Two of the prominent string theorists working on ideas about holography and cosmology featured in Amanda Gefter’s new book are Tom Banks and Willy Fischler, who have a new paper out on the subject, entitled Holographic Space-time and Newton’s Law. Besides the usual sort of thing, this paper contains a rather unusual acknowledgments section (hat-tip, the Angry Physics blog):
The work of T.B. was supported in part by the Department of Energy. The work of W.F. was supported in part by the TCC and by the NSF under Grant PHY-0969020. However, the authors do not thank either of these agencies, nor their masters, for the caps placed on their summer salaries, nor for the lack of support of basic research in general.
It seems that while debating philosophical issues concerning holography and cosmology can put one at the upper end of the current academic star system pay scale, it doesn’t stop one from getting embittered that it’s not enough. The authors did revise this text a few days later to remove the complaints.
For those who don’t know what this is all about, prominent theoretical physicists (and mathematicians) in the US generally have research grants that pay them not only research expenses, but “summer salary”. Historically, the reasoning behind this was that academics needed to teach during the summer to make ends meet, so agencies like the NSF would get them more time to do research by paying them to not teach. That was long ago, in a distant era. At this point the typical sums universities pay for summer courses are so much smaller than the academic-year salaries of successful senior academics that few would consider dramatically increasing their teaching load this way to make a little extra money.
Taking the NSF as an example, the standard computation is that an academic’s salary is considered just pay for nine months, with the NSF allowing grants to pay for up to two months of summer salary. In other words, grant applications can include a request for 2/9ths of a person’s salary, to be paid as additional compensation in return for not teaching summer school. As the salaries of star academics (who are the ones most likely to get grants) have moved north of 200K/year, these additional salary amounts have gotten larger and larger, crowding out the other things grants pay for (post-doc salaries and grad-student support are the big items).
Several years ago the mathematics part of the NSF instituted a “salary cap” on these payments, limiting them to about \$25K/year. This year, in response to declining budgets, such a cap was put on payments to theoretical physicists, at \$15K/month. So, any theorist with an academic year salary of over $135K/year saw a reduction in their additional compensation (although as far as I know only two were so outraged by this that they complained in the acknowledgments sections of their papers). The report of this year’s panel on the future of particle theory in the US includes the language:
This past year, the DOE instituted caps on summer salaries, and the NSF is following suit. We agree that this is preferable to further cuts in student and postdoctoral support, but it should be noted that still lower caps will have implications for research productivity, particularly if they reach the level of junior faculty (assistant or associate professor salaries). Many researchers may have to supplement their income with further teaching or other responsibilities in the summers.
Since Banks and Fischler work at public universities, one can check for oneself that they are seriously impacted by the new caps. Fischler is at the University of Texas, Banks has positions at UC Santa Cruz and Rutgers (I have no idea how the two institutions split his salary). Some of the grant information is also publicly available, for instance the NSF grant referred to in the acknowledgment is this one. It expires soon, but was supposed to provide \$690K over three years, presumably including summer salary for Fischler, Weinberg and three others. One anomaly here is Weinberg, who at over $500K/year is likely the highest paid theorist in the US. The same people have a new grant recently awarded, for \$220K.
They have a newer version of the paper, which leaves the last bit of the acknowledgements out.
Research physicists need Porches too…
Does the AdS/CFT correspondence put a quantum limit on how small can a violin be?
Uh, do Banks and Fisher have some sort of special deal? UT Austin pays full profs 144K and UC Santa Cruz 128K, according to the AAUP, so the effect of this on either of them would be either very small or nonexistent. Of course UTA pays Weinberg who knows how much, so maybe both have some special deal. In any case it’s honestly gross, I mean if they wanted money they could have worked in industry, or become lawyers, or worked on Wall Street.
I don’t know what those AAUP numbers mean. The Texas state info I linked to has a range of about \$90-240K for tenured faculty in the UT physics department, with Weinberg way off scale at \$536K (Fischler is at \$192K).
Banks was one of the four string theorists brought in by Rutgers in 1989 with a very special deal (I wrote about Dan Friedan’s description of how it happened in my book…). At the time he would have been one of the highest paid theorists in the country and this is probably still the case (UCSC lists him at $243K, Rutgers at \$235K, but I think those are full-time salaries and he’s not full-time at both places).
In any case, the amounts lost to the salary cap should be quite significant for both of them.
Questions for anyone angry that some theorists might be paid well:
1) Does it worry you that our society rewards a lifetime of Nobel-quality academic work less richly than a few years playing for the Yankees?
2) Would we really be better off if Weinberg had gotten a JD instead of a PhD?
3) How does our culture benefit if the only way to make significant contributions to science is to sacrifice some or (quite possibly) _most_ of your potential future earnings?
4) Who exactly are we selecting to be our scientists if we insist that a scientific career necessarily involves a monastic lifestyle?
5) Is there any evidence that limiting the pool of scientists in this fashion produces better science?
6) Are we better off as a society if we take care to pay scientists poorly enough that they are underrepresented among the politically influential social classes?
The “summer salary” issue is a different one than that of whether theorists are paid well. Weinberg’s \$536K/year from UT should be compared to the fact that their football coach is getting $5.266 million/year (although he doesn’t have tenure and the team is not doing well…), so you can easily make the case that Weinberg is a bargain for UT. But would it really make sense for the NSF to pay him an additional 2/9ths of this (about \$120K), taking this money away from hiring postdocs and funding grad students, with the excuse that otherwise he wouldn’t have time for research because he would have to teach summer school? Or, maybe there should be a salary cap of some kind?
Of course I have no idea what Weinberg’s actual “summer salary” arrangement is, and he’s not known to have complained about it getting capped.
I don’t think CIP or Jeff M were talking about summer salaries. In any case, I think it’s worth keeping in mind that we’re talking about a system with a power law distribution of rewards where the most highly compensated are barely on the bottom rung of the top percentile.
Yes, top string theorists in the US are “barely on the bottom rung” of the top income percentile. Or put differently, have a great job, perfect job security, and are making about six times as much as the average citizen. Lots of ways to look at this, but you don’t have to be an Occupier to think that people in such a situation probably shouldn’t expect sympathy when they complain about their sort, or that scarce NSF funds shouldn’t be given to them proportionately to how far up they manage to climb in that top income percentile.
The string theory salaries discussed here don’t get you to that bottom rung. I was talking about Weinberg.
I don’t know where you’re getting your data. I was looking at
where, if I put in \$210,000, I get 99.0% for income percentile ranking “among all U.S. individuals with incomes”.
Household income is something different, there it looks like \$310,000 gets you to 99.0%. From what I remember, Weinberg’s wife is or was a lawyer, so likely to contribute significantly to their household income. I have no idea what the average spousal income is for top string theorists, but it could very well be at least \$100K, putting them in the top 1% for household income also.
Let’s grant that you’re right about the percent cutoff. (Although iirc wikipedia puts the cutoff at about $350K for household income, which could be slightly above the sum you’re discussing.) It doesn’t change my point: These guys may look a bit silly for griping about summer salary in particular, but the big problem here isn’t that one of them might be well off enough to afford an expensive car. It’s that we’re massively underfunding scientific research.
My original comment was an attempt to get people to realize that there’s some tension in thinking that we don’t pay scientists well enough and being angry that some of them are quite well paid. You can try to level compensation within science, but a) I suspect this doesn’t really work without at least massive changes in the grant system, and b) I think the message it sends to the rest of society — namely, that we put a limit on the value of scientific research — is a bad one.
I just don’t agree that “we don’t pay scientists well enough”, especially when you’re talking about star-academics. I don’t have time to go get precise numbers, but typical salaries for tenured faculty have for many years more than kept pace with inflation, and outpaced typical salaries for the average person. Yes, people in a few other fields (finance, professional basketball, university administration) have done even better, but I don’t see how that’s an argument for higher academic salaries, any more than it’s an argument for paying everyone in society more.
What has changed since the 60s in the academic pay scale is an increased income spread, including a huge growth in the number of very poorly paid adjuncts, with stars at the top doing better and better. I don’t see how taking limited NSF resources away from those at the bottom of the research hierarchy (grad students, post-docs) to make sure that those at the top keep climbing into the top 1 percent sends a good message to the rest of society. And academic 1 percenters whining about a summer salary cap doesn’t help the image of science either, it definitely won’t convince the average US citizen that they need to support better science funding.
To the extent it’s true the US is underfunding scientific research, it’s not because it’s not writing large enough checks to successful academics for additional compensation. It’s because it’s not doing things like spending the $10-20 billion it would take to build a higher energy collider. That’s not spending that would go to the 1 percenters of science, but spending that would go to building things and employing lots of more modestly paid people.
Sorry that this is becoming a bit of a rant, A.J. You’re welcome to respond, but other than that, from now on I’ll do my best to restrain myself and others and try and stick to the summer salary issue rather than larger issues of inequality in US society.
Well I should probably stay out of this, but I’m with Peter on this one. Academics (with tenure, at good schools, a quickly shrinking group) do quite well in relation to almost anyone else. They have of course worked very hard for that, but plenty of people work very hard, and many of them make very little money. Try teaching little kids sometime. And the adjunct issue is quickly becoming a major problem, which most of us with tenure unfortunately just ignore. Imagine spending years getting a doctorate, and then ending up teaching 6 classes at three different schools each semester, and earning (if you’re lucky) 40K for the year. My brother in law, who has tenure at a very small school, where he doesn’t make much, sometimes ends up teaching an extra class if he was worried something wouldn’t run because enrollment would be low (and keep in mind he teaches a 4-4 load). When this happens, he just gets paid adjunct rates for the class. I think nowadays it’s about 2K per class. There is no way I would ever do that, but I’m lucky, I have tenure at a pretty good state school.
Just saw PW’s last response and as a result have painfully swallowed the diatribe I was writing…suffice it to say I agree with PW (after decades of experience in academia, industry, and government).
Not much to add. I’ve already belabored my point, I don’t have any easy fixes for the 80-20 problem, and I’m happy to see the conversation turn from “why are these few guys getting tens of thousands of dollars?” to “why aren’t all these folks getting collectively billions?”.
” From what I remember, Weinberg’s wife is or was a lawyer, so likely to contribute significantly to their household income. ”
She’s a professor at the UT Law School, so her income (at least what she gets from UT) is readily available: $229,949.
I wouldn’t mind a world in which top physicists get paid better than football coaches, but given the choices that DOE and NSF had (cutting faculty summer salary or student support) its pretty tawdry for pretty well off guys like Banks to whine in their acknowledgments – the whines might be better appreciated in the science faculty lounge – and they might not get hella sympathy from humanities types either.
A tenured academic making \$200k is dramatically better off than a professional making \$200k: The academic has much greater flexibility, and is typically required to work only about twelve hours per week (most work much more, but some don’t, and none lose their job for working the bare minimum). The academic can spend a significant period of each year at conferences of his choosing, which are often just a quasi-holidays. Every few years, the academic gets an even bigger holiday in the form of a sabbatical. As the academic ages into his 50’s, 60’s and 70’s, he continues to receive his full salary – there is no concern about losing his income after being ousted by smarter, harder working youngsters, and no real concern about losing his job to layoffs or corporate bankruptcy.
Many academics in their 20’s and 30’s do work extremely hard. But they know that at any time after tenure, they can reduce this workload to whatever level they desire, and preserve their same income stream into their 80’s. For a hard working professional to aspire to the same cushy lifestyle and high degree of security, the only means is to stuff a lot of cash into the bank during their young productive years. I would therefore argue that an academic salary of \$200k is actually equivalent to a professional’s salary of around \$600k.
Don’t get me wrong – I’d certainly like lawyers and athletes to be paid a lot less. But I do feel academics tend to miss the big picture of financial life in the real world.
Very interesting discussion. I think, the whole grant system for research has corrupted American universities beyond any reasonable limit. While it is true that one needs grants to buy equipment, to pay graduate students and to get release time from teaching to do research and poor students should not pay for that, it has become excessive capitalist enterprise for universities. Administrators like it because they get free overhead money without doing anything. In turn they pay professors who get large grants large sums of money. Usually even promotion and tenure are also related to how much money you bring in. As it is well known, for faculty at small universities, it has become extremely hard to get grants in basic physics. The success rate for a proposal is usually less than 10%. In India, it is certainly not like that. I wonder how it is in Europe. Most likely they have a different system.
Paying someone on a salary for only 9/10 months of the year is a ridiculous practice. Either they are a full time employee or else they are on wages. And if they are on wages, institutions should reduce their expectations of work output as a result.
Discussion of tenure do not enter into this. Employers cannot be allowed to operate with such bizarre practices while still claiming to employ people full time.
The “9 months” salary scheme at US universities is a bit of a fiction, there for the convenience of the faculty. People are almost always paid this “9 months” salary in 12 equal installments monthly. The reason for the 9 month business is to allow people to claim not to be paid during the summer, so eligible for “summer salary” for up to three months. In some fields there probably are ways for faculty to make significant sums working at something else over the summer. Among most of the mathematicians and physicists I know though, the only viable summer job is teaching, and that doesn’t pay enough to make it worthwhile except sometimes for the most junior people. So, in practice they’re collecting the 9 month salary just like a 12 month salary, with some hopes of additional compensation of up to 2/9ths of their yearly salary via the grant system.
These are important points (which may be coming at a particularly awkward time, for academic funding in general). I’d like to know more about the issue of overheads, raised by Kashyap Vasavada. What is the argument for universities just getting a fixed percentage overhead on all (standard) grants? Do they have to account for the money?
As far as I know, universities have an overhead rate they have negotiated with the granting agency, and rules as to what things the overhead rate applies to. Summer salary is something that they get overhead on, so it’s not just individuals benefiting from the summer salary arrangement, but their institutions also.
The standard argument for overhead is that the university is providing facilities used by the faculty member to carry out their research (office, staff, lights, library, computer network, etc. etc.) so they should get part of the grant money to pay for this.
In my experience (as a department chair) overhead money goes to the department to use as they see fit. Some will let the PI choose what to do with it, some just put it in the general department fund. I assume there are schools where the money never gets to the department.
Peter, Jeff, thanks for your replies.
I had heard the arguments about offices, lights, etc., before, but it is not obvious how they are justified at a quantitative level (and there are reasons I can think of for being skeptical about the whole idea). Does anyone know what is involved in universities’ negotiations with grant agencies in setting the overhead, and in simply taking it to be a percentage of the grant?
For a recent news story about this, see
“Overhead rates are calculated based on a complex formula that takes into account an institution’s historical costs for construction, maintenance, utilities, and administration.”
Critics point out that this formula is leading to the wealthiest institutions (e.g. Harvard) getting the highest rates since they have the most expensive construction and the highest paid administrators.
There was a scandal about this back in the early 1990s, see
when it became known that the president of Stanford was charging personal expenses (yacht, wedding ceremony, flowers) as “research expenses”. I haven’t heard much about this though in the past 15-20 years.
After the Stanford scandal, universities have become much, much more careful about charging inappropriate expenses to research grants. I don’t think we’ll see something like that again any time soon.
Does it make any sense to lash out at the NSF for the inadequate overall levels of research spending? I’m sure the NSF would be happy to dole out as much funding as congress would give them.
Anyhow, if you happen to be working in string theory in the US, you should also factor in to your compensation calculations the significant probability your turn will eventually come up for the $3M Milner payoff.
Peter, thanks once more. (There’s an NSF page.) A couple things come out of this:
(a) Indirect Costs are not readily identifiable with the research project (they are not the costs of keeping the researcher’s lights on, etc.) but are part of the general cost of keeping the organization running.
(b) Indirect Costs are negotiated (or set by the granting agency). I have not (so far) seen any “complicated formula” which determines what they are.
It is completely unclear from this why Indirect Costs should be a percentage of the grants. The whole thing seems to be a general way of subsidizing organizations (mainly universities) which produce research, but without really specific objectives.
The questions of what exactly is being rewarded, and what effects these incentives have, really deserve some attention.
If you take a look at
especially Attachment IIIa, you can see a sample computation of an indirect rate.
“I wonder how it is in Europe. Most likely they have a different system.”
Not really. In some universities in Europe getting an important grant (like ERC) is sometimes the ONLY way to get a tenure-track position. Furthermore when positions are open “demonstrated ability to acquire external funding for research projects” is one of the first things people look at here. Mind that some universities don’t even pay full salary, they simply expect you to get grants. And I’m not talking about the ” “9 months” salary in 12 equal installments monthly.” that Peter was mentioning. You need to get a grant at some point otherwise you risk getting fired.