Friday, September 26, 2008

Why government MUST invest in fundamental research

So why are we so concerned with government funding levels. What if, say, the next president of the US announces that ALL federal spending on research is effective immediately cut to zero - NSF, NIH, DOE, DARPA - you name it.

What's so bad about that? Just let the free market with its invisible hand work things out. Want to study RNA interference, stem cells, supersolid Helium, new Fe-based superconductors, spintronics? Formulate a business plan, pitch it to Venture Capitalist fund, and if they are willing to invest - good. If not - too bad, this means your ideas are obviously wrong.

There are several problems with the model where "free market" takes care of fundamental research. First problem is the time scale. Free market works wonders if the feedback loop is relatively fast. If your invest $1,000 dollars today, on a chance to make it into $10,000 in a year, you might feel good about it. What if you have to invest today, but wait for 10 years to learn whether or not you get a return? 20 years? How about 50 years? 100 years? Even if the return rate was 1,000 fold, or even 1,000,000 after 100 years, you may not care because you will be dead by then.

Problem number one is that it typically takes 20-30 years - in best case scenario to see "returns" on fundamental discoveries. Often discoveries pave the way to other discoveries, and so forth, which eventually trickle down to technological applications - often unintended or unforeseen.

Find me a company that is willing to invest in a project with no hope of return for 20-30 years and I have a bridge in Alaska I want to sell you.

The second problem is that currently there are very few intellectual property mechanisms to protect fundamental discoveries. Think about discovery of GMR which is the basis of every hard drive, or NMR (aka MRI), or transistors. How much money do you think Gruenberg and Fert got from hard drive industry? What share of industry do they have? What about Stuart Parkin - surely he must make a few dollars on each hard drive sold, right?

Similarly, Einstein family must be getting residuals for every GPS unit sold, including iPhones, correct?

Unfortunately (for scientists), but fortunately for the world economy and progress of innovation, most fundamental discoveries are entirely "open source". And that's the way it should be. The downside is that investing in fundamental research is not going to be very profitable if we were to share the results with the rest of the world community. And sharing of results is crucial for scientific progress.

So what would happen if the government funding disappeared? I am sure scientists would survive by doing applied research, and probably be A LOT richer as a result. I have worked in VC-funded startup company and I think most scientists would do very well in that environment. And there will be no short-term negative consequences, if anything, the economy will get a substantial boost from a lot of smart people making "ready for market" devices instead of curiosity-driven discoveries.

The big problems will start in about 10-15 years, when China, or EU, Singapore, Korea or Japan will suddenly find themselves as the new economic force, fueled by investments in fundamental discovery. Not only US won't be manifacturing anything, they won't be inventing anything either. Just like biotech companies cluster around academic centers - like San Francisco, Boston and North Carolina triangle, the future economy will follow where the technology goes, and the technology will go where science is valued.

The incredible thing is that US government can boost science on the cheap - incredibly cheaply - in all the letters we write to congress, we are arguing for a mere billion dollars or so - across ALL disciplines. I dare you to find another example where a billion dollars can make such a significant difference in the future of the country.

9 comments:

Mark said...

Speaking of condensed matter physics -- there are no fundamental laws left to be discovered here. We can say a lot of things about future returns on fundamental science but condensed matter physics is no longer part of the latter. I will say an unpopular thing but we have reached an enormous overproduction of researchers who basically simply hung around a new promising direction (like graphene, which hasn't produced a single true puzzle but attracted herds of theorists and experimentallists for a simple sad reason that they have nothing better to study). One billion would not make us happier. It will simply lead to us reproducing new bigger herds of researchers. As one nice article from mid-90's (can't remember the title, but many have read it) pointed out, exponential growth always comes to an end. Federal government spends a lot more on fundamental science now than 40 years ago, but we make much fewer fundamental discoveries now.
There is an element of hypocrisy in our begging Congress for more money, but come on, let us ask ourselves honestly -- are we worried about benefits to the society only, or may be we harbor some personal agenda (want to live well)?

Zodia said...

Mark, If we really want to live well, then why did we choose to do physics in the first place?? It's hard, competitive and the its monetary outcome is nothing compared to what engineers, doctors, lawyers, computer scientists get doing their jobs!

I feel you're just playing devil's advocate :)

The drift to graphene research is natural. It is due to the extreme difficulty of the current main problems in CMP (think about HTS). A new problem might help shedding some light on the nature of long existing problems.

PonderingFool said...

In terms of where a billion could do a lot of good, I would argue lead clean-up. My wife is a special ed teacher. It amazes me how many kids she gets who are severely intellectual challenged due to high lead exposure. Where I went to graduate school, people were still signing waivers that they understood the paint on the walls was lead based.

Science though is also a great to fund. Studying the unknown yields results we can not even imagine. One of the downsides of the genomics work is though the short term, easy frames to get people to get behind it were oversells. It is beginning to bite us in the rear now.

okham said...

Speaking of condensed matter physics -- there are no fundamental laws left to be discovered here.

That's right, otherwise you or I would have discovered them by now... :-)
Mark, what are "fundamental laws", anyway ? Can you give me an example of any scientific discipline or subfield that would be worth pursuing, at that rate, other than particle physics or string theory ?
Are there new fundamental laws to be discovered in biology or geology ?

Mark, with all due respect, if you have lost interest in your field of study why don't you just quit, and leave room for someone else who hasn't, instead of advocating the demise of the entire enterprise, simply because it no longer excites you ?

Mark said...

Zodia -- quality of life is bigger than just salary. We enjoy having a relative freedom in our choosing of what to work on (even within constrains of funding agencies, tenure-track requirements, constant pressure from competitors, etc.), we like not having a boss, flexible schedule, travelling, respect that comes with a professor's job and many other little things. Many of us are simply scared by the unknowns of changing a field/profession late in our careers, so we prefer just to talk about how we would have been richer had we chosen a more lucrative job.

You are absolutely correct about graphene. The troops of CM physicist surrounded graphene not because it is of any relevance to the society now or in the distant future, but bacause that's what they can do it, nothing more. Graphene is at most applied physics but we prefer to pretend it is of great fundamental importance and should be funded by government. Let me be silent about spintronics bubble which evaporized as quickly as it inflated, except for some more reasonable (=applied) things in magnetic semiconductors.
And for the record -- I wrote papers both on graphene and spintronics.

Okham -- I am easy on fundamentality. Kondo, weak localization, conductance quantization, quantum Hall effect -- I agree to count those discoveries as fundamental. Please name discoveries of similar level made in XXI century.

Quit... you know, I am thinking about it. But leaving my personal happiness aside, do you think not having dissenting voices would benefit our profession? Besides, who knows what career I could chose then? May be I will become a scientific adviser for a new President. Will it be better to have me pissing into the tent from above then?

Incoherent Ponderer said...

Mark - come on, seriously?

It is usually string theory types who declare that once they unify all forces there is no scientific questions left to resolve.

How much of condensed matter do we actually understand? Just because we can solve Schroedinger equation for a particle tunneling through a square potential doesn't really mean we can say much about 10^23 interacting particles.

There is a HUGE amount of condensed matter discoveries to be made in 21st century, and I hope you are just trying to be a contrarian. Do you seriously argue that we know (and understand) everything there is to know when it comes to condensed matter systems? Surely you can't be saying this.

I don't think you have to be a professional physicist to understand how far we came economically in the last century. Just compare the average lifestyle in 19th century (which was not all that comfy even for super-rich) vs. now. Much of it due to advances in technology. Much of the our GDP is due to technology. But unfortunately most people don't understand and don't see the connections between, say, what Gruenberg and Fert did and the fact that they can store huge amounts of information on their PCs.

So this point can and should be made repeatedly.

Pondering fool - good point on lead poisoning. There is a broader similar point that can be made though. There are many more examples of environmental challenges that we surely will face, and their solution will require new technologies. If we were to disinvest from science, these technology challenges will be much more difficult if not impossible to resolve.

okham said...

do you think not having dissenting voices would benefit our profession?

Mark, "dissent" is when you pointedly and reasonably criticize my choice of problem to investigate, or investigative approach, or methodology, or take issue with my interpretation of the data or theoretical framework... that is indeed a necessary contribution, and much good can and does come from it.

But saying that "it's all boring", "nothing is left to discover", that herds of people with nothing better to do are just wasting their time and taxpayer money, is not dissenting.
It's gratuitous, wholesale demolition for its own sake; it's like setting the house on fire to enjoy looking at the pretty flames... in the process possibly demotivating and turning away from science many young people.
That, to me, is neither necessary nor particularly noble.

Please name discoveries of similar level made in XXI century.

You mean, over the past eight years ? Supersolid helium (if confirmed, and I have to admit I am skeptical) is at least as important as the ones you have mentioned.

Incoherent Ponderer said...

Supersolid He, new class of Fe-based high-Tc superconductors, metamaterials, quantum optics/information/teleportation experiments, quantum oscillation studies of fermi surface in high-Tc SC, quantum spin hall effect, new physics at oxide interfaces, and yes there is this thing called graphene - and these are just some from hard condensed matter. Never mind soft matter, granular physics, biophysics etc.

Looking back there will be many more discoveries over the past 8 years that we may not appreciate yet.

What is the best movie of 2008? What is the historical significance of bailout package? Was picking Palin a mistake for McCain?

Too early to tell.

Doug Natelson said...

Somehow I missed this discussion while traveling. IP and Okham don't need any help from me, though, and IP has listed a bunch of interesting CM developments in the last 8 years. Indeed, as far as I can tell, more exciting stuff is going on in CM right now than during most of the 1990s.