Perverse incentives that encourage unethical waste

If you had a federal agency in charge of getting scientific discoveries commercialized, wouldn’t you want them out talking to industry?

So when the agency is NIH, what do they do? Thanks to the 2005 NIH ethics rules, they are discouraged from collaborating with industry.

Derek Lowe writes

The reason I'm talking about all this is that I've heard of instances where people from NIH have refused (or felt as if they have had to refuse) invitations to give talks in industrial settings, because they feared conflict-of-interest problems. This seems perverse, especially for an agency that's talking about getting heavily into translational drug research. That'll have to lead to numerous contacts with industry, I think, in order to be much good at all. So how will the NIH manage that if the drug industry is seen as contaminating their Purity of Essence?

and then in the comments “JAB” replied:

As an NIHer, I pretty much agree with Derek that we're actively discouraged from interacting with industry by the current ethics rules. Formal consultancies were prohibited several years ago, and I don't believe that's changed. Most of my colleagues shy away from anything that might require ethics office approval. It IS possible to give a simple seminar at an industrial site, with prior approval, and I know of folks who do so. Formal collaborations under a CRADA are permitted, but that's two orders of magnitude more work to set up.

I don’t want to minimize the importance of being as blameless as Cæsar’s wife, or the very real problems of any private entity (whether business or activists) illegally influencing government decisions.

That said, the idea of — on the one hand — pressuring (or exhorting) collaboration between government, industry and university scientists to collaborate — while on the other hand adding red tape to make that nigh impossible — is just crazy. Crazy.

And while i understand the central role NIH plays in funding, evaluating and disseminating medical discoveries, they’re not the FDA. It’s one thing to say we don’t want regulators mingling with dirty industry — it’s another to say researchers can’t actually go out and promote real translational research.

It’s not clear how to fix the problem without re-opening the same problems that led to the 2005 rules in the first place. In any endeavor, government regulation and red tape comes along because of a bad apple (or barrel or orchard) that (often) leads to overreaction in the other direction. So eliminating the rules is begging for trouble.

Still, once upon a time the government had rules about de minimis benefits. Is providing someone free lunch in a cafeteria — rather than spending $25 in labor to reimburse a $10 meal cost — really going to lead to unethical outcomes? Some of this is just common sense.

The problem is that some of the outside lobby groups really don't want close industry-government collaboration. (The group pressuring NIH director Francis Collins seems to fit into this category, led by all the usual suspects).

To me, it seems unethical to waste taxpayer money with excessive regulation, delay necessary therapeutics and diagnostics, and even perhaps lead to people dying who didn’t need to.

But then, much as David Friedman points out in his economics primer Hidden Order, people focus on the direct benefit of a given government intervention (e.g. preventing one case of fraud) but not the indirect costs (wasting thousands of person-hours of labor for compliance).

Buying biotech firms to kill them

(Cross posted from the Engineering Entrepreneurship blog)

At #IndustryStudies2011 this week in Pittsburgh, I heard an interesting talk about what happens to biotech startups after they are acquired. Panos Desyllas of the University of Manchester presented his study (with two Manchester co-authors) of UK biotech firms acquired 2006-2010 by non-UK companies.

The team studied in depth six acquisitions, interviewing executives from both sides of each transaction and also analyzing five years of trailing patent data. They also traced what happened to the key scientists after the merger by noting their affiliations in subsequent patents.

From this data, they came up with a simple (but useful) 2x2 typology: are the two firms similar in technology and are they similar in capabilities?

The firms might be exploring different technological frontiers. Or the acquired firm might have something that the acquirer does not — or vice versa — whether it be UK marketing by the acquired firm or global marketing by the acquirer. The (plausible) intuition is that complementary acquisition is more likely to create ongoing value than a more directly competing one.

The typology worked as predicted. In the case of acquisitions where both the technology and capabilities overlapped, the buyer closed the acquired company, keeping only an IP expert or two as a temporary consultant to transfer the tacit knowledge.

In discussion during and after the session, we discussed the case where the buyer bought a rival with the sole purpose of killing it. This happens all the time, and in some ways it seems like a special case with an utterly predictable outcome.

The other case I brought up was when the acquisition starts out as being complementary — but the acquired firm gets killed anyway.

In April, Cisco killed the Flip camera line that it bought for $590 million in 2009. Pure Digital founder Jonathan Kaplan was sorry to see Cisco knife his baby rather than put it up for adoption, particularly when it remained profitable.

The other example (from the life sciences industry) was Biogen Idec, billed in 2003 as a merger of equals between two biotech startups, Boston-based Biogen and San Diego-based Idec Pharmaceuticals. However, the failed merger brought the closure of the former Idec operations in San Diego last November, and the layoff of some 300 employees (including a close personal friend).

During Desyllas’ session, we discussed whether the closure was a good thing or a bad thing for the local economy. In true Schumpeterian fashion, the creative destruction makes available skilled talent to the local economy for other ventures. On the other hand, some off the displaced workers may never have a similar opportunity again.

But in the end, we agreed that the pattern proved a familiar point: companies get sold when the owners want to sell — usually when they want liquidity for an illiquid investment. Whether the founder (such as Kaplan) or the venture investors, once the company is sold all bets are off.