Our Daily Bleg: How to Justify Long-Term Scientific Research?

In response to our bleg request, Rafe Petty of the University of Chicago chemistry department wrote in with the following question(s). Let him know what you think in the comments section, and send future blegs to: bleg@freakonomics.com.

I was recently at a lecture by George Whitesides, one of the most well-known living chemists. He gave a very interesting lecture at the University of Chicago outlining some of the most pressing problems that we as chemists face in the near future.

A very interesting problem that he focused on was the difficulty in justifying long-term research funding in a capitalist economy, which recognizes financial return — not social return — thereby assigning little value to long-term research.

My question centers around if it is possible to “fix” this problem.

Is there an economic model that justifies long-term research spending? Whitesides mentioned in passing something about option theory. Is this the only model that works? Or are there other ways of encouraging/promoting long-term research spending?

Thank you for your insight. This is an important question for any number of scientific disciplines, and of particular interest to young scientists who will be soon entering their respective fields.

Eric G

Who says that science is under funded? Scientists, of course, especially young ones who face an uncertain future.

I don't think that there is "problem". (I say this as a former scientist, BTW.) Governments and corporations must allocate resources. It's called economics. In a state of equilibrium, everyone feels short-changed.

Not that being a graduate student and worrying about the future is any fun. The memories ....

Actually, one economist who has argued that
medical research is under funded is Kevin M Murphy of the University of Chicago.
From http://www.press.uchicago.edu/cgi-bin/hfs.cgi/00/15531.ctl:

From an economic perspective, they find, the answer is a resounding "yes": in fact, considering the extraordinary value of improvements to health, we may even be spending too little on medical research. The evidence these papers present and the conclusions they reach are both surprising and convincing: that growth in longevity since 1950 has been as valuable as growth in all other forms of consumption combined; that medical advances producing 10% reductions in mortality from cancer and heart disease alone would add roughly $10 trillion-a year's GDP-to the national wealth; or that the average new drug approved by the FDA yields benefits worth many times its cost of development.
Without basic science, no medical advances.



Ignoring without comment those who believe that the solution involves (1)substantial alteration of human nature; and/or (2) more taxes to be spent by politicians and bureaucrats whose ignorance of science is legendary; and/or (3) fixing some systemic failure of an economic system to do something or address some issue which is not economic but rather subjective, e.g. "happiness"; and/or (4) some combination of 1, 2, and/or 3.

If basic scientific research is necessary and produces theories, ideas, technologies, etc which lead to or create practical applications leading to wealth creation, then whether there is "an economic theory" to justify funding for such research or not, funding will appear in the market.

It is entirely unnecessary that a theory be constructed prior to action. Especially in economics, construction of theories regarding how markets appear, operate, disappear, etc occurred well after the appearance of the markets which the discipline of economics was/is intended to discribe and explain in rational terms, written down in books. If enough freedom of economic action is retained by individuals within the American society - economic, political, what-have- you - funding will be found for productive activities. How do I know this? Because that is what has happened in the past. You don't truly believe that all of the growth of the last 200-plus years in the American economy occurred because of altruistism, government funding, or some pretentious political philosophy; do you? If you do, re-read any one of the many economic histories of the U.S.

Short reply: "Theory-Schmeory; work for it."



Two different questions, as I see it. First, how can anyone (including the gov't) place a high enough economic / financial value on long-term research? There's no inherent problem here, it's just that using the most common discount rates (3-5%/year) ends up placing a very low value on any benefits that accrue in, say, 30-100 years time. This is a common issue in health and environmental economics. Note that this is not a low moral valuation of the future, but rather a statement that if real interest rates are actually going to be at that level for 50 years, it would be better to invest the money and simply give it all back in the future, rather than spending it on science now.

However, for various reasons (namely the impossible nature of making political commitments for 50 years), this is not a satisfactory answer for most people. One alternative is to use a lower discount rate (e.g. 1%) but that gives weird implications for the short-run. Hence many economists (myself included) suggest using non-constant-rate discounting instead (e.g. starting at 4% and declining to 1% in the far future). Try googling this term, or "hyperbolic discounting", or look at the Stern Report on the environment. Problem solved, if you can convince Congress :)

The second question is whether or not a capitalist system, in which firms are actually trying to maximize their NPV profits (but at what discount rate??), can reproduce these outcomes. As several people have pointed out, the problem seems to be that most firms (or rather, potential stockholders) seem to care more about quarterly earnings than anything else. Google claims to be an exception. That's probably a mistake on everyone's part, but the CEO is only planning to stay for a few years anyway (the stock price is bound to go up faster than the S&P sometime -- so he'll stop there, claim credit, take his bonus, and move on). In theory one could align these incentives, but I think it's going to be very hard in practice, which is why (even with patents) the government is the natural home for basic science.



Like some of the others, I think you're asking the wrong question. I don't think we should justify science - that sounds an awful lot like we're trying to get away with something, to pull something over on people.

I think the real answer is to do a lot more (and a lot better) scientific and skeptical education and outreach. If people want science, they'll get science.


As a young scientist who will be soon entering my respective field (chemical engineering), these questions strike me as fascinating. In order:

Q: Is there an economic model that justifies long-term research spending?

A: Yes, its called the german-style research institution (AKA a university). Organize a bunch of researchers, then pay them to solicit money from others (students, donors, government, etc.) in exchange for services (books, papers, talks, lectures, etc.) A quick look at Whitesides's own institution will show that this model can work quite well in making money.

Q: Is option theory the only model that works?

A: My understanding of option theory is very limited, but I know that it relies fundamentally on accurate quantification of risk and rewards in the quantity being optioned, in this case long-term research. I think the reason we call it "long-term" is because we're admitting that we can't quantify its benefits or even bound its stochastic variation, so I think even option theory will fail in this case (sorry to disagree with you, Rick from post #14).

Q: Are there other ways of encouraging/promoting long-term research spending?

Outreach and education. People have mentioned it multiple times in the comments above; scientists must communicate with the public what they are studying so that they (the public) can figure out how to apply it to benefit themselves. The scientists can then solicit money from the public (students, donors, government, etc.) to continue to do their work, train new scientists, buy new equipment, etc.

I'll be the first to admit that I'd love it if the government just gave me the money to run a lab. But then where's my economic incentive to do research?

As for the disconnect between social and financial return, my personal belief is that once going to a university becomes affordable again (the gap in college graduation rates between rich and poor is currently larger than its ever been and increasing) a lot of the social returns of research will be re-realized. Either that, or the revolution will come and I'll have to find a new job.

P.S. For more information on the rise of the current American higher education model, see "Killing the Spirit" by Page Smith or "To Advance Knowledge: The Growth of American Research Universities" by Roger Geiger.



"My model" would answer: as long as science and finance is connected so close we will have problems of financing long-term research. It shouldn't be like this. In a society where esteem rules and not money, this isn't a problem anymore.
www.sikantis.org and www.sikantis.net


There's a potential model for this in patents procured by and licensed to the government free of charge. Patents procured by inventors working at Aberdeen Proving Grounds, China Lake, or other military facilities may be used by the Department of the Army, Navy, etc.

The ROI is theoretically in the reduced cost of national security to the taxpayer. But then again, economics isn't my field. Maybe it doesn't work that way.

Maybe positing open licensing of the results of government funded research can provide a means of constructing a model for cost recovery.

El Conejo

As a mystic, I see need for neither long term or short term research. Silence and contemplation reveal all that is known and unknown. The problem for the rest of the world, is there is very little money to be made in this avenue or pursuit of knowledge. Therefore the world chooses the worship of Mammon as justification for how it lives. Show ME the MONEY.

geoff v


At my university, I think that the 50% overhead goes to making those nice little name plates that hang above professors doors. Oh yea, and buying the cardboard toilet paper. Other than that, I haven't got a clue.


As a chemistry student at University of California, San Diego, many professors will tell you similar things. However, it seems that funds are available for research regarding human centric information. Such information is likely regarded as affecting society's perception of human ability or origin, revealing the egocentric attitude of current civilization.

However, long term pursuits, which often start as a purely scientific anomaly, are often ignored. Too much high level science and lack of immediate applicability has probably doomed many projects. However, quick short term research, in my opinion, uses failed or interesting long term data that was largely ignored the first time around due to dissenting opinion. Ultimately, long term research is very useful in probing unresearched areas of science.

The most approachable case would be the discovery of stem cells. In the early 1960's, there were research projects underway in the process of neurogenesis, most notably Altman. His work was initially ignored due to previously held theories countering neurogenesis in adults. Eventually, during the late 1990's, neurogenesis was discovered to be caused by plasticity of stem cells. Once the application of stem cells was understood, research took off.



I keep seeing references to Bell Labs when the subject of business financing pure research is discussed.

Bell labs was not a product of a free market as we tend to think of it. Bell Labs existed largely because the telephone company was a "cost plus" regulated monopoly. They could fold the expense of Bell Labs into their rate base and the regulators would assure that customers had to pay this cost.
Notice that once the regulated monopoly model of telephone service went away so did Bell labs.


Sure. If land is a resource and there is unknown land out there, of course you would invest money to go stake it out. Knowledge is a resource too, and research is the only way to explore future expanses of knowledge.

Why do we need new knowledge at all? Simple. In any industry, if sales do not consistently increase or quality consistently decrease, profits go down. Eventually sales and quality must saturate. So you need a new industry to start over in.

And if there was no new knowledge getting stockpiled by research, there ain't gonna be no new industry when you need one.


To start, I would suggest looking at where the gov't funding for basic science actually goes. In my institution and discipline (applied physics) the bulk of research is done by graduate students and the bulk of research costs go to grad students. A grad student requires 100k/yr (20k tuition 30k salary 50k overhead) of grant money (not including lab expenditures). Of that roughly 70k goes to the university (tuition+overhead) and 30k goes into the student's pocket. Hiring a highly motivated, talented, educated, research scientist for 30k a year is an excellent deal for the gov't. I'd like to know how the university spends the rest of that grant money, and determine whether it's an efficient use of tax payer dollars.


We probably need to rethink Discount Rates in order to give more value to future economic benefits - this was one of the innovations fo the Stern Report on Climate Change.

The problem is not dependency on government grants - we can make them as flexible as we like.

One change since the 50s is a more aggressive and short-termist financial system. Shareholders are in charge and they demand short-term profit - this is bad for R&D investment.


I think the answer is obvious: all scientific research should be modeled after JJ Abrams' renowned long-term project, LOST. Americans have willingly (and enthusiastically) invested upwards of five years into this endeavor. And, I think I speak for many when I say: I'd HAPPILY up that number to ten if it would increase the number of questions answered by the final results.

...a ridiculous answer...or, is it?

James Hardaway

One good quantifiable explanation to this question would be to take a look at the work done by economist Paul Romer in the early nineties. He hypothesized a relationship between the productivity (A) from the Solow model of economic growth in which he includes variables for the number of people in the labor force actually working in labor, and another for the number of workers involved with research and development projects. He points out that countries with higher numbers of research and development workers have faster rates of growth; additionally, unlike the solow model, which has the growth rate eventually fall to zero, Romer argues that a society which has dedicated resources to research and development, will have potentially limitless growth. Romer's works are available on his website as a professor at Stanford, as well as on JSTOR; most helpful would be his work on Endogenous growth.

Hope this helps



I'm looking forward to possible answers, but one complication I'll raise is the question of how the funding model might prefer low-risk/low-reward research with high-risk/high-reward research on some 'big questions.' Currently, people are offering major prizes to get people to go after the big questions.

The other thing is this: while this questions addresses individual project funding, the other thing shaping research interests is career impact. If universities (and other research centers) continue to reward 'lots of papers' over one really good paper that takes longer, then the project-funding approach will only go so far.

(Obviously, I oversimplify the issue, but that's just to draw the distinctions.)

Randy Pistor

Rafe--why don't you pose this question to your cousin Randy? Get back to doing your homework!
Uncle Randy

David Heigham.

ANK is getting to it. The underlying economic justification for research is that people value the output. For example, quite a few of us are happier now that we can pin down the age of this universe reasonably well. If that information ultimately enables an increase in measured Gross World Product which would not otherwise have occurred, fine; but that gain would be a by-product.

The primary need is a variety of economic modes by which we can pay for looking further over the frontier of existing knowledge. Taxation is one. Organised voluntary effort is another. Offering prizes for solving Fermat's Last Theorem or through the Nobel Foundation is another.

All these modes and more have been around for some time. What has got people hot and bothered now is that the by-product consequential economic gains of fundamental research have turned out to be pretty terrific. That makes us even more willing to pay for increases in knowledge; so people ask how much more? and what is the most effective way to pay it? The answers are: pay as much as you feel like, because there can be no useful way of estimating how much economic return you get for the marginal billion until well after the event; and spread your bets on the ways you pay because in advance (or even with hindsight) you can only guess which method of paying or combination of methods is likely to produce the highest possible economic return.

How much do we feel like paying? There are established methods for roughly estimating that in aggregate; like there are methods of finding out how much we feel like spending on defense and security. But these estimates have and will have no relation to the ultimate aggregate economic benefit that is delivered in the future. Inferring that the benefits of the marginal chunk of desired spending is equal to the value of what would be foregone to fund that spending is tautologous, not a guide to decision.



I'd phrase the question as: Looking at the great benefit to mankind that scientists who were not capitalists produced; has capitalist science been as productive? After all, we have 400 hundred years of history to examine for an answer. I'll defer to better historians, but my amateur opinion would be that capitalist science is practically an oxymoron.