Are they wrong? US research funding

It seems to be part of the job description of anyone in the sciences to periodically complain that scientific research funding is insufficient, with the situation going from bad to worse. For some recent examples, see this from Bruce Alberts, the Editor-in-Chief of Science, and this endorsement from Professor Matt Strassler.

In the contrarian spirit of this blog, I want to suggest that the situation is actually quite a bit more complicated, and the story of research funding is not completely a one-sided one of the oppression and impoverishment of scientists. Also in the spirit of this blog, I want to avoid topics I don’t know much about, which in this case includes the vast majority of scientific research and how it is funded, especially outside the US. The biggest component of R&D funding in the US is the military, and I have no idea what this money is going towards and whether it is being well-spent. I’ve also heard that there are increasingly vast sums being spent by the US on classified research, not necessarily accounted for and showing up in obvious places in the budget, but I have not idea whether this is even true or what the size of this is. While ignorant about what military R&D spending is going to, I confess to a general prejudice that it seems to me to be huge and if I knew more I’d probably be strongly in favor of there being less of it.

The next biggest component of R&D spending is biomedical, and again, I’m woefully ignorant. Unlike spending money to find better ways to kill people, biomedical research is inherently something worthwhile, so more of it undoubtedly is better. But whether it is now being spent well, or whether taking away from some other priority to spend more in this area would be a good idea, I haven’t a clue.

On overall US federal spending levels, Alberts compares a level of .87% of GDP in 2013 to a level of 1.25% of GDP in 1985. He’s getting his data from here, but those numbers do tell a more complicated story. Measured in constant (2012) dollars, non-defense R&D/year went from $32 billion in 1985 to a maximum of $67 billion in 2004, and has been relatively flat since then, with $64 billion projected for 2013. Defense R&D went from $65 billion in 1984 to a maximum of $90.5 billion in 2008, has dropped significantly in recent years to $76 billion for 2013. Another set of overall numbers from the same source are for the NSF budget, which went from $4.6 billion in 1998 to $7.25 billion in 2013.

For a while on this blog I used to try and keep track of the US budget situation and periodically report on it, at least the numbers I could find and understand for math and physics. The most important thing to say about the situation of recent years is that the US federal budget process has completely broken down. Budgets have gone from being passed late to never, with government spending now allocated by some baffling system of continuing resolutions and last-minute “cliffs”. There appears to be nothing anymore like a sensible process for making future plans and sticking to them. Those responsible for managing research facilities are not only in the dark about how much money they’ll have to spend over the next few years but sometimes don’t know how much they’ll have to spend next month or next week. No matter what you think spending priorities should be, trying to run organizations this way is completely nuts and a disgrace to the country.

Getting close to fields I do know something about, here are some other numbers (also 2012 dollars): NSF yearly spending on math and physical sciences has gone from $924 million in 1998 to $1,323 million in 2013. DOE Office of Science has gone from $3.3 billion in 1997 (including $895 million for HEP) to $4.5 billion in 2013 (including $764 million for HEP).

Theoretical physics is very much small potatoes on the scale of science funding in general. For FY2012 the DOE spent $67 million on theoretical and computational physics, the NSF $13.6 million (+6 million for Physics Frontier Centers), up from $11.7 million (+6.3 for Physics Frontier Centers) in FY2008 (real, not inflation adjusted dollars). Increasingly, large amounts of funding are coming from the private sector. The Simons Foundation spent $40 million on grants for math and physics in 2011. The Perimeter Institute has gotten $150 million or so from Mike Lazaridis over the years, and the Templeton Foundation has recently provided $2 million to Perimeter, after $8 million to FQXi, and millions more in other grants such as $2 million for the philosophy of cosmology. Yuri Milner has in the past few months handed out about $37 million in checks to physicists, with one goal that of supporting their research.

The overall pattern seems to be that science in general has not been doing that badly, although HEP funding in the US has been cut significantly, as the US lost leadership in HEP to CERN with the LHC becoming the focus of attention. US experimental HEP faces huge challenges in the future, but they have more to do with the SSC debacle of 20 years ago and the lack of a compelling technological way forward to higher energies than with general federal science budget cutting. Funding for theory from conventional government sources has been fairly flat, with new sources of private funding starting to have a major impact.

As for the work conditions of US academics, Matt sees the situation as:

Whereas before the year 2000 it was easy for U.S. universities to attract the best in the world to teach and do research at their institutions, and to train the next generation of American scientists, the brain drain since that time has been awful.

On the other hand, my own experience at Columbia (a wealthy private institution) and in mathematics has been that the post-2000 period has been one where the US in general and Columbia in particular have done very well in competing for talent. While the middle class in the US has been in decline, top-flight US academics have seen significant salary increases. The AMS compiles yearly numbers for salaries (see here), which show the mean academic-year salary for a mathematics full-professor to be $127,674 at large public research universities, $148,074 at large private research universities. Back in 1999 the numbers were $85,571 (public) and $95,977 (private). Comparing to median US incomes, the ratio has increased from 4.26 to 4.73 during this time in the public university case, 4.77 to 5.49 in the private university case. At the top of the profession, average salaries for full professors at Harvard (in all fields) were $122,100 in 1998-9 (6.07 times US median), $198,400 in 2011-12 (7.36 times US median). The general pattern is that of the rest of US society, with the rich getting richer, and staying very much competitive for talent with the rest of the world.

As usual, informed and on-topic comments are welcome. If you just want to rant though about the evils of government spending, or go on about how in a just society scientists would get lots more money, please do it somewhere else.

This entry was posted in Uncategorized. Bookmark the permalink.

49 Responses to Are they wrong? US research funding

  1. X says:

    Firstly, the very topmost universities such as the Ivies will be the last to depopulate. Secondly, much of the impoverishment of the university system is being done at the state level and is destroying state public university systems. Thirdly, the defunding of the system is done by removing the bottom tiers of the system: low pay and few hires for assistant professors, sending young people to the eternal damnation of sub-professor-level migrant academic and lecturer jobs. This does not cause pay for full professors to decline. (Learn more about such trends by looking at “wage stickiness” in the econ lit.)

  2. Peter Woit says:


    I don’t think what is going on is “wage stickiness”, since what’s happening is not salaries at the top of the scale going down more slowly, but salaries at the top of the scale going up. You see this throughout the US economy. In the universities, another place you see it is with administrative salaries which have been growing even faster than star professor salaries (Columbia’s president makes nearly $2 million/year). This isn’t only the Ivies: Berkeley may have massive budget problems, but it still came up with a big increase in its chancellor’s salary for their new chancellor.

    The phenomenon of exploitation by universities of the labor pool that they themselves flooded (by dramatically increasing the size of Ph.D. programs) is by now a very old one, and the destruction of wages at the bottom of the scale has coincided with large pay increases at the top of the scale.

  3. X says:

    Hi Peter,

    I wonder if in a twisted sense, those crazy salaries for top bureaucrats might be worth it. Since a main job of top bureaucrats is to use political savvy and shenanigans to funnel money from various sources into the university, a university president who gets paid 2M$/yr but pulls down 100M$/yr in donations is a good investment. Of course, then you have to wonder about the marginal costs and whether a 500k$/yr president would pull down the same donations anyway, but the optimal salary seems non-obvious to me.

  4. Yatima says:

    “I confess to a general prejudice that it seems to me to be huge and if I knew more I’d probably be strongly in favor of there being less of it.”

    Woah I got the impression of a witness having to make a statement in front of some Popular Court on his current view on whether the governement is, in his opinion, indeed working for the Greater Good of The People. Or not.

    Luckily the flames of passion are being fanned in the next paragraph.

  5. Peter Woit says:


    That’s certainly the argument being made to justify huge increases at the top of the pay scale, in all industries. The guy running Lehman was using it to justify his $22 million payday right up to the moment they went bankrupt.

  6. srp says:

    This post opens up a huge can of worms, many of which you probably will not want crawling around on this nice, clean science blog. For example, your distaste for military R&D (“better ways to kill people”) could easily lead off into a complex discussion about pacifism, whether it’s better to kill people using worse ways, US foreign policy, etc. I’m betting that comments about this aspect of the post would get snipped as being “off-topic” even though, as the lawyers would say, you opened the door to these issues. For the record, though, that kind of statement about military R&D does not increase the credibility of the rest of the post. End of on-topic though off-topic comment.

    In general, you are right that the US government and private donors have not been chary in supporting science of all kinds over the last decade. The increasing difficulty in winning grants is more the result of the well-known “sorcerer’s apprentice” effect caused by the requirement in most grants that new PhDs be trained. (And of course those PhD students provide the cheap workforce for the senior investigators.) The more grants given, the more PhDs spawned later, and the more competition for whatever level of funding has then been appropriated. With this system, no level of spending will ever recreate a situation where a high percentage of grant applications are funded.

    In my opinion, the reason why this training policy is maintained is that the government and science bureaucracy prefers to have a large and lower-paid STEM workforce. They are not entirely irrational in this regard; President Obama’s former economic advisor Austen Goolsbee has published econometric studies showing that increases in government R&D spending tend to mostly raise the wages and incomes of researchers rather than producing greater quantities of research. The training requirement partly neutralizes this effect, which would otherwise be even more severe. Lots of cheaper researchers helps the government get more research done for less money, gross of training costs; the interesting question is whether using initially low-skill PhD students as the front-line workforce in labs is really the most efficient way to get the research done.

  7. Peter Woit says:


    My disclaimer that I have little idea what military R&D spending goes for was meant quite sincerely and intended to provide appropriate background for any of my comments on the subject. But, yes, I’d rather not host or moderate a “military spending, good or bad?” argument here.

  8. Peter Woit says:

    As for the “efficiency” argument for the desirability of a massive low-wage work force, surely the most efficient work-force structure would be slave labor, with a small overclass keeping armies of needed scientific researchers chained in labor camps, kept in line by brutal physical methods, and fed just enough to keep them working long hours optimally.

    While this would get the most work done at the cheapest price, perhaps there are other desirable goals in a society more important than extracting from human beings as much as possible at the lowest possible cost.

  9. Felipe Pait says:

    I am not very knowledgeable but I share your view of military research. I worked on military-funded R&D for a short while, at what I judge to be a top-notch operation, and my impression was that the funds were not spent very efficiently – the US was not getting a lot of national security per dollar spent. Classified research appears to be less effective than the less secretive components. There is a lot of waste on questionable projects. Rerouting military funds to science research or to engineering development would, for the admittedly little I know, be good for the country.

    On the other hand, perhaps we should not be comparing the effectiveness of military research with science research. National security by brute force is massively expensive, and painful. To the extent that military research can make defense operations even a little more effective or deadly, or prevent wars, it is justifiable in its own terms.

  10. Bob Levine says:


    There’s a kind of no-win situation that seems to have come about in the sciences. Yes, the induction of young researchers with bleak prospects into these fields is from one angle cynical, callous and exploitive. But on the other hand, we know—the relatively few Max Planck examples notwithstanding—that it’s the younger investigators who are likely to make the major breakthroughs. A field which is aging demographically is in danger of withering; older established scientists are unlikely to make the kind of huge conceptual leaps that the likes of Einstein, Feynman, Dyson and von Neumann achieved. If you don’t actively recruit the best of the youngest, your field has a bleak future indeed. So the idea of having Ph.D. training as a component of grants isn’t bad in itself; the difficulty is at the other end. A lot of the blame has to go to the universities themselves, and university administrations in particular, for the gradual elimination of a permanent, reasonably well-compensated professoriate as its core faculty, with jobs waiting up the line for those new Ph.D.s, in favor the disgraceful ‘Kelly prof’, migrant-labor model that academic managers have been doing their best to create.

  11. lun says:

    the administration overhead is definitely a large part of the problem.

    The processes you describe is a good experimental refutation that the “market”, in this case, the academic job market, is a “statistical system of many equal players” rather than a power structure of generally hyerarchical power relations. The problem is, in the latter case a decrease of funding just makes the system more corrupt instead of leaner and higher average quality.

    So, paradoxically, the author of “Am I wrong?” is correct, but the “anonymous scientist” commenting on page 2 (a very thoughtful comment, starting with “Yes you are wrong”) is correct as well. More funding might make the system less corrupt and bloated, and less funding will worsen the “politics”.

  12. Peter Woit says:

    Bob Levine,

    One should keep in mind that typically it’s not faceless administrators that want to expand Ph.D. programs, but department faculty. And when faculty go to the administration to request lower teaching loads (so they can do more research), and are told, “OK, you can teach less and we’ll hire some adjuncts to cover the courses”, do you think faculty say no to this?

  13. Bob Levine says:


    “when faculty go to the administration to request lower teaching loads (so they can do more research), and are told, “OK, you can teach less and we’ll hire some adjuncts to cover the courses”, do you think faculty say no to this?”

    There’s no denying that senior faculty are often complicit in perpetuating the system as it’s developed during the past 30 years. But what I’ve seen at my own university is a disappearance of the tenure-track jobs themselves, as faculty retire and their lines are frozen, or cancelled, or merged together. And *that* initiative is coming from administrative management policy, whose creators have figured out that if you can strip as many benefits as possible from the majority of your academic work force, you’ll save enormous portions of your budget. At public institutions such as the one I teach at, this makes senior administrators look good to the legislators who supply our funding (although way, way less than they used to), and, maybe even more important, it increases the value of those administrators in the eyes of other institutions looking to hire cost-cutting, efficient managers in what now seems to be the private sector business model that many (most?) academic institutions seem anxious to follow. The jobs that are sacrificed to make our provosts, chancellors and executive deans look look like CEO material and keep their bosses in the Statehouse happy are the jobs that our newly trained Ph.D.s aren’t ever going to get….

  14. Peter Orland says:

    Peter and Bob Levine,

    Both of you are right. As more adjuncts replace faculty, less research is done, resulting in fewer grants, hence less need to hire full-time faculty, and so on… There are some enlightened deans/provosts/presidents who see that flushing faculty lines down the toilet leads to for-profit U. of Phoenix-style degree mills.

    I sometimes vainly try to explain to non-academic friends that research-oriented faculty are the best people to learn from (that was my experience anyway. Give me a scowling teacher, who would rather be in the lab or busy doing calculations. He or she can answer questions, not BS through them ).

  15. Ray says:

    Once for amusement I compiled a list of quotations of scientists over the last 50 years. They always said, funding is being cut and the job market is getting worse. I guess whining is a permanent feature of being an academic.

  16. Lowell says:

    Perhaps the production of PhD students is too large and not determined by their quality.

    At a place where I once taught, the number of entering graduate students was determined by the number of graduate student TA’s needed to teach undergraduate labs and grade papers. It was not determined by the number of outstanding students that could be attracted to enter the program.

  17. dark says:

    What role do you think string theory and current LHC results has had in attracting funding, research, academic positions, and graduate student applications/interest?

  18. Z says:

    David Helfand had a similarly austere editorial on the state of science funding last month, as it relates to Astronomy:

    and broaches the idea that, perhaps, there is some birth control necessary at an early stage.

  19. Chris Oakley says:

    Sorry – but I really do not accept that training a large number of Ph.D.’s when there is little possibility of them getting an academic job is a valid one. Only about one in five of my peer group (Oxford theoretical particle physics circa 1984) got a commensurate academic job, and I was not one of them. But even if I had known that I was not going to be able to continue I would still gladly have entered the Ph.D. (or D.Phil. as they call it in Oxford) program simply for the privilege of doing it. In the U.K. there is a also a persuasive demographic reason for doing more degrees. When I took my degree, 5% of eligible students went to university. Now it is more like 40%. Having a masters or Ph.D. is thus a way of distinguishing one from the crowd when applying for (non-academic) jobs.

  20. chimpanzee says:

    “Better to have Brain Drain, rather than Brain-in-the-Drain”
    — Indian prof, Re: India talent studying abroad in USA

    There is a Reverse effect in progress (as pointed out by M. Strassler), talent is going away from USA. I went to famous HS (University High on UIUC campus)

    [ produced 3 Nobelists (Philip Anderson/Physics..Higgs mechanism, Hamilton Smith/ hired by Dr Craig Venter for JCVI “1st organism designed by computer”, James Tobin/Economics. Other notable alumni are Bill Bardeen/Fermilab (son of John Bardeen, Nobelist), Shamit Kachru/Stanford ]

    which has the similar formula of high-end Academia: recruit top-students with top-faculty..GOOD THINGS will happen. 1 of my HS classmates (Dr Dave Albin, NREL/National Renewable Energy Lab, leading PV/Photovoltaic researcher) told me something alarming:

    “I’m losing 2 of my staff to a Chinese solar company, who offerred them $300K/year salary”

    This is American STEM talent being drawn away to foreign shores. China leads in Wind/Solar, the Obama’s so-called “next Sputnik moment”. The Solyndra failure (DoE, Dr Stephen Chu/Stanford & Dr Stephen Koonin/Caltech/Physics/Provost & BP Chief Scientist)..500 million wasted, since they couldn’t compete with market flooded with low-cost Chinese solar cells. This is a sign of “too little too late”. That OMB (Office Management & Budget) email that got leaked

    “This [ Solyndra ] loan is not ready for prime time”

    indicates they were RUSHING THINGS. Another sign of “too little too late”. A123 (Battery Tech) failed..bought by Chinese company (250 million loan wasted)

    The KEY to Economic Recovery is a STEM based stimulus package. Not only Alternative Energy (due to politicization of Global Warming hype/hoax), but Biotech, Nano Tech, Computer Tech. Apple Computer is the pinnacle of Tech which found mass-market monetization (#1 market capitalized company, exceeding “old Oil” Exxon Mobil)

    It’s a well known fact that Curiosity/Pure Research has long-term mass-market potential. The CD/DVD was a mass-market entertainment device, that was rooted in Theoretical Physics

    “Einstein & Botha realized that photons in the same state..[ laser ]”
    — Murray Gell-Mann, “From Student to Scientist” PBS Documentary

    I view HEP (High Energy Physics) & Space Exploration as “push it to the limit” challenge-frontiers, which spinoff mass-market Technology. Silicon Valley was a product of

    1) Physics research
    Bardeen/Shockley/Brattain, invention of Transistor at Bell Labs

    2) NASA Apollo program
    pressure to miniaturize electronics created the Integrated Circuit (chip)

    It’s clear that these 2 sectors need to be exploited (“seed funding” by US Govt), for further “Emerging Economy”..Economic Growth/Recovery. As part of America 2.0, NASA 2.0 (really hit hard by Sequestration), Science 2.0

    “I have never tried, in even one single little instance, to help cultivate the cultivated classes. I was not equipped for it either by native gifts or training. And I never had any ambition in that direction, but always hunted for bigger game–the masses. I have seldom deliberately tried to instruct them, but I have done my best to entertain them, for they can get instruction elsewhere.”
    – Mark Twain, a Biography

    Leadership in STEM needs to take the above to heart, & figure out a short/medium/long term strategy to leverage their KNOWLEDGE for mass-market (“masses”) monetization.

    “I’m interested in KNOWLEDGE, not Product”
    — Dr Misha Mahowald, Calltech PhD CNS (Computational Neural Sciences), Neuromorphic Engineer, “Discovering Women episode”

    [ “Physics of Computation” field by Dr Carver Mead (VLSI pioneer), kinda involved Feynman ]

    The above illustrates the tunnel-vision 1-sided nature of Science, it is IMPERATIVE to keep an eye on Product (“monetization”). In order to demonstrate to Funding Agencies the RoI (Return on Investment).

    “You have to sell yourself”
    — S. Hossenfelder, advice by her professor

    The whole Fiscal Crisis in Washington DC could be a “blessing in disguise”, to FORCE a Science 2.0 recovery that has more Perspective (“integrated solution”)

    Time for Vision, Leadership, Execution to “kick in”.

  21. jackj says:

    I found this article interesting, but still feel that the references to Perimeter Institute are somewhat misleading. Perimeter is in Canada, which is a separate country with its own funding situation.

  22. Peter Woit says:


    That’s true, and I suppose should have been pointed out. But it’s also true that Canada and the US are quite closely linked economies and cultures, with Waterloo just over the border. Matt is raising the alarm about “our ability to defend ourselves, especially since some of the most active spending on science is being done by countries that are hostile or potentially hostile to the free world.” It doesn’t seem to be Ontario he’s concerned about, but who knows…

  23. Peter Shor says:

    When Matt says “active spending on science is being done by countries that are hostile or potentially hostile to the free world”, by “countries” does he mean China? Certainly science funding has decreased substantially in the countries of the former U.S.S.R., and none of the Islamic countries have large research budgets.

  24. Peter Woit says:

    Peter Shor,
    You’ll have to ask him, I’ve no idea what he’s talking about.

    Looking at the Milner prize recipients, almost all the senior ones are in the US, but it is true that the young “New Horizons in Physics” prize recipients aren’t. They are in Canada, Switzerland and Israel, but I don’t know if Matt considers those places to be hostile to the free world. The fact that they’re not in the US may just reflect this being kind of a string theory prize, and string theory hasn’t been very popular among US physics departments hiring permanent people recently.

  25. Statistician says:

    On something slightly unrelated: is scientific blogging going to die? I’ve looked at google trends and it seems to be the case:

    Interest was great around 2005 – 2007 but slowly people just don’t bother reading all these science blogs. I’ve included Peter, Lubos, Sabine Hossenfelder, Cosmic Variance and Clifford Johnson in the comparison chart.

  26. SpearMarktheSecond says:

    Concerning experiments at the high-energy frontier… the US has always had the land and once (and probably still) had/s the expertise… that led to Robert R. Wilson’s `Desertron’, the name for what turned into the SSC. But certain practicalities (who wants to live in the desert? Which desert state has a congressional delegation as powerful as that of Texas?) took the SSC to Waxahatchie. And then the money problems killed it.

    Novel acceleration ideas have been researched for at least 30 years, with meager results.

    So probably we are stuck with a physically large machine. What makes little sense is putting the next large machine in Japan, where land is scarce, although perhaps their government is ready to spend for a linear collider.

    Can’t say CERN makes sense either, where is the land? Quite a densely populated place.

    A really grand concept would be an international collaboration that builds the next big one somewhere in Africa. The collateral benefits might be incalculably large.

    But there seems to be little appetite for grand projects in the US. We seem to have leaned more to our finagler, swindler, litigious, squabbling side than our grand side in the past years. A giant international project costing $30 billion with major contributions from all the world’s economies just sounds foolish in our current situation. We can only do $1,000 billion wars, nothing idealistic.

  27. Kernel says:

    I’m a graduate student at the Institute for Quantum Computing, another Lazaridis initiative, at the University of Waterloo. I like to ask questions about where the money for IQC and Perimeter Institute come from, and have found out enough to say that your statement that “Perimeter Institute has gotten $150 million or so from Mike Lazaridis over the years” is misleading.

    I have heard the sum of $150 million quoted a few times, though I’m not sure where it comes from. However, the figure probably comes from the total funding received by PI, of which Lazaridis contributed only a part. In fact, at least half of that money has been provided by the federal government of Canada, and a substantial portion was also provided by the provincial government of Ontario. In other words, PI is largely taxpayer funded. Lazaridis and others (such as the Templeton Foundation) have made important financial contributions as well, but the sum of their contributions is probably far smaller than $150 million. As far as I can tell, the situation is similar for IQC.

    I think an important difference in culture exists between Canada and the United States of America. In the US, there is significant political pressure to cut any government expenditure unless it is for something everyone agrees on, like military. Rationally speaking, a government agency that wants to protect its budget is well advised to claim that it is a military organization. Therefore, the fact that much US science funding comes through the military should not be too surprising. Just because spending is described as military, doesn’t mean it actually is.

    In Canada, people pay a lot less attention to what the government actually does, so the government can get away with spending money without coming up with arcane justifications. I wonder, for example, what value Perimeter Institute really provides to the Canadian taxpayer. I think a major reason that Canada can get away with hosting a place such as PI is that no one is really asking the pertinent question. It’s of course in my best interests to keep my mouth shut, but I suspect that PI does very little for Canada despite the fact that it is mostly funded by taxpayer dollars.

  28. Peter Woit says:


    2006-7 was the period of the String Wars, after the appearance of my book and Lee Smolin’s which got a lot of media attention. I can’t say that I was sorry to see that period go, since most of the heavy blog traffic then came from people interested in arguing about things they didn’t really understand. To some extent I think the story of the string theory debate was that it did get well aired on the blogs, and anyone who cared made up their mind back then, with few people still interested in revisiting the debate now.

    The numbers I see on my web server have been pretty constant for quite a few years now. I do think some activity has moved from blogs to twitter/facebook/google+ (and google wants to encourage this by killing google reader).

  29. Peter Woit says:


    Here’s one source for the $150 million number

    I believe Ontario provided another $50 million, and there have been contributions from other sources. Details of the Perimeter funding should be in its annual reports, latest one is here

    I think you’re dramatically oversimplifying the US science budget issues. For one thing, many fields of science get little to no funding through military channels. Biomedical research is a good example of something which the US pays for a lot of (the NIH budget is about $30 billion/year) and has wide support from many quarters. The NSF has long enjoyed political support from both parties. I doubt the average person or average politician in the US knows or cares any more about government science spending than their counterparts in Canada.

  30. srp says:

    Peter: Your slave-labor comment is odd. The government pays for people to get PhDs and conditions funding of research on new PhDs being trained. This is called “subsidizing” or “investing” in the creation of human capital. We may both agree that the policy is quantitatively excessive, but the rationale for it is as I stated earlier–the government is responsible for getting the most research output it can for its expenditures, and increases in research spending tend to raise compensation much more than research output unless something is done to make the supply curve of researchers more elastic.

    It’s fine to complain about the loss of tenure-track jobs and the switch to a contingent workforce in academia–folks over in the humanities have felt the brunt of this for decades, and its bad effects as well as its good points are pretty well understood. But the “exponential model” also doesn’t make sense–you can’t have an exponentially growing set of tenure-track jobs to match the exponential growth of PhDs implied by the current grant system. If each researcher pushes out two or three new grads each year, and each were to then get a tenure-track job and funding for research which entailed turning out two or three grads, and so on…pretty soon the number of researchers would exceed the world population. People are taking an atypical post-World War II situation in the U.S., where growth was exponential for a time, as normative.

  31. Peter Woit says:

    I’m in complete agreement about the fundamental problem of too many PhDs. My “slave-labor” comment was about the argument that the government is justified in using its power to intervene in the labor market in order to keep down labor costs. I’m curious whether this is done in other parts of the economy.

  32. dinotroll says:

    I feel at some point in time birth control has to be done. But I also agree that we cannot do it right after undergraduate since students are very raw at that time. Already PhD in Physics takes 6 years on an average. I would advise breaking it into 2 stages of 3 years each. The first would be something like a comprehensive masters, getting done with all graduate level courses and publishing a paper or two. Then when you apply for the next three years, birth control can be applied and the students themselves might realize whether they are cut out for physics research or no. At this stage the application procedure can also be made a bit more rigorous with interviews and research plans. With the present system it makes no sense that a person takes 6 years to finish PhD and then has to take up job in a software firm.

  33. srp says:


    “My “slave-labor” comment was about the argument that the government is justified in using its power to intervene in the labor market in order to keep down labor costs. I’m curious whether this is done in other parts of the economy.”

    Great question. Answer part one: The government (which is not a unitary actor, but it’s syntactically easier to discuss that way) is very interested in keeping down the prices of stuff that they have to buy, where they are a customer. They aren’t very good at it, for a variety of reasons (especially the political influence of their suppliers), but there’s a vast apparatus that tries to keep down the prices paid to military contractors, for example. Since most of the cost of these items is labor, especially white-collar labor, the effect is to depress the wages of defense industry workers. We also used to have this thing called a draft that enabled the government to confiscate the labor of every able-bodied male in the United States and pay almost nothing for it. (Thank you Milton Friedman and Richard Nixon for ditching that one.) But to this day, the government tries to get the cheapest volunteer force it can subject to quality and quantity constraints. Nobody wants the taxpayer to give government suppliers rents (returns in excess of what is necessary to call their resources into service).

    Answer part two: Physicians getting Medicare and Medicaid payments are a different case, because the customer is ostensibly the patient. But price controls aimed at lowering spending have been included in the legislation for many years; they haven’t been implemented because of political pressure from physicians and some patient groups, leading to annual “doc fixes” that put money back into the budget to cover the suspended price rollbacks. One school of thought on PPACA is that the various “accountable care organization” and IPAB innovations it includes are intended to stealthily push down the incomes of US physicians in the name of fiscal responsibility.

    Answer part three: Politicians have (insincerely and cynically in my view) spent quite a bit of time arousing public indignation at the compensation of specific managers and financiers. One interesting example was the AIG case, where a group of mid-level workers who were not responsible for the initial problems were specifically retained by the government to do the toxic security clean-up job with the promise of substantial bonuses for not deserting the sinking ship. Congresscritters went nuts over those bonuses, to the point where one of the drones wrote an open letter that I think was published in the New York Times trying to explain the situation. And of course we have the continual bloviation over executive compensation, trying to lower the pay of that sector of the work force.

    So I think it is fair to say that the US government regularly intervenes in various labor markets to reduce the compensation that specific types of workers receive. It is especially prone to do so when it has monopsony power, but shows little compunction about trying to beat down the wages of others it holds as undesirables or political untouchables.

  34. Peter Woit says:

    Thanks srp,
    This is getting too far off topic, but I’ll just comment that I think I see a pattern here, with the effectiveness of efforts to drive down wages inversely proportional to the size of these wages. From my experience here in Manhattan, I can assure you that those in the financial industry and at the top levels of other industries based here seem to be holding up pretty well under the federal government’s lash, whereas those working in biomedical research labs are a different story.

  35. Stephen Olsen says:

    A serious problem is the fact that many senior faculty members at Universities and National labs have lifetime tenure & never retire, thereby locking up positions and large salaries that would otherwise go to young people. In fact senior faculty that do require usually free up funds for more than one junior faculty position. Lifetime tenure should be replaced with tenure to some fixed age. Faculty members over that age that are still productive should be eligible for subsequent fixed term contracts that they could negotiate with chairpersons & deans etc.

  36. PhysGrad says:


    Any thoughts on the relationship between funding for graduate students in physics, the growing cost of undergraduate education (particularly at state schools where the cost is growing faster than at private institutions), and the ability of programs to attract and train talented graduate students who are coming in with higher debt loads and seeing very little growth in grad and postdoc stipends?

  37. Low Math, Meekly Interacting says:

    I’ll briefly add to some comments above about military R&D. I happened to work with a “couple” people in my first job out of college who transferred to the NCI from another military facility not too far away. The reason they moved on was not because of some deep moral objection to the potential military applications of microbiology (otherwise, they never would have worked there in the first place, though I suspect their work was not on weapons, but countermeasures). Rather, they came to be dissatisfied with the professional isolation and, more to the point, the perceived lack of high standards for deciding what projects were to be tackled in the first place. Apparently military R&D of that time (late 80’s, early 90’s) wasn’t be run by the sharpest knives in the drawer, and one needn’t be so concerned that money was being spent finding better ways to kill people. Rather, be concerned that money was being spent by fools to do foolish, and ultimately useless, work.

    I have little confidence the situation has changed much. Add to the idiocy the endemic waste of the Military Industrial Complex, and one can almost ignore humanitarian concerns altogether, and focus on the scandal of funding a jobs program that produces nothing. That money could be better spent on almost anything.

  38. Peter Woit says:


    The student loan situation in the US is horrific, with universities more and more funding themselves by getting their students to take on increasingly large amounts of debt, putting them in a situation Bruce Springsteen described as one of having “debts no honest man could pay”. Students with this debt are now graduating and entering a very rough job market, with ugly results. In this larger societal context, I’m not sure that Ph.D. stipends and postdoc salaries keep people from going to grad school, since the alternatives may be more dismal. One popular alternative to physics grad school has sometimes been law school, but that’s becoming another sad story.

    I don’t know much about the situation in physics. In the field I do know about, math in general, Columbia in particular, Ph.D. stipends are around 30K/year, and postdoc stipends I think have done pretty well. In recent years at the top places, you have things going on like the Simons Foundation postdocs paying 70K/year, and this does have an effect on bringing up what universities offer everyone, since they are trying to compete with these.

    Both Math and Physics grad programs in the US are now producing Ph.Ds. in record numbers, so there’s no numerical evidence that the loan problem has suppressed these numbers. The situation in Math and Physics is rather different, with Math Ph.D. students funded typically as TAs, Physics students funded out of grants. With government funding flat and increased faculty salaries and tuition, I would guess that has kept downward pressure on money available for student fellowships, so physics student stipends may be doing worse than in math. I have heard complaints from theorists in physics that the government is funding fewer of their students, but have to confess that this sounds to me like a good thing, given the absurdly small number of permanent jobs in the field.

  39. Al says:

    “Ph.D. stipends are around 30K/year”

    is that including tuition? because if yes, it must be hard living on 12K/year in NYC.

  40. Peter Woit says:


    If the Ph.D. students were paying tuition out of their stipend, they’d get a negative stipend, since full tuition is nominally something like $40K/year. But Ph.D students are admitted on fellowships that pay their tuition, and also the 30K or so stipend. Some students make extra money teaching extra courses, courses during the summer, or tutoring.

    The master’s degree programs are a very different story. There typically the students are paying tuition, even full tuition.

  41. Otto says:

    But whether [biomedical research money] is now being spent well, or whether taking away from some other priority to spend more in this area would be a good idea, I haven’t a clue.

    From everything I’ve read, NIH R01/R21 paylines have been extremely tight lately (e.g., NCI at the 7th percentile). Meanwhile, the utterly useless NCCAM get $128 million (FY 2012) and things like the (equally useless) TACT trial gets $30 million. One could argue that these are pittances in the overall NIH budget, but it’s not money well spent. Science recently reported that the sequester could cut success rates to ~15%.

  42. Bobito says:

    A country which has a lot of physics PhDs working as computer programmers is doing fine. The “liberal arts” training of our era is a broad technical background. To find the best physicists you have to fund based on potential, and inevitably only some subset of those funded will produce really creative work. However, if you had not funded the 100, you would not have have found the 10.

  43. StevenC says:

    Dear All:

    I share Peter’s worry about students – either their loan or their quality. I have supervised or help manage some graduate students before, and they produce work that is rather low quality. Unforuntely, I have no say on their graduation nor I am a thesis committee member; but I have made silent protests for skipping defenses of students that I myself have managed – laugh. Sadly, the poor-quality student problem falls under as a new spin to Publish-or-Perish: Have-Students-with-Thesis-or-Perish.

    Student loan is going to be a time bomb. Not only debt levels are high, but youth employment is on the high across the developed world (not just in US – it is as big as a disaster in Europe and Asia). That is a dangerous combo: the young folks get in debt, and cannot develop their career to pay back.


  44. StevenC says:

    Excuse my spelling – I mean “unemployment” instead of “employment”. I now work in the UK, but I was talking about the work I did when I was back in the US few years ago.

  45. StevenC says:

    Sorry one more point:

    I think the increase number of poor quality graduate students do have relationship with the problem with student debt and poor career prospects – going to graduate school become fads and quick fixes for poor employment prospects, and plays right into career expectations of senior research staff and professors. That is actually quite bad, imo.

  46. srp says:

    I worked for a while at a university that was on the edge of losing Carnegie Tier I status because it didn’t have enough PhD students. University administration repeatedly pressured our department, which was one of the strongest at the school, to start a PhD program. Faculty resisted because a school at that level would never have been able to attract graduate students near the quality of the faculty themselves.

    You want to think of your doctoral students as future colleagues. Nothing is more soul-sucking than a bad PhD program turning out marginal degree holders.

  47. Gilgit says:

    While the absolute numbers look OK, I think the chart on this page tells a lot:

    The chart shows how the percentage of applicants for federal funding that actually gets funded is much lower than in the past. It also shows, at least for the NIH, that the number of applicants is much higher.

    Which brings me to two points:

    1.) One of the reasons America has continued to be a superpower for the last 30 years is because we lead in science and technology. And one of the big reasons for that is because of government funding. We have a lot of scientists/engineers and the decision was made after WWII to keep funding science (with a bump up after Sputnik).

    To stay on top you have to keep spending. Since there is large core of scientists already out there, if you want to keep growing you have to keep increasing the funding.

    2.) I notice the number of applications in the chart has kept growing. This may be because it is for NIH and the biological/pharmaceutical fields keep growing. But I keep getting the impressions that big companies don’t plow their massive profits back into research like the used to. If we are going to keep leading in science then government money is going to have to take up the slack and needs to grow more than in the past.

    Of course America doesn’t have to keep being a technology powerhouse. The financial industry is still profitable. I’m sure they’ll do us all proud.

  48. Peter Woit says:


    From what I see of the numbers, there was a massive increase in the NIH budget 1998-2003, with more money going into the system causing a massive increase in the numbers of grant applications during the years after that. If you wanted the success rate for grant applications to stay constant, it looks like, after doubling the NIH budget in those years, you would have needed to double it again in recent years.

    Exponential growth is just not sustainable, it has to stop at some point.

  49. Gilgit says:

    (My post went long. If you don’t think I followed the rules you can delete it.)
    OK, I followed your link and looked at many graphs and concede your point that we are not at low levels of funding.

    I did think about it a bit, though. Are there strong reasons to increase spending right now? Some thoughts popped out at me.

    Looking at the non-defense R&D chart, I see that we really were spending a lot on energy in the late 70s. Luckily, Ronnie Boy worked very hard and after 6 years was able to gut many of the programs. This lead to the current world were no one thinks about energy or any massive side effects of using energy. 🙂

    Energy research might mitigate the effects of fossil fuels. Offsetting the amount of money going to drought relief, hurricane relief, and [you favorite catastrophe here] relief. And, thanks to Ronnie, we’re 20 years behind.

    Another thing that came to mind is the number of articles lately on diseases that have been showing up that are immune to every single antibiotic we have and how these strains have been appearing in western countries more and more. I’m thinking that if this trend continues and we have some deadly outbreaks, we might wish we’d spent more money in the preceding years.

    And as for private drug companies: than_expected_because_the_drug_pipeline_has_dried.html

    that says that the number of new drugs coming on the market is much lower than in the past. I’ve read that people are trying to get public funding specifically allocated to develop new antibiotics. I hope they succeed. In any case, private companies think its not profitable enough because the people dying today live in poor 3rd world countries. Gotta love that private sector.

    Finally, I can’t help but notice that the economy isn’t as strong as it was. Maybe you’ve noticed too. It would be nice if – when the economy does pick up – we have lots of scientists and engineers trained and fresh for the fight. Instead of lots of ex-scientists who have moved on because they couldn’t find work.

    If the $64 Billion was increased to $74 or $84, would that really be unaffordable? The total budget is $3.5 Trillion.

    By comparison, I’d point to a story NPR did a couple weeks ago about how the Americans with disabilities act, which was passed in the 80s has change the country. Many people who have become unemployable have been put on disability because we don’t have any jobs for them. Not because they are too ill to work. They can’t do manual work anymore or the only employer for miles around went bust. 14 million people that cost $260 Billion a year. I’m not against this, but as the report pointed out: we are spending $260 Billion to have people not work and almost no one noticed and few people even talk about it. It wasn’t a plan, just how things have worked out.

    I bring it up because if we can afford something that big by accident, we should be able to increase R&D funding NOW without breaking the bank. Even if the increase was only temporary. That chart (following your link) that shows “total non-defense R&D” spending – it says that in the late 70s, after Apollo was done, we spend $39 billion, so we still aren’t double that. So spending doesn’t sound unsustainable to me.

Comments are closed.