From David Sinky on subsidies for science, by email

by on March 24, 2013 at 4:16 am in Economics, Education, Science | Permalink

Since it seems that the supply of talented researchers in any specific area is likely fairly inelastic in the short term, to what extent do you see cash funding (as opposed to supply of talent) as a major constraint to specific scientific research in either the short and medium terms?

>Even if we believe that this funding will lead to a proportional increase in clean-tech research, I suspect that returns may be quite low since the impacts of this funding would seem to be:

1. Pulling smart people from their private sector efforts into publicly funded research

2. Funding marginal projects by lower quality researchers where returns are likely to be significantly lower than average returns to research funding (which may be quite low already)

3. Increasing the funds available to established, high-status labs and researchers.  If a large percentage of a lab’s output is due to the abnormally high human capital of its lead researchers, the binding constraint is their time and mental resources rather than cash so the returns on additional cash would not be very high.

4. Allowing institutions that were already going to fund this sort of research to direct additional funds to other priorities such as undergraduate academics (stem or otherwise), student amenities or other unrelated research initiatives.

I suspect much of this logic also applies to donations to “cancer research charities” which I believe may be one of the single least efficient use of charitable dollars.

In general, I am disappointed that neither the right nor the left seems interested in trying to estimate the return to marginal government spending on research (either in aggregate or for specific programs).

The points above lead me to suspect it is quite low in aggregate but I’m open to being convinced otherwise if you think there is good evidence to do so.

1 So Much For Subtlety March 24, 2013 at 5:02 am

3. Increasing the funds available to established, high-status labs and researchers. If a large percentage of a lab’s output is due to the abnormally high human capital of its lead researchers, the binding constraint is their time and mental resources rather than cash so the returns on additional cash would not be very high.

This is likely to be entirely counter-productive. Little really interesting science comes from established, high-status labs and even less from established reseachers. Really good science comes from young unknown researchers. The more money you pour into the Establishment, the more young scientists will do Established science. Which may be mildly interesting but it won’t be great.

The good example is Einstein who produced excellent papers while outside academia, but then awards rained down on him. Once he was given the Nobel prize – at a fairly young age too – he produced what? He commented on Bose’s work. He produced a refrigerator with Szilard. That’s about it.

It would be a great idea to recreate something like the Bell Lab where scientists could do more or less what they liked. But how to make them productive and not lazy is a tough question. Perhaps the best way to spend science money is to scrap the Giant Projects approach, and to take the best student from every physics faculty in the country, every year, and give them five years to sit around with other graduates in New Haven or somewhere doing whatever they like.

2 dearieme March 24, 2013 at 9:19 am

“But how to make them productive and not lazy is a tough question.” I doubt it: my experience of research in good universities is that the only way to stop good people working is to lock them out of the lab. The trick is to find a way of ensuring that their effort isn’t all footled away trying to suck up to the bigwigs in their discipline, or in the destructive status-seeking of doing work of no practical utility.

3 adam March 24, 2013 at 12:57 pm

Biology has this and it’s called Janelia Farms (look it up!).

(and +1 to the dearieme comment)

4 Michael Gemmink March 24, 2013 at 5:30 am

Many of the break-through technologies of the 20th century came from government projects. The private sector is good at improving existing technologies but will skimp out on more ambitious research projects because of the risk. There is a reason nuclear energy, space travel and the internet did not have their roots in the private sector, yet the welface gains because of these inventions is incredible.

5 So Much For Subtlety March 24, 2013 at 5:45 am

Actually it looks to me like two of those did have their roots in the private sector. There is no denying that rocketry started out among hobbyists who did all the theoretical development a long time before the German Army became involved. Governments simply took existing technology – even the liquid fuelled rocket – and improved them. The theory for nuclear energy was also laid out in the private sector before governments got involved in building bombs.

6 Honza March 24, 2013 at 8:26 am

whoa! where did you hear that? nuclear sciences have fitrm roots in govt funded research. as does quantum mechanics.

7 Andrew' March 24, 2013 at 9:37 am

It’s all hysterical. We could argue about the internet. Or how the pneumatic radial tire that has provided multiples of benefit compared to nuclear was purely private. Or we could talk about how a handful of times the government actual got its hands dirty with actual public goods are the exception that prove the rule and get on with the business of beating the government into doing their job.

8 Andrew' March 24, 2013 at 9:38 am

The government is holding back nuclear, btw. They want nothing more than to screw the internet, and did inestimable damage with rockets before the private sector made lemonade with that tech.

9 Therapsid March 24, 2013 at 10:44 am

The government is holding back nuclear today, but that’s because of the development of a regulatory state which largely post-dates the Manhattan Project and the heyday of government funded science in the immediate post-war era.

The federal government did a great job funding science until the creation of things like the EPA and Nuclear Regulatory Commission the passage of regulatory legislation like the Kefauver-Harris Amendment. All well-intentioned, but all contributory to the TGS which not coincidentally followed their passage.

10 mulp March 24, 2013 at 3:00 pm

“The government is holding back nuclear, btw. They want nothing more than to screw the internet, and did inestimable damage with rockets before the private sector made lemonade with that tech.”

So, you are arguing that Nixon and Reagan were big government liberal leftists who destroyed the rocket and nuclear industries by their total government takeover of the entire private sector rocket and nuclear sector, only since Obama freed the private sector by getting government out of nuclear and rockets is the private sector finally free to privately fund new nuclear power plants and private rockets free of government intervention??

Not a single private funded nuclear power plant broke ground after Reagan until Obama.

Not a single new venture in rocketry was formed in the US after Nixon which marked the collapse of US domination of rocketry, until Elon Musk IPO’d based on his privately funded rocket serving the ISS.

😉 😉

All we need is the rise of the private sector with small government under Obama! Right Andrew??

11 JWatts March 24, 2013 at 4:02 pm

Your version of history seems unique:

Not a single private funded nuclear power plant broke ground after Reagan until Obama.

The last private new Nuclear power plant in the US broke ground in 1974. It had nothing to do with Reagan. Furthermore, all of the recent construction (“expansion of existing plants”) started the permitting process while Bush was President.

“The “Nuclear Power 2010 Program” was unveiled by the U.S. Secretary of Energy Spencer Abraham on February 14, 2002 as one means towards addressing the expected need for new power plants. The program is a joint government/industry cost-shared effort to identify sites for new nuclear power plants, to develop and bring to market advanced nuclear plant technologies, evaluate the business case for building new nuclear power plants, and demonstrate untested regulatory processes leading to an industry decision in the next few years to seek Nuclear Regulatory Commission (NRC) approval to build and operate at least one new advanced nuclear power plant in the United States.”

Not a single new venture in rocketry was formed in the US after Nixon which marked the collapse of US domination of rocketry, until Elon Musk IPO’d based on his privately funded rocket serving the ISS.

Armadillo Aerospace: founded 2000
Bigelow Aerospace: 1999
Orbital Sciences: 1982
Scaled Composites: 1982
SpaceX: 2002
XCOR Aerospace: 1999

12 Andrew' March 25, 2013 at 6:28 am

What is the utility of rocketry? We can put satellites in space. Had the government not killed thousands and decided to have the ability to kill millions with rockets we don’t know how long it would have taken to get the lowest hanging fruit of rocketry.

13 Bernard Guerrero March 25, 2013 at 2:39 pm

“All we need is the rise of the private sector with small government under Obama! Right Andrew??”

Actually, mulp, Nixon was pretty weak sauce as a Small Government type. Google “wage & price controls”. What he was, was an *opportunist*. Also, if Obama is willing to get the government out of the way in those fields, well, more power to him! I’m glad we agree on what needs to be done. :^)

14 So Much For Subtlety March 24, 2013 at 9:21 pm

The theoretical basis of nuclear energy was laid a long time before governments got involved. Hell, Robert Heinlein was writing stories about nuclear power a long time before governments got involved – Blowups Happen was published in 1940. H. G. Wells was writing about nuclear bombs even before that.

What government funding has quantum mechanics ever got from the state?

15 ThomasH March 24, 2013 at 6:24 am

I think we need a big push for
asteroid collision protection,
beginning to create the successor to the LHC perhapts a larger version of the SSC
new principle antibiotics to address resistance
replication research to correct publicaion bias

16 So Much For Subtlety March 24, 2013 at 9:25 pm

That is exactly what is wrong with modern science funding. We should fund nothing big. Too many large projects take too much money from ordinary scientists for grand prestige programs that may or may not work. The LHC should never have been built. The SSC neither. Rather the government should be handing out lots of small grants for more modest work. That will fund the young Einsteins of the future. It will also lessen the need to publish once the grant granting process is not disproportionately rewarding to a small number of projects.

17 Zach March 24, 2013 at 6:52 am

The supply of researchers isn’t as inelastic as all that. After the PhD, there’s a major winnowing as some people get permanent jobs and some don’t. Something like 2/3 or 3/4 leave and go into industry. So I would actually say that the supply of researchers is completely elastic in the short term.

The question of cash funding doesn’t directly affect the supply of researchers, because you have to have a permanent position (ie, post-winnowing) to apply for grants. The fraction of successful grant applications is very low (~15%), so it’s risky to create a position in anticipation that a grant will come along. So I think your correspondent is descriptively wrong, but functionally right.

There’s a complicating factor in all of this, which is that many grants have earmarks for a PhD student or postdoc. So they do reflect an influx of researchers, but the funds are earmarked so that the new researchers have to work on old researchers’ projects.

My general impression in physics is that most groups can add a few people or start a new project with no noticeable decrease in quality. The supply of quality researchers is large, and the lead time in carrying out projects vs thinking up new ones is large enough that there’s a significant backlog of quality projects. But the translation between new sources of funding and new researchers working on new research directions (ie, not doing established researchers’ work for them) is indirect and slow.

18 Ben March 24, 2013 at 9:51 am

Going into industry does not stop someone from being a researcher, and it is not a winnowing. Some of the best R&D goes on in industry.

19 dstraws March 24, 2013 at 1:16 pm

r&d on applied science. Most of the pure science industrially funded research is gone. You want to see what is on the pure science end of things broadly go read Science or Nature.

20 mulp March 24, 2013 at 5:44 pm

“Some of the best R&D goes on in industry.”

Ok, name some examples that have occurred in the US in the past quarter century that were not funded by the US government or by non-US national industrial policy initiatives.

DTV in the US was Clinton era national industrial policy that many hoped would put the US private sector back in the display manufacturing business. Government policy was ensuring a huge consumer demand at the beginning of the 21st century which would be so large that dozens of manufacturing plants would need to be constructed, manufacturing plants that would be heavily automated using industrial robotics tailored to high volume large size clean room manufacturing. With the Bush administration, the DTV transition was hands off with the Bush administration letting the industry delay and delay and delay the final change-over. It was Congress that set the deadline based on the desire to free up spectrum for wireless, something that Congress expected to happen circa 2005.

It was the 21st century when Steve Jobs gave up on the US for manufacturing because the R&D on how to manufacture has not been done in the US. The US can create materials and ideas, but turning them into products is beyond the abilities of US manufacturing because the R&D has not been done in the US by either government or the private sector. I remember Jobs being ridiculed in the 80s for building a factory to make computers in the US at a time when US computer manufacturing was moving to Asia – he was soon booted out of Apple.

And if you can not manufacture a product, you almost certainly can not design a competitive product.

21 Silas Barta March 24, 2013 at 8:06 pm

Ok, name some examples that have occurred in the US in the past quarter century that were not funded by the US government or by non-US national industrial policy initiatives.

Okay: Pretty much everything related to our understanding of how to scale up large informatic systems and process big data sets, machine learning on such sets, recommendation systems, etc. Then Google’s development of driverless cars … That’s just off the top of my head.

22 Dan Weber March 25, 2013 at 6:53 pm

Not that I want to encourage him, but the driverless car was a direct result of a DARPA challenge. No one succeeded until Thrun finally pulled it off.

23 Zach March 24, 2013 at 7:02 am

With respect to clean-tech research, the connection between new money and new ideas is likely to be very weak. Most clean tech ideas actually build on very mature technologies where people have already tried most of the good ideas. Also, the thermodynamic limits and economic price points are very unforgiving.

Even if your marginal dollar produced the next Einstein, would that really help photovoltaics become economically viable? Photovoltaics are as old as the last Einstein! The first solar cell was produced in 1880!

24 JWatts March 24, 2013 at 10:44 am

would that really help photovoltaics become economically viable?

There’s a misunderstanding of the economics of solar cells that makes smart people continuously overestimate the overall utility. The focus is almost alwayson the cost of the solar cells themselves and it fails to look at total system cost. In the US you could give away 10% efficient solar cells for free and the levelized cost of power would still exceed the current wholesale cost of electricity.

The US should concentrate Federal funds on research of high efficiency cells, not production of existing low efficiency cells.

25 paul March 24, 2013 at 7:20 am

Of course the short term returns are low! The whole point of federal research for science is that it can fund research for time horizons that are not viable for industry. So it’s silly to impose a private sector type of short term view on this thing.

The only thing that matters are the long term returns. Publicly-funded science has a pretty good track record on long term returns over the last 100 years.

26 BC March 24, 2013 at 12:22 pm

We don’t see the whole context of the email but, given that the email mentions clean-tech, maybe the context is publicly-funded research in “targeted” areas. In that case, your point that publicly-funded science tends to produce results on longer time scales — and I would add in areas and ways that are not predicted at the outset — would also argue against directing public research funding towards politically targeted areas like clean-tech.

27 dstraws March 24, 2013 at 1:24 pm

I think we are conflating science with technological development. You may disagree with targeting clean energy technological development (of which Solyndra was the poster child), but that is not the “science” that is funded by US. The science with long term down stream effects is that being funded by DARPA, NIH and NSF among others.

28 BenK March 24, 2013 at 7:23 am

The scientific enterprise is very inefficient, yes. However, one of the largest inefficiencies is that sustainment requires constant begging and shilling for funds. The people who should be thinking about hard problems in science are instead thinking about hard problems in funding. The NIH, Wellcome Trust and HHMI have all hit on the same solution – give the promising researchers 5-10 years of uninterrupted funding without rounds of competition or threats to lab group survival (ie. enough funds for the PI, staff, administrative support, facilities, equipment, supplies). Lower the stress, increase the productivity. They might even be able to do better if it were 7-15 years – but to systematically improve the state of science, this paradigm needs to be extended massively, to embrace the majority of the enterprise. In an environment which has even the Nobel Prize winners trimming their labs, the premise that science is presently saturated with funding is pretty silly, even were there a point at which talent inelasticity came into play. There’s tons of science to do, many scientists willing to do it, and large numbers of very useful proposals are unfunded.

29 prior_approval March 24, 2013 at 7:45 am

‘ I suspect that returns may be quite low since the impacts of this funding would seem to be:
1. Pulling smart people from their private sector efforts into publicly funded research’

Well, that certainly describes how the Internet was created – all the smart people pulled away from their private employers while then being funded by ARPA.

Imagine what the Internet would have been if done by a private company.

The funny thing is, it is actually pretty easy to imagine what the alternatives to ARPA’s funding brought into being. It is just very rare that anyone actually has any experience with them, though old time IBMers are the most likely.

30 Ben Webster March 24, 2013 at 8:49 am

I guess I’m just piling on here, but I think your correspondent is being totally wrong-headed here.

1. Science is constantly losing huge amounts of human capital. At the moment, we’re training more PhD’s than can be absorbed in academia, and a lot of people are hanging around in short-term jobs, waiting for something permanent to open up, and many of these people are very smart.

2. At least in my experience, the best grant applications being turned down are pretty good. Funding rates are in the low double digits! Also, I wouldn’t be so confident that NSF panels are really distinguishing a 90th percentile proposal from a 80th percentile proposal that effectively.

3. This certainly doesn’t have to happen, though as mentioned above, even very established labs could benefit a lot from having a more reliable future funding stream. I certainly think funding agencies want to fund a broader range of people, especially younger people, in a broader range of locations (keeps the congress-critters happy!), and I think that’s where a lot of the marginal dollar would go.

4. This might happen in some cases, but I suspect deans would be more likely to react by working harder to get and keep people with external funding, not resting on their laurels. Sad though it is to say, external funding is a good way for colleges to “keep score,” so I think they tend to run after it in a way that doesn’t necessarily make financial sense.

31 Andrew' March 24, 2013 at 9:48 am

“I suspect much of this logic also applies to donations to “cancer research charities” which I believe may be one of the single least efficient use of charitable dollars.”

I beg your pardon! Well, on the bright side, it’s possible that that money is mostly completely wasted, so that it doesn’t even have the distortionary effects you refer to. Maybe The Fed should give cancer charities all the new money first.

32 BC March 24, 2013 at 2:00 pm

I have also wondered about the wisdom of donating to research that can already compete for public research dollars through the “normal” research funding process (competitive grant proposals, peer review, etc.), cancer research included.

One area, however, to which I have donated in the past was embryonic stem cell research. My reasoning was that, due to the political controversies surrounding it, it’s quite likely that fewer research dollars have been allocated to it than otherwise would have been based purely on its scientific (and economic) merits. If one extends this reasoning, one concludes that private charitable research dollars should be allocated to the least politically popular research. Donating to politically non-objectionable research is actually redundant. Like TC though (or is it Sinky?), I am “open to being convinced otherwise if you think there is good evidence to do so.”

33 Andrew' March 25, 2013 at 6:30 am

I suspect you have it bass ackwards on stem cells. Their potential was widely exaggerated as a funding drive, IMHO.

34 prognostication March 24, 2013 at 9:53 am

The entire premise strikes me as a little ridiculous, because it assumes a more-or-less efficient labor market, and there’s plenty of work showing various ways that the labor market isn’t all that efficient. Some talented people want public sector jobs even when private sector jobs pay much better. Some have spouses with better jobs locking them into labor markets with limited opportunities. And so on and so on. My view is that those inelasticities matter much more than modest changes in federal funding (and let’s be realistic — modest changes are most of what has been on the table in recent memory).

35 mw March 24, 2013 at 10:16 am

While we’re spending $13 billion / year on the carried interest loophole, $20 billion / year in farm subsidies, and at least $500 billion / year on “defense,” there’s clearly a deep need to debate how “productive” federal research dollars are.

Clearly “libertarians” would be happier if we went back to implicit rather than explicit federal research funding, a la the Bell Labs monopoly juggernaut, that way they can’t put a price sticker on it and, in any case, they can tell themselves it’s the work of the “private sector.” For that matter, that’s what we *still* have in medical device manufacturers and pharma R&D departments, which certainly would not exist in any recognizable form if we didn’t allow them to rape the price mechanism.

36 Joe Smith March 24, 2013 at 4:47 pm

MW – agree completely on the implicit subsidy to Bell Labs through the monopoly. A lot of good work came out of Bell Labs. It sure would be nice to achieve those kinds of benefits in the future without the cost the AT&T monopoly imposed on consumers.

37 Surse Pierpoint March 24, 2013 at 10:20 am
38 Ray Lopez March 24, 2013 at 10:23 am

The problem is not with the short term, it’s with the long-term. The US government needs a better patent system. Presently the rewards go to incremental improvers not fundamental breakthrough inventors. The law tries to compensate for this with a “doctrine of pioneer invention”, but we need better statutes like: (1) a stronger presumption of validity for ‘true pioneer inventions’, a designation to be given to truly pioneer inventions as determined by expert panel. Presently about 40% of all patents are invalidated and that’s too high; (2) a statutory ‘prize’ for inventors to go to them regardless of what their employment contract says for pioneering inventions (presently in the USA most employees get nothing for inventing something); (3) a statutory prize for industry for inventing something really important, for example a prize for the flying car to the first team to meet certain milestones: (4) anti-trolling statutes (not a big deal actually, but it’s icing on a cake); and, (5) special “German style” fast track courts for IP. The present ‘district court overloaded with petty drug offenses and review by the CAFC’ is inefficient–hence it costs $200k+ in legal fees to litigate a patent. Should be 10x less. With the above society will really profit long term–we’d be living forever with zero radiation nuclear fusion and traveling faster than light in a flying car if we had the above, instead of worrying about petty “marginal” issues like how much to cut Social Security and Medicare for rich seniors. Nuff said.

39 JWatts March 24, 2013 at 10:58 am

To sum up many of the comments: Special interest groups are always interested in protecting and expanding their money. And their work is always more important than anyone else’s work.

Ask a farmer if farming need more or less money from the Federal government and she’ll say more.
Ask a cop if police departments need more or less money from the Federal government and she’ll say more.
Ask a scientist if research labs need more or less money from the Federal government and he’ll say more.

40 jseliger March 24, 2013 at 11:06 am

If a large percentage of a lab’s output is due to the abnormally high human capital of its lead researchers, the binding constraint is their time and mental resources rather than cash so the returns on additional cash would not be very high.

Sinky might be right overall, but even high-status labs apparently must spend a lot of time, energy, and money on grant writing and other fundraising efforts. If you throw some cash at them and lower the barriers to getting that cash, more of their time and mental resources might go to their work instead of seeking money for their work.

41 Bill March 24, 2013 at 11:18 am

What you need is a new paradigm for funding and scientific research:

Kickstarter for Science.

You put up a research proposal online. Some experts critique it. People fund it as a non-profit contribution, or fund it taking license fee rights to discovery.

42 prognostication March 24, 2013 at 11:46 am

iAMscientist and PetriDish, at least, possibly others, already exist for this purpose.

43 Bill March 24, 2013 at 11:59 am


Then you’re ready for another idea:

Remember the 19th century when country gentlemen were part time scientists, collaborating with academic scientists or clergymen scientists, or others interested in science…small discoveries, and even some big ones (evolution?), came from this. Remember in the 20th and 21st century how some folks make their computer time available for other projects.


What about experts collaborating outside of work with others in their field, or in fields that they need a collaborator in, to conduct research as a hobby, or as part of an academic research project.

44 Bill March 24, 2013 at 12:03 pm

prognostication, Thanks for the web link names. +1

45 anon March 27, 2013 at 1:11 am

Microryza is another major effort, actually. This space is very much in flux.

46 mw March 24, 2013 at 11:52 am

…and then we get the NYTimes-ification of science, lowest common denominator where only projects with immediately comprehensible possible outcomes get funded.

47 Bill March 24, 2013 at 12:01 pm

That’s why I mentioned having expert comments and review that is part of the process..

However, sometimes having stupid ideas researched and definitively put down has value: I would pay for research on the effectiveness of copper wrist bracelets on “curing” arthritis.

48 anon March 24, 2013 at 11:18 am

I just don’t see this inelasticity at least in physics. Lots of very able and enthusiastic young phD’s graduate every year and leave academia just because it’s too risky to keep doing research. Those who don’t quit find themselves fighting over grants to allow them to do research one year at a time.

2. As if there was some way to know beforehand which research projects are likely to yield large returns. It is exactly these “marginal projects” that we’d need to be funding since the biggest findings lie where no-one has looked before. Many times in science underappreciated ideas by less known researchers have turned out to be revolutionary dozens of years after their discovery. For this reason assuming that there are some “high quality researchers” who already have all the funding they need and “low quality researchers” who couldn’t do anything worth funding anyways is absurd.

49 PNP March 24, 2013 at 2:40 pm

anon–I also finding it trouble that grants tend to be for such short durations. Good research is often a multi-year process, but most grants fund only short-term projects. I feel that we also have a bias towards funding projects that are actually not that original or that are by more established researchers, because familiar projects seem more likely to succeed at producing incremental results. But this is likely not the best way to push along real scientific breakthroughs.

50 Addicted March 24, 2013 at 11:32 am

How many smart science people in the private sector are doing science as opposed to writing HFT algorithms to squeeze out a few more fractions of a penny out of each transaction? The latter being a simple transfer of funds, as opposed to productive work.

51 BC March 24, 2013 at 1:00 pm

The latter increases the efficiency of financial markets, significantly reducing transactions costs for all market participants. The HFT guys are the reason bid-ask spreads are so much narrower now. The “transfer” of funds is from the old-style market makers, who have been replaced by the HFT guys, to the investing public that gets the benefit of the HFT-provided liquidity, which is to say that it is not a transfer but a cost savings. The social value of that transactions cost savings is huge when one considers global transactions volume.

None of this is to argue against science funding. In fact, one could make the opposite argument. Funding of university-based research subsidizes graduate education. Even if future research needs are unpredictable, STEM PhD students tend to have abilities, enhanced by their research-based education, that can be adapted to add value in a variety of areas in the unpredictable future, HFT being just one example. There is a positive externality to having a human capital base that can be flexibly redeployed to those areas where market demand develops in the uncertain future. Research funding buys a kind of option on human capital.

The fact that scientists have become HF traders is good thing and is an argument for funding scientific research. If we subsidize MBA or Masters of Finance education, for example, we may end up with a lot of finance professionals that we won’t need in the unpredictable future, and they will most likely not be able to become physicists. If we subsidize graduate STEM education through research funding, we will be in a stronger position regardless of whether it turns out that we need more scientists or more HF traders (or more management consultants, patent attorneys, data crunchers to determine things like which Target customers are pregnant, or any other of the variety of jobs that STEM PhDs do).

52 jm March 24, 2013 at 2:04 pm

While narrower bid-ask spreads are a good thing it’s not clear that HFT is the cause of decreasing spreads or that it benefits other market participants. See, for example,, whose authors suggest that investors bear significant costs from HFT, with the example that the spreads on Citi reduced from 20 basis points to two following a reverse split that made it less attractive to HFT.

53 PNP March 24, 2013 at 2:35 pm

I don’t think it is true that increased public funding would pull people from the private sector. That seems to imply a shortage of researchers. But the data I’ve seen suggests just the opposite–that there are more people with advanced STEM degrees in many fields than jobs and that many private companies are shedding R&D.

Public science funding strikes me as inefficiently low at the moment. Grant levels are so low that I believe we are seeing very talented people leave the field entirely rather than put up with the funding process, and scientific researchers are having to spend almost as much time writing grants as doing science.
All that said, I do think an interesting question is how best to increase funding (e.g., to we fund a few people very well or try to spread the wealth).

54 sceptic March 24, 2013 at 3:34 pm

Public funding strikes me as inefficiently high. Ever talk to a science PhD student outside of the top ten schools? Not the sharpest tools in the shed.

55 Alan March 24, 2013 at 5:52 pm

Wow! I’m impressed! Newly minted science Ph.D. graduates in fields from molecular spectroscopy to immunology and you are competent to judge how sharp they are.

56 PNP March 24, 2013 at 7:26 pm

Well, they probably aren’t getting the bulk of public funding since they usually don’t get jobs as professors . . . They might benefit derivatively as grad students, but in that capacity they are basically assisting profs who did graduate from top 10 schools.

57 PNP March 24, 2013 at 7:35 pm

Actually, my comment is pretty dumb itself. There are people we perceive as smart and dumb everywhere, so I don’t think this is a productive line of inquiry. I guess I would ask, sceptic, what you see as the alternative to public funding that would be less inefficient. I agree that we could administer public funding much better than we currently do, but I am skeptical that private industry will invest in basic research, and I find it unlikely that the public at large is qualified to evaluate really advanced science.

58 David March 24, 2013 at 5:06 pm

As an alumnus of MIT and a current doctoral candidate researching Cleantech at Caltech, I have a few observations:

Re point 1: a lot of top end quantitative talent goes into finance and non-technical consulting. Money might fix this.
Many research mentors are terrible at improving the human capital of their students. They wouldn’t last a week as managers in the private sector. This isn’t obvious because their students and employees are smart enough to develop themselves.

Point 2 and 3: Our system is very wasteful of this talent. Running a top research group is a full-time job, yet we give these people obligations to teach and serve on academic committees. How much of your own time is spent doing these things, Tyler?
It takes a lot of time for a professor to get all that money. That is most of the job of top research talent… technical sales. That’s a waste.

Point 4: just earmark the funds. Or found Max Planck institutes.

Our system is very very good at identifying top research talent. It is very terrible at then actually USING it for research. And those problems are systemic.

59 Bill March 24, 2013 at 5:39 pm

+1. In the battle between research skills and grant application and tech marketing skills, the latter wins out. But don’t presume it is different in private enterprise. Who wins in the office contest: the person adept at office politics, or the person with a good research idea?

60 Andrew' March 25, 2013 at 9:33 am

The difference is that that is a known issue in business and people often try to impose systems to combat it (e.g. According to Charlie Munger Sam Walton forbade purchasing agents accepting so much as a hot dog from vendors)whereas academia treats those bugs as features. I recently read a claim that your favorite whipping post tenure is such an example- (among other things) it allows the PI to behave idiosyncratically while allowing the University to claim an inability to discipline them.

61 PNP March 24, 2013 at 7:24 pm

Our system is very very good at identifying top research talent. It is very terrible at then actually USING it for research. And those problems are systemic.

+2. That’s exactly right.

62 James Davies March 25, 2013 at 9:02 am

Agree with David above. Sinky doesn’t even get #1 correct. Our system trains many bright researchers. The problem is there’s a huge oversupply of scientists who are fresh PhDs and post-docs who can’t find long-term work in their field. Depending on the field, half or more of these people cannot find jobs after completing apprenticeships (post-docs) and leave the field, often leaving science too. So the scarcity is in funding for people doing research long-term, not a scarcity in human capital. Do search for “STEM PhD oversupply” on Google. The idea that adding public funds would somehow pull “smart people from the private sector” is ludicrous.

63 Mark B March 26, 2013 at 2:34 pm

All the scientists that I know, including high profile ones with long records of success, spend huge amounts of time on grants and grant-related activities. Even for these established researchers, it is stressful, time-consuming and there is always a push toward doing conservative, guaranteed to “work” (i.e. lead to papers) research.
In biomedical research here in Canada, the system is very conservative and definitely only rewards young researchers who have been in the same field for their entire careers (or nearly). It is very hard to get any idea that is “out of the box” funded and you must have a long track record in any case.
I can’t recommend going into biomedical scientific research to anyone, unless you view it like becoming a musician or artist. It is that sucky.

64 Jacob AG March 27, 2013 at 11:06 am

Um, re: #1, the private sector is relatively likely to do science that is privately useful, not publicly useful. There are lots of positive externalities to inventing the internet for example (thank you DARPA), hence we subsidize basic research.

Comments on this entry are closed.

Previous post:

Next post: