Arnold Kling  

Speaking of Science Funding

PRINT
Labor Market Commentary... The Locavore's Dilemma...

I attended a set of two panels this morning at the American Association for the Advancement of Science, touting "convergence," meaning joint work involving engineering and the biological sciences (with other disciplines as well). The scientists were talking their book, which they do with even less subtlety than Wall Street portfolio managers. Although the panelists touted more funding for science, there were a lot of arguments made that suggested that the right issue is not "more" vs. "less" but smart vs. dumb.

1. Alan Guttmacher of NIH said that he gets tremendous resistance from the scientific community in trying to re-orient NIH to be able to fund more innovative forms of research. The existing structure of NIH and its funding methods he regards as anachronistic, but they have developed powerful, determined political constituencies.

2. Keith Yamamoto of the UCSF school of medicine and Tom Kalil of the White House office of science and technology policy both praised funding models that gave administrators at universities a few million dollars each to allocate themselves, using local knowledge, rather than forcing everything through the Federal grant process.

3. Just as a point of information, several of the panelists were part of something called the Koch Institute for Integrative Cancer Research at MIT, which just opened, seeded by $300 million from the infamous Kochtopus.

If I could affect government science policy, I would distribute some portion of the government research budget through venture capital firms. That is, instead of having the grants approved by a panel of scientific peers centralized in Washington, grants would be approved by venture capital firms dispersed around the country.

VC's are used to hearing pitches about creative projects. They understand that although you try to fund the most promising teams and ideas, you can get failures.

There is the issue of how to set the goals for the VC firms. I would not think that one could use dollar profit as if they were funding business start-ups, because we are talking about research that could be a long way from leading to any practical implementation. Instead, we would need some other metrics. Surely, there are metrics that one can use to measure the success of research projects (citations by future researchers, for example). Or ex post evaluations by panels of scientists.

I would rather see decentralized funding with centralized metrics for success than the centralized funding that we have now.


Comments and Sharing


CATEGORIES: Business Economics



COMMENTS (14 to date)
Silas Barta writes:

*relabels my tuna factory as a VC firm to get the funding from Arnold_Kling's proposal*

*hires lobbyist to make it plausible*

Vipul Naik writes:

From what I remember and can find on the Net, the Koch Institute got _only_ $100 million, not $300 million, from David Koch:

http://web.mit.edu/newsoffice/2007/koch-institute-1009.html

I don't see any additional $200 million gift.

fundamentalist writes:

We need to keep in mind the Soviet experiment with state funding of research. Remember the 60 minutes episode on their research village with the top Soviet scientists and how little of their research was useful?

Hayek's "Counter-Revolution in Science" explains the current fetish with scientific research. Socialists set up scientists as the high priests who would lead humanity to salvation. In reality, pure science has given us cool toys and little else. Most productivity enhancing inventions come on the job and are made by engineers, mechanics and tinkerers. Most pure science research is an amazing waste of wealth, not to mention the even worse waste in social science research.

jeppen writes:

I've been following this blog for a long time now and, and I think this is as close as I may get to be on topic when asking: Do you deliberately avoid talking about nuclear power? I see very little on this topic here, but lots on just about everything else. A pity, I think.

It seems nuclear may need government R&D funding and support to progress, and the US is currently handing over nuclear leadership to China, who is performing broad research and is building some 25 nukes. They will probably start selling nukes around the world in the coming few decades, outbidding everybody else. They will also have spent fuel enrichment and reprocessing, control of lots of uranium production, a far advanced breeder concept and more. Are we fine with just buying the stuff from them and not competing?

Also, is the reason nuclear needs government support in the US that the government has put up too many obstacles? I think so! So, would it be ok to let two wrongs make a right, or should we just leave it be and try to protest the obstacles?

drobviousso writes:

What are your thoughts on using awards for primary research goals?

MernaMoose writes:

jeppen,

I pretty much agree with you on nuclear power. The US has been stupid on this front for a long time. But sadly,

Are we fine with just buying the stuff from them and not competing?

Yes, our green minority in this country is quite happy to let the technology go over seas.

The US is loosing its edge, and this is just one example. It is predominantly a consequence of our own stupidity.

wd40 writes:

"Keith Yamamoto of the UCSF school of medicine and Tom Kalil of the White House office of science and technology policy both praised funding models that gave administrators at universities a few million dollars each to allocate themselves, using local knowledge, rather than forcing everything through the Federal grant process."

I strongly doubt that my university administration, my colleagues in other departments, or even my colleagues in my own department are as good as judging my research potential as the reviewers for NSF who are experts in my area. In this case local knowledge is not geographically local but intellectually local. The fact that the grant giving is centralized should not be treated as prima facie evidence against NSF or NIH.

Q writes:

To say that grant decisions are made in washington is somewhat true in that experts are flown in from everywhere to meet and score applications and that there are govt employed Administrators to help manage funding, but it is a stretch to say that funding decisions are made by beauracrats in Washington.

Ian M. writes:

Unless you have the better metrics in hand, this post is your typically worthless conservative wishcasting. Get real. Basic research takes years to bear fruit and the privatization of such discoveries would vastly increase the cost of research. The savings of federal research is that all the knowledge produced is public, allowing private firms (with the engineers and technicians, etc...) to innovate productivity enhancing devices from the amazing pool of public knowledge. Besides, since all the increase in productivity over the last 30 years hasn't been shared with workers, why would anyone outside of corporate owners care about increases in productivity? It isn't going to my pockets (insert hand-waving supply-side drivel "rebuttal").

Contra fundamentalist, research science is one of the most competitive fields I have seen (including being sole proprietor of my own business). Each NIH grant has 8-10 PhD researchers jostling for the money. The labs are stocked with graduate students, Post-Docs and technicians all with college degrees all making less than $40K/yr. This is a highly competitive, highly efficient system. I maintain a core equipment facility at a major research hospital. Cost, cost savings, cost reduction and sharing of expensive equipment is mentioned *continuously*. The easiest productivity enhancement to this field would be to up funding, because right now the best grant writers, not the best scientists, get the money. VC firms would just switch that to the best pitch getting funded.
I could go on, but the author's prescription is so hilariously misinformed no corrective is possible in this space. But one last note - who would be the judges within a VC firm? Hopefully they would hire experts in the field to analyze proposals for malarkey. Wait, that sound like what we do now...

Kyle writes:

"Unless you have the better metrics in hand, this post is your typically worthless conservative wishcasting."

Hah yea, so true. This sentence was my favorite from the post:

Surely, there are metrics that one can use to measure the success of research projects (citations by future researchers, for example).

Exactly. Just ask them how many future researchers will cite their paper and then write a check to whomever gives the biggest number. Maybe the author meant that, years and years after the research has been funded, you can look back to see how good your investment was by citations. Not very helpful either.

If you really study the issue, you're going to find that it's extremely difficult to distinguish between promising research that will result in value for society and promising research that won't. This isn't to say there isn't some obviously dumb stuff out there to cut.

Kyle writes:
In reality, pure science has given us cool toys and little else.

Of course this is spectacularly wrong. The opposite is actually true – historically you get diminishing returns from engineering a well developed product, but can get large discontinuous jumps from improvements in scientific understanding. I’d also like to stress that the 3rd grade notion that scientists make discoveries and engineers apply them falls apart with new technologies; it can be very hard to tell the difference between the groups when the underlying principles are fresh.

Two easy examples off the top of my head are organic semiconductors and giant magnetostriction. Our understanding of these phenomena results from basic scientific inquiry that has resulted in products that people find very valuable. It’s hard to imagine a world where giant magnetostriction hadn’t been discovered; is someone really telling me that increasing hard drive storage capacity by more than an order of magnitude (with almost no change in production cost) doesn’t result in an improvement in productivity? Sometimes studying weird materials for no other reason than to catalog their properties leads to surprising results; that’s how much of 20th century science started, and it still pays off – high temperature superconductivity comes to mind. In a few years we’ll be able to add negative index metamaterials to the list – maybe we can already, and I’m just not aware of it.

There are two typical reasons that people think this little of “pure” research. Sometimes they do so because they’ve defined all research that doesn’t result in something that value as pure, and thus pure research is bad. Three cheers for tautologies. More often they simply don’t know anything about scientific research – they hear “pure” science and think of the LHC. To use physics as an example, everyone has heard the term “string theory”, but in reality almost no one works in string theory. In fact, more physicists work in condensed matter and materials science (how many books are there in Barnes and Noble on those topics?) than all other areas of physics combined.

I will give the commenter this: it’s refreshing to have someone be totally wrong on that side of the scientific fence. The population of Michio Kaku-reading scientific sycophants is much larger.

DixieFlatline writes:

Arnold,

Sounds like you're trying to put lipstick on a pig.

-Dix

Chris T writes:

All of our modern technology originates in what was pure science at one point or another. Einstein's theory of relativity was an explanation for a number of experimental observations that were at variance with Newton's laws. Those experiments did not have a practical application in mind; they were conducted to better understand the universe.

Today, accounting for relativity is critical in the accurate operation of the Global Positioning System.

DNA was first discovered 1869, but it took a century of research before the first practical applications could be devised.

Nuclear physics and quantum mechanics required all kinds of fundamental science experiments before practical uses were found.

Penicillin was found completely by accident while performing exploratory research on Staphylococcus.

Pure research can take a long time to provide dividends, but the dividends it can end up providing are enormous.

Kyle writes:
giant magnetostriction

Duh, that should be giant magnetoresistance. Sorry, that kind of flub is one of the dangers of posting quickly from work.

Comments for this entry have been closed
Return to top