# 10% of US Postdocs negotiate pay raise.

by on August 18, 2010

UC postdoc protest, from prouaw.org

On Friday we told you that the 6,500 postdocs of the University of California just ratified their first union contract, after 18 months of negotiation.  The five-year contract requires that all postdocs be paid at least $37,740/yr, which is$340 above the current minimum at UC.  (Postdocs with 5 yr experience must make at least $48K.) These salaries are floors; one-third of UC postdocs are already paid more than$47K/yr.  1.5–3% pay raises are built in, as are ceilings on health insurance premium increases.  In return, the postdocs agreed not to strike.

This is a big deal because the ten UC campuses employ 10% of all postdocs in the United States. The 6,500 postdocs, spanning ten University of California campuses, unionized in 2008 as the Postdoctoral Researchers Organize, affiliated with the United Auto Workers (UAW).  UC grad students are also unionized. Any UC astronomers want to share their perspectives from over there?

Now, context. As Kelle pointed out, astronomy postdocs…

Median salaries, from "Doctors without Orders" from Sigma Xi.

get higher salaries than biology and chemistry postdocs.  That’s true, though it’s a dubious distinction to be less badly underpaid than ridiculously underpaid colleagues.  The US median salary for postdocs, in 2004 dollars, is $34,700. At their reported 51 hr/week, that’s$14.90 per hour — about what Harvard janitors make.  (Reference: the report Doctors without Orders from Sigma Xi.)

I find the low bio/chem postdoc salaries particularly curious, because those fields have a direct connection to industry, where Ph.D. chemists 6–9 yrs from their B.S. degree have a median salary of $90K/yr in 2008 dollars. I got interested in the subject of low postdoc pay because of an embarrassing experience at Caltech. At a mentoring event, a grad student asked a room of postdocs and grad students, “So, how much do postdocs make?” After an awkward silence, I tried to get the discussion started by volunteering a stipend of$45-50K for grants postdocs, and $56K for Hubble/Spitzer/Chandra fellows.* There was an audible gasp in the room. Then came a chorus of numbers in the 30-40K range. At an institute with a$1.4B endowment, $270M/yr in grants, and$106K/yr average salary for assistant professors.

A recent article, “The Real Science Gap” (Miller-McCune, via The AstroDyke) posits a root cause for these low salaries:  that there are far too many junior science PhDs relative to the number of permanent jobs in science.  The article suggests the “scientist shortage” is a myth that keeps the supply of young scientists high, and wages low.  Definitely worth a read.

While I’m impressed that the Decadal Survey acknowledged the reality of this overproduction, I’m dismayed that it suggests we not worry because “training in astronomical research appears to be well matched in practice to much broader career opportunities.”  This proposition seems highly dubious to me, especially presented without evidence or a control sample.  I suspect that a selection effect is at work:  people who go to astro gradschool are smart, so they’ll probably succeed in industry.  It’s not at all clear to me that ~6–8 yrs in grad school gives any advantage, especially not compared to, say, a 2 yr MS program plus on the job experience.  (Cosmic Variance and commenters suggest that 2 yr of grad school is a sweet spot that maximizes subsequent salary in industry.)  The Decadal Survey’s assertion strikes me as well-intentioned but self-serving wishful thinking.  While grad school teaches valuable problem solving and other analytic skills, so do a range of other industry experiences and (shorter) professional degrees.

So.  Are there too many postdocs for the number of jobs?  Is there a way to increase funding for STEM without creating a “lost generation” of junior researchers with poor job prospects in academia?  Did the UC postdocs’ unionizing really achieve much?  Is such organizing repeatable at other institutions?  Why are chemistry postdocs paid poorly, while new PhD chemists in industry are highly paid?  Do the NIH funding guidelines make sense?  Why don’t fellowships come with cost-of-living allowances?

*  Those were the numbers a few years ago.  2011 Hubble fellowship stipends will be $63K/yr. { 11 comments… read them below or add one } 1 Casey August 18, 2010 at 12:25 pm This is an interesting question. In fact, it pops up in other places, such as the oft-quoted Freakanomics: http://freakonomicsbook.com/freakonomics/chapter-excerpts/chapter-3/ They ask the same question about drug dealers (no affront intended, honestly). But it is the same kind of problem: “So if crack dealing is the most dangerous job in America, and if the salary is only$3.30 an hour, why on earth would anyone take such a job?”

The answer, which may be relevant to us postdocs, is that drug dealers accept low pay on the promise of future success. Those that survive the low-level work are sent up the ladder to very lucrative work. This is similar to the process for postdocs moving on to tenure track. Now a better question is whether this is fair in an age when postdoc terms are getting longer..

2 Jane Rigby August 18, 2010 at 12:43 pm

Yes, I thought that the Economics of Drug Dealing was, by far, the best chapter in Freakanomics. But I hadn’t thought about it in the context of postdoc prospects — thank you, Casey, for making that connection!

If I remember, only a very small fraction of street-level dealers advanced to management — a smaller fraction than postdocs who advance to faculty. Of course, the high-management payoff in the drug trade is higher than than in academia.

I wonder, are postdocs and street-level dealers equally well informed about their chances of advancement? Do they do the cost-benefit analysis rationally?

3 Chris August 18, 2010 at 1:08 pm

I was also dismayed to see that the decadal survey did not take the problem of overproduction of astronomers so seriously. For the last 15-20 years the argument has been that a graduate education in astronomy is very good preparation for industry. Most of my friends from grad school who left after obtaining a Masters degree have been successful either finding jobs in industry or teaching at community colleges. My friends who obtained a Ph.D. and had one or more postdocs have had more trouble getting into industry (in economically lean times). I would like to see some data which actually supports this assertion of industry success before we spread this to another generation of astronomy students.

I am glad to see postdocs organizing. In my postdoc, at the prestigious Center for Astrophysics, we had difficulties just getting paid in a timely manner. At one point as I was expecting my first child and trying to get the nursery ready our pay was two weeks late. The CfA justified these delays by saying we were all basically independent contractors.

4 Ben August 18, 2010 at 1:55 pm

When I was a UC postdoc I participated in a few postdoc-organizing discussions, before the unionizing effort was seriously under way. At that time the problems of, for example, biology postdocs, were more serious than just low salary. A lot of people were on NIH training grants that paid of order 28K without benefits. While astro postdocs were full UC employees, the people on NIH grants were in a non-employee status that didn’t give them UC health insurance, and even things like parking permits and using the gym became a PITA. Since then NIH adopted some standards that tried to persuade PIs who had an NIH postdoc/indentured servant to kick in for benefits, etc, but there were still problems.

My understanding is that astronomy postdoc salaries were dragged upward when the Hubble fellowships started – other fellowships had to increase salaries so as not to look bad by comparison. I think the reverse happens in other fields – low salaries for e.g. the NIH postdocs kept standards low. I think higher industry salaries do not push postdoc salaries up because there is non-liquidity in the job market – jumping between academia and industry is a significant mental barrier and people don’t tend to move back and forth.

While worrying about postdoc salary scales, also consider what junior faculty salaries are – it’s probably not as big a jump from Hubble fellow to assistant prof at public university as many people think. One of the reasons we don’t get paid more is that the overall society is not interested in paying for academia.

5 KT August 18, 2010 at 9:29 pm

Jane, Casey, Chris, Ben, all–

Thanks for this great post! I was also shocked on Friday at the Astro2010 Report to hear L.H. and D.E. reporting on the Decadal solution to career problems (good question, Travis M! –the response was quite unsatisfactory, though — Check out http://www.tvworldwide.com/events/nas/100813/, starting at time 61:25).

The Decadal Survey definitely misunderstood our concerns about over-production of PhDs, because their solution was to open up our non-research-track options (after we get a PhD in Astronomical Research!). I did not hear them address the issue of PhD Astronomers who want research careers (permanent positions) not having good odds of getting such a career.

In fact, I was at the Town Hall last June (2009) in Pasadena at AAS/Summer and again, the same professors were there, misunderstanding my concerns viz. offering the mitigation of more off-shoot career options. But many PhD Astronomers want astronomy research career options, permanent positions, and those are in short supply and that is the concern that has yet to be addressed.

As Jane mentions, why would I get my PhD to join a career I could have started 6 years ago with a MS in engineering?

6 Casey August 18, 2010 at 10:02 pm

Perhaps the solution is to abolish tenure…?

7 Christian August 19, 2010 at 1:55 am

Quote: ” I got interested in the subject of low postdoc pay because of an embarrassing experience at Caltech. At a mentoring event, a grad student asked a room of postdocs and grad students, “So, how much do postdocs make?” After an awkward silence, I tried to get the discussion started by volunteering a stipend of $45-50K for grants postdocs, and$56K for Hubble/Spitzer/Chandra fellows.* There was an audible gasp in the room. Then came a chorus of numbers in the 30-40K range. At an institute with a $1.4B endowment,$270M/yr in grants, and $106K/yr average salary for assistant professors.” (sorry for the poor quoting style; I couldn;t fine a style guide on how to quote ‘right’) Jane, here is the thing: As you surely know, universities pay for very few postdocs directly. Those that get paid from endowment are prize fellows who _are_ well paid in any area. The majority of postdocs at any university, however, are paid through grants from NSF, NIH, DOE etc. The budgets of these grants and the budgeting policies/traditions (!) of the granting agencies are what is really controlling general postdoc salaries. So that’s what is getting me: How is the unionization of postdocs in the UC system going to help them? UC isn’t really paying them anyway. If I have X$ in my NSF grant and the union is telling me I have to pay my postdoc Y>X $, what am I to do? Is the UC system (that’s broke anyway!) going to chip in the money so that I can keep my postdoc? Unlikely, right? Why not try to work with / fight (or at least create awareness at) the funding agencies instead? Reply 8 Peter August 19, 2010 at 10:13 am I am surprised at how fast the Hubble/NASA-based fellowship stipends are going up (~$10k/yr over the last 5 years), and not sure continuing these increases will help in the long run. Soon many astronomers may find they’ll need to take a pay cut to take a faculty position.

9 Ben August 19, 2010 at 1:56 pm

If you have X in your grant and the postdoc salary+benefits is Y>X, then you can’t hire a postdoc. (Actually, what’s likely is that you can only afford to pay the postdoc for 2 years instead of 3, etc.) People will have to budget in their grants for livable postdoc salary+benefits rather than cramming the number down.

A real issue is with postdocs where the money comes directly from the funding agency, like the NIH training grants I mentioned above, and the funding agency was impoverishing the postdocs. Because UC employs such a large percentage of the entire US postdoc corps, this puts pressure on the funding agencies to pay enough to support a UC postdoc. This is likely to be more effective than a loose bunch of postdocs trying to pressure funding agencies directly.

10 Amanda August 19, 2010 at 5:52 pm

To me, the bigger picture issue here is that a large portion of the science in astronomy is being done by people with short term (2-3 years) positions. With that kind of work force, it is difficult to have the continuity that projects like those described in the decadal survey need. Would astronomy research productivity be even greater with a fewer number of longer term research positions rather than more postdocs?