February 16, 2009

The Bubble In Academic Research

And history is quite clear on this: once a bubble pops, it never reflates.

From Science:

A controversial policy requiring researchers to make their papers freely available to the public at a U.S. National Institutes of Health (NIH) Web site is facing a potential roadblock. Last week, members of a powerful House committee held the first-ever congressional hearing on the policy and floated a proposal to overturn it.
Breathe.  Think about this. NIH sponsored research was supposed to be freely available in an open access format (PubMed Central), under the intricate logic that if taxpayers funded it, they should be able to read it.  But

Law professor Ralph Oman of George Washington University in Washington, D.C., argued that NIH's policy is a "dilution of the rights of the copyright owners" and "will destroy the commercial market" for science and technology journals.
Well, duh.  That's obvious, and by obvious I mean, of course, a red herring.

Don't make the mistake of thinking this has anything to do with maintaining a journal's revenue stream.  They certainly didn't complain when Pharma sponsored studies had to have their data open access.  So why would they care about NIH open access?

Apparently, it costs a journal $4000 per article to publish it, at 50% gross margins.  I have no idea where that money goes-- the authors get nothing for writing the article, and peer reviewers get nothing for reading them-- but that's enough reason to destroy this unnecessary system.

The mistake is thinking that open access is a threat to a journal.  The trickis: subscriptions are paid by big buyers (e.g. universities), and they don't get to choose what they are subscribing to.  You buy the Elsevier Package, and you get whatever crap journals they happen to be offering.  Say you want only American Journal of Obstetrics and Gynecology- $600/year.  Too bad-- it's coupled to Brain Research, at $23,000/year.  I defy anyone to disagree that there has never been, and never will be, a finding in Brain Research worth $23,000.

Universities will pay whatever it costs to get the crap journals in which their NIH sponsored research is published, even if they also have to buy the other crap journals.

Open access articles isn't nearly as big a threat to publishers as simply unbundling the journals from each other, letting universities decide which ones to buy.

So open access doesn't threaten subscription revenues, it threatens the number of journals they can publish.  It's the multitude of journals, each with its inflated subscription rate, that brings in the real money.  Also, each crap journal carries advertisements; even a crap journal with a small niche can have higher ad rates because of the bundling.

Elsevier alone made $3B in revenue last year.  That's a lot of business to be at risk, by science.


But wait: this is a bill blocking the block the open access policy of the NIH that was put in place last year.  Why weren't researchers putting their stuff up on the internet before that, anyway?  Just for fun?

Why didn't each lab simply create a website and reprint everything?  It would take no work-- I assume they didn't write out their findings longhand with a quill.  It is technically a violation of copyright, but have you ever heard of a journal coming after the author of a paper for violating copyright?  Me neither.

The answer is that a lot of the payoff of research is the publication itself.  Where it gets published often matters much more to the researcher than what was actually discovered.  Research is often conducted with a journal in mind and "what's hot" in a field--- as decided by the members of the editorial board of the journal, who decide based on their particular interests (read: funding streams and prejudices.)   BTW, these are often the same guys who are grant reviewers.  This almost completely drives decisions "on the ground" on what to study and what not to study.  (Or did science sufficiently crack the mysteries of lithium to warrant a complete move to Depakote?)

Would anything in AJP be worth reading if it wasn't actually in AJP?

If you think about it, the entire past 15 years in psychiatry have produced no discoveries at all.  None.  We have different medicines, okay; but they're not better, just different.  We don't have a better handle on the anatomic or genetic or anything causes of anything-- we're not even any further along in defining our terms.  Thousands of articles rehashing the same old ground have kept thousands of academics employed, to no benefit whatsoever for mankind.

Take away the journals and the system collapses.  Force researchers with NIH grants to publish their findings without the marketing and packaging of a journal, and you've effectively halted half of the NIH research, until another generation of researchers with a different research model show up for work.  For sure, unquestionably, you've killed off psychiatry as it functions today.

The academic research system is flawed because it does not incentivize research, it incentivizes the process of research. 

Academic research is a bubble, money keeps flowing into it as long as it produces quality research.  Who decides quality?  Journals are the rating agencies, Moody's, they keep it sustainable by giving it AAA rating.  The ratings agencies are precisely what keeps the bubble inflated, just like with the mortgages, they are what keeps research money pouring into the system. 

If someone could look behind the ratings, and take measure of the actual value of the research, the bubble would pop faster than, well, you get the idea. Then there's the "systemic risk."   Journals collapse, academic centers collapse from lack of funding, Pharma loses the AAA rating on their studies which are done by academics, published in journals, etc.

Research would be forced to change completely-- and for the better.  But you'll have a decade or so recession in science and education while the old generation dies out and the new one becomes old enough to start work.