Science Investment and The Horizon Problem (Updated)

Apparently, it's "economics of higher education" day here at Uncertain Principles. This time out, we have Steve Hsu on Larry Summers.

(Update: I should also link to this post by Brad DeLong discussing the same article, with good stuff in the comments. And while I'm bashing free-market advocates, here's Jonah on the worst assumption in economics, and Dave on funding research with prizes.)

Steve quotes Summers saying:

In today's economy an outstanding graduate of a leading business school earns a substantially higher salary than a potential Nobel prize winner graduating with a PhD in biology. Several years after graduation the differences are even more pronounced. It should not be a surprise that in light of this economic reality more of our talented young people are not headed towards careers in basic research in the life sciences.

and asks:

Gee, is there a market failure? If life sciences are so important to society, why doesn't the market reward researchers the way it does hedge fund managers? Does government really need to interfere? Obviously I agree with Summers, but I don't think strong believers in markets as resource allocators for society can do so without questioning some of their beliefs.

I'm not a strong believer in markets as resource allocators, but I agree with him. I don't think the current arrangement is at all surprising, though. The market doesn't reward researchers like hedge fund managers due to the same structural flaw that killed Bell Labs.

Bell Labs, for those who aren't steeped in science myths, was the original source of a surprisingly large number of the things that make modern life what it is-- in addition to obvious developments in telecommunications technology (the fax, wireless networking), and electronics (the transistor), Bell Labs researchers pioneered fields from radio astronomy to laser cooling, and invented a host of computer-related technologies, including Unix. It was a large and well-funded research operation run by the Bell Telephone company for much of the previous century, and basically collected really smart scientists, gave them tons of money, and set them loose to invent whatever they could.

It was a fabulous engine of innovation, and it's gone now. There's still something calling itself Bell Labs, but it's a shell of its past glory, and much more tightly focussed on commercial applications. The old system was dismantled years ago, for market-based reasons-- Bell shelled out a huge amount of money to keep the operation going, but didn't necessarily reap the profits of the inventions made in their labs. That doesn't look good on the balance sheet, and so it was more or less gutted in the 80's and 90's, to make the company more profitable. Nobody else has stepped into the gap for the same reason-- there's no way to ensure that a research lab will turn a profit, and thus the investment is really hard to justify to narrowly focussed business managers.

The problem here is that the market system as currently constituted is an absolutely miserable way of allocating resources for long-term projects. The horizon for most corporate evaluations is much to short-- an investment that might pay off in the next five years will be discussed in the finanacial pages as a gigantic gamble, and most decisions appear to be made based on what the effect will be on the stock price this quarter.

That's why a hedge fund manager gets ten times the money (at least) of a top researcher. If a hedge fund manager does his job really well, somebody gets a billion dollars right now. If a researcher does his job really well, his employer gets ten billion dollars ten, fifteen, even twenty years down the road. If the standard for performance evaluation is how much money you made for the company in the past year, it's obvious that the hedge fund guy is going to get paid more, even though his net benefit to society is almost nil. Basic research might transform society for the better down the line, but in the short term, it's all investment and no return.

I don't disagree that market systems are a good way of allocating resources in principle, but the way we're going about it at the moment is badly broken. The short horizon has decimated private research, and is among the root causes of the fact that the US health care system is a joke, and is largely to blame for a number of environmental catastrophes in the making.


More like this

Water confirmed on distant planet - Any lingering doubts about whether water exists on a planet orbiting a star 63 light years away have been quashed by astronomers in the US. The team measured infrared light emitted by the planet, known as HD 189733b, and found distinct spectral…
A couple of weeks back, DougT won this year's Nobel betting pool, and requested a post on the subject of funding of wacky ieas: could you comment on this: and the phenomenon of the uber-rich funding science in general. It seems…
Eugenie Samuel Reich is a reporter whose work in the Boston Globe, Nature, and New Scientist will be well-known to those with an interest in scientific conduct (and misconduct). In Plastic Fantastic: How the Biggest Fraud in Physics Shook the Scientific World, she turns her skills as an…
An oldie but goodie for the connoisseurs of my long political rants (May 11, 2005): -------------------------------------------- I am not an economist. Actually, I have no background in economics at all, if one ignores "Capitalism for Beginners" and "Marx for Beginners". Like all Yugoslavs of my…

While I agree that the time horizon problem is one of the issues that killed what we knew as Bell Labs (and I've always wondered if they call whatever the sub-establishment was that invented the Crunchwrap Supreme "Taco Bell Labs"), I think that Bell Labs had a further problem of "how much money did it make AT&T at all, rather than over a long time frame. AT&T made its money off of a phone service monopoly, which is related to but not the same as technological innovation.

Markets do suck for allocating resources to basic research, but it's in a more interesting way than just having a short time horizon.

Hedge fund managers make a lot of money because no one really wants to be a hedge fund manager, and there are not many of those positions available. (People want the bonuses, not the job.) I'd guess the job is hugely stressful and has a life cycle of a few years, if you're good at it.) Of course, Chad's point about short term and long term money making holds, but that just plays into how stressful the jobs are. Do you want the stress of managing millions of dollars each day? That's thousands of peoples retirement's savings. Screw up and you screw them.

This is in comparison to your average basic science researcher, which is a time-consuming job but not stressful if you're doing it right. Also, research jobs are easier to get, and many are tenured or effectively permanent. I'll stick to computational physics, killing only electrons on occasional nonsensical calculations.

By Upstate NY (not verified) on 29 Jan 2007 #permalink

I've seen too many recent reports that hedge funds don't perform for their clients, i.e. they don't really earn their fees.
Of course ultimately the investment buisiness determines which corporate projects get funded, so an incremental improvement in that selection process could have a large benefit for the economy.

Harvard Business Review 61(6) 195 (1983)

"Toward the end of What does the R&D Function Really Accomplish? [HBR 61(4) 1983] Ulrich Merten and SM Ryu state that 'subsequent performance analyses indicate that the R&D program is more vigorous and more sharply focused' [emphasis in original].

"I submit that if their efforts resulted in anything more than another headache for researchers already beset with corporate politics, that statement would have read, 'Subsequent income statements indicate that the R&D program is more successful and productive.'

"Remember that Detroit eliminated 95% of its lockwasher inventory by substituting a drop of Loctite, not by inventing more lockwashers."

Who was that Korporate Kulture terrorist demanding a full 23 years ago that R&D management is self-serving incompetence? Hey... who's your Uncle?

"That's why a hedge fund manager gets ten times the money (at least) of a top researcher."

For successful ones, more like between one hundred and one thousand times. Of course, some go broke.

By PhysioProf (not verified) on 29 Jan 2007 #permalink

Jake writes:

... I think that Bell Labs had a further problem of "how much money did it make AT&T at all, rather than over a long time frame. AT&T made its money off of a phone service monopoly, which is related to but not the same as technological innovation.

Indeed, query whether the innovations created by Bell Labs didn't undercut the rationale for the AT&T monopoly.

In my experience, the "invisible hand" of market forces usually points in the right (that is to say, desirable for society) direction, but the magnitude of the market forces and the characteristics response times of economic systems are often very far from what one would like. For example: suppose that one wanted to do away entirely with the FAA, and let market forces control airline practices completely. Clearly if an airline's planes fall out of the sky all the time, market forces will act to either shut down the airline or encourage them to fix their planes. However, do we really want to wait for that to happen? Industrial R&D is somewhat similar. The market does reward innovators in the long haul (though even that's debatable), but if the shareholders and board only care about a quick return this quarter, where's the market incentive to invest in innovation?

What Doug describes is part of the larger issue of Market Failures. These are essentially the rough edges where the theoretically perfectly efficient free market meshes with the reality of every day life. Scienceblogs really needs a policy analyst here to go over the basics of public policy, as it has a profound impact on science.

There are several basic types of market failure, where the free market alone is an inefficient allocator of resources. The most obvious of these is the concept of Public Goods. The military is an excellent example of a public good, they will defend American land (including your property) regardless of whether you personally would choose to pay them, thus military defense requires collective action (ie government funding through taxes)...the same is true of fire and police departments. The environment is of course the mother of all public goods (although environmental problems also generally tend to involve an additional market failure called negative externalities, but there's not enough room here).

To an extent, what dooms a lot of R&D funding is the fact that when scientific research develops a truly world-shattering innovation, that innovation often becomes something of a public good. Discovering the ability to transmit television signals over UHF/VHF was an amazing breakthrough, but broadcast airwaves themselves are public goods, and television technology is ubiquitous enough that it may as well be. The paradox is that most paradigm-shifting discoveries have such a great impact specifically because they quickly propagate across the world. The "Open Source" movement is itself an admission of the fact that the vast majority of IT intellectual property may as well constitute a public good.

The one place where this market failure has really been addressed is the pharmaceutical industry. The granting of short-term exclusive patent rights allows a given company to profit off of their discovery, essentially granting a short grace period between innovation and public-good status. A limited-time monopoly is granted to provide incentive for innovation, while also providing for the eventual attainment of semi-public good status for the discovery in the near future.

Government-granted monopolies have also been attempted (with varying degrees of success) in certain transportations (airline, train) and utility (phones, cable, power) industries. Obviously the government provides actual grants for research, but perhaps there is some way to tweak the system so that companies might be allowed to more directly profit from broad-reaching discoveries. I know this sounds too much like a "let's give more handouts to big businesses," but who do you think is going to sign hefty paychecks for those researchers?