? Additionally, if there is such an effort underway, what is it and how can I contribute?

> According to the latest data release, 62,750 postdocs were employed at U.S. institutions in the fall of 2022, a 1% drop compared with the year before. But the trend diverges sharply by citizenship. From 2021 to 2022, the number of U.S. citizens and permanent residents working as postdocs dropped from 29,755 to 27,289.

> The 8% change is the largest year-to-year percentage-wise drop in the history of the survey, which has collected data since 1980. Meanwhile, the number of postdocs with temporary visas increased by 6%, from 33,573 to 35,461, about the same number as in 2020.

It doesn't seem like a completely drastic drop but definitely not insignificant.

Obviously the solution is to pay them competitively.

But universities seem to reserve the good salaries for superfluous administrative staff and other dead weight business school positions.

But then again, who gets to write the budget and set salaries.

Here's an excellent video about this problem from a first-person perspective.

https://www.youtube.com/watch?v=mExlPihH3jk

No shit. Who wants to still be treated like a grad student after defending. I didn't even consider looking for a post doc, and I graduated more than 10 years ago. Stayed on as a staff scientist for about five years then went to industry.

Completely unsurprising. Why get a postdoc position with wages near minimum wage, long working hours, high stress all for a small chance at a vanishing number of Tenure Track jobs when Industry pays literally 10x more in some cases with vastly better work life balance and work that is often frankly more impactful and interesting. Someone working as a postdoc sends a significant anti-signal in my opinion.

I am a bit weird, because I have a regular software engineering job while also doing a part time PhD. I love my advisors, and I think the work I'm doing is pretty cool, but when my advisors asked if I was considering doing a postdoc after I finish the the PhD, I had to figure out a polite way to tell them "absolutely not".

I know not everything is about money, but fundamentally a regular industry job is just going to pay categorically better than nearly anything in academia. It's one thing for me to do the PhD for nights and weekends while taking home a yuppie software engineering salary because I want to formalize my "learning for fun", it's another for me to dedicate myself full-time to continuing the research for peanuts. It honestly kind of feels like a scam from universities; by definition, nearly anyone with a technical PhD is qualified for decent-paying work in industry, so the fact that universities pay so little for post-docs is borderline-insulting.

Bit of a personal anecdote on this:

I've not done a post-doc, but I used to work for a relatively prestigious private university in NYC as a "research scientist" [1]. I put "research scientist" in quotes, because that was my official title, but my work largely boiled down to software engineering, mostly JavaScript and a bit of Haskell. In 2015, I was being paid $90,000/year, which may seem like a lot of money until you realize that it's NYC; taxes and cost of living are pretty high. The university refused to pay any of my moving expenses, and they barely paid any of my health insurance (basically the bare minimum they were allowed to with the ACA).

For perspective, most entry-level software engineering jobs in NYC at the time were paying at least $120,000 even at startups, and big organizations were paying more like $130,000, and I wasn't entry level. I had been a senior-level engineer at my previous job in Dallas.

Upon moving to NYC, and starting my job, and having a fiance (now wife) that wasn't able to work at the time due to some visa issues, I was having a lot of trouble making ends meet, and I was having trouble paying rent [2] and medical bills. After three straight months of having to borrow money from my parents to cover my bills, I talked to the management team for the project I was working on, trying to explain to them that I was having a lot of trouble holding my head above water, I had maxed out my credit card moving to NYC and that they were paying well below market rates, so I would really just like an advance to at least avoid paying huge amounts of interest since they hadn't covered any relocation expenses.

Instead, I got a long, indignant response explaining how important the work they were doing was, and then told me that I should be grateful for even having the job, and that I was already being paid higher than [other person on team] who had been there longer, so I'll just need to figure things out myself. It really rubbed me the wrong way, because they acted like I was living this life of luxury when I felt like I was barely surviving, and they treated me like I was fucking Gordon Gecko for simply Googling "market rate for software engineers in NYC".

I eventually got laid off from that job (really fired, long story), and sort of swore an oath that I would never rely on academia as my primary source of income ever again. I've still done work for universites since then (I was an adjunct lecturer for a public university for about a year), but I always made it clear to them that I will be doing work for a private employer as my primary income.

A part of me loves academia, but they can be the embodiment of the "this is more than a job, it's a family" employer red flag.

[1] It's not hard to find my work history, so feel free to figure out which university with some clever searching, but I politely ask that you don't post it here.

[2] Before you give me a lecture, I wasn't trying to rent a luxury apartment in Greenwich Village, it was a pretty crappy apartment in Washington Heights infested with mold and mice. It cost me $2100/month in 2015.

Well yeah they're a scam.

In the replication crisis era it's a good thing I suppose.

https://en.wikipedia.org/wiki/Replication_crisis

I'm not sure why this is being down voted.

The classic argument in favor of taking a (potentially major) pay cut to work in academia as opposed to industry was always that academia is an environment for true learning and doing proper research uncorrupted by profit-seeking motives of corporations. Essentially, you make the classic trade-off between income and integrity.

In recent years we've seen that there's not a whole lot of integrity in academic research. Most of it is politics, grant hunting and working to inflate someone's ego; integrity of the actual science being done be damned.

At least with the current state of industrial research your results have to pan out enough to convince the market. Maybe LLMs are a overhyped, but they at least work enough to impress the people using them. Meta might never create the metaverse Zuckerberg dreams of, but at least people doing research on the hardware will hold in their hand lighter, higher performing, and truly state of the art VR headsets.

I'm guessing that it's getting down voted because it implies that the solution to bad science is to stop doing science. It's unfortunate if the poster was trying to say something different.

There are other solutions to "fixing" science: you demand more openly published data, look more critically at peer review, reexamine the tenure process, incentivize open source development, etc. It's possible that the poster was suggesting some of these things, but that wasn't clear.

Industrial research is great if you have a product you can sell, but there's a huge swath of science research where that doesn't work: you can't monetize JWST images or LHC data. Zuckerberg, whatever you think of him, isn't going to dump all his money into something that won't pay out at all in the near future, and will be shared by everyone when it does.

The researchers who would replicate studies would be postdocs themselves.

..and having a lot of them didn't improve the situation.

There was a shortage of them already. The shortage only got worse

Postdoc system must be abolished. It is meant for individuals seeking some skills outside their discipline and has crept under the carpet to a toxic level. A PhD degree essentially states that the individual is strong enough to conduct independent research. It is oxymoronic that additional training is mandated after that.

Shortage should lead to a better pay and(or) a higher vacancy. However those are not the case.

The wages are set by the US Government (who tend to be the largest grantmakers).

> a higher vacancy

This absolutely IS the case. The US used to fix this through immigration, but Chinese researchers are de facto blacklisted now and Indian researchers are increasingly deciding to return to India instead of waiting 15-20 years for naturalization.

Isn’t the vacancy rate quite high?

Quick google search says no?

Reply on news.ycombinator.com