An interesting point from Ezra as an afterthought on his recent appearance at Cato:
Towards the close of the event, someone in the audience argued that moving towards a national system would hugely retard medical innovation. I never understand the evidence for this claim. A huge amount of the tech advances come from public institutions and grants even now. The Veteran Administration’s development and integration of the medical information software VISTA is one of the great delivery advances in recent decades. Medical development only increased as Medicare has expanded, in large part because bringing more seniors into the pool increased the potential profits to be made. Indeed, you could create a national system and plow the savings directly into NIH grants, hugely accelerating innovation. There’s just no evidence on the other side of this innovation question — and there’s quite a bit that the current system encourages a lot of wasted R&D, like into molecularly dissimilar but functionally identical copycats of patented blockbuster drugs.
First, the argument that Medicare had a positive effect on innovation is at least contestable. But supposing that it did, it seems a little odd to, on the one hand, suggest that this boost to innovation was the result of having “increased the potential profits to be made,” while on the other hand professing befuddlement at how proposing to lower costs by “squeezing the providers” might be thought to put a damper on innovation.
Ez proposes plowing the savings realized by government bargaining into the NIH, but for this to achieve net savings, you’ve got to believe that the NIH grant process is going to be more efficient at the margin than market incentives. This isn’t my issue, so I don’t know how likely that is to be true, but it does seem reasonable to expect diminishing returns to budget expansion here. That is, the first dollar of NIH grant money presumably goes to the easiest call, the most obviously worthy effort to conduct unpatentable basic research with broad benefits, etc., but progressively less so for the subsequent dollars.
Anyway, what I found particularly interesting was the point about duplicative research. On the one hand, stronger patents provide more research for R&D. But if they’re too long, many companies may determine that if a drug is going to be profitably under patent for a while to come, it makes sense for them to devote time and resources to tweaking a few molecules and developing a functional equivalent. That’s not only wasteful in itself, it dilutes the additional profits longer terms are supposed to provide, since the makers of the original drug now have to compete with the copycat. So you might imagine conditions under which shorter terms actually yield a triple benefit: Consumers get cheap generics sooner; pharmaceutical companies don’t invest as heavily in creating unnecessary copycat drugs; and the returns to the original drug maker are only slightly diminished, because the shorter term is offset by the uncontested monopoly they enjoy for the whole of the term. The trick, of course, is figuring out just what those conditions are, and what the optimal term is.
2 responses so far ↓
1 Timon // Apr 17, 2007 at 10:27 pm
The thing about patents in a hyper-deflationary area like biotech or software is that no matter how you align the incentives (and remember who is doing the aligning) they will never stay coherent for very long. Today’s vast investment is tomorrow’s bauble, and this is not only true of items with low or zero marginal cost. If the concept is to use government brain- and legal power to reward investment in technology then there is no principled difference between allowing Senators to award grants or monopolies (or laurel wreaths). If anything the grant strategy is at least likely to be used more efficiently, since a monopoly incentivizes inefficient production; for example, Jeff Bezos credibly claims the infamous 1-click patent took something like 35000 engineer-hours to produce. If there were no monopoly to capture he could have just waited for the next version of the standards that now make that procedure trivial, saved the labor and spared himself the ridicule. There is no reason to suppose biotech won’t go through the same intense deflation as the tools to analyze compounds for medicine become ever cheaper, as they have with genes. When that happens it would be nice if there wasn’t a 20-year government blackout on lifesaving ideas.
2 Mike S. // Apr 18, 2007 at 7:36 pm
I posted this in comments at Ezra’s blog:
As a former pharmaceutical research chemist who has just started working as a patent lawyer, here are my thoughts regarding the copycat drug issue:
(1) The most egregious type of market-gaming by pharmaceutical companies occurs when a patent on a successful drug expires, and a successor drug which is incrementally different is launched by the same company. Clarinex/Claritin and Prilosec/Nexium are good examples examples of this. By itself, this is not a problem, because the original will have gone generic, but aggresive marketing practices will lead some doctors to push the new version, and some patients will always demand the new-and-improved version. However, this is not really a problem of “wasted R&D,” but of an seemingly undeserved windfall to pharma. From a patent policy standpoint, these improvements could be easily treated as non-obvious and therefore unpatentable. Of course, development and regulatory costs are so high in pharma that it is questionable whether such improvements would make it to market in the absence of patents. A compromise might be to grant certain types of drug improvements patents having shorter-than-usual terms.
(2) A tougher issue is what former FDA head David Kessler called me-too drugs – chemically distinct drugs with the same biochemical mechanism of action. Here, the objection is that there are too many entrants in big-dollar therapeutic classes where the medical need has already been met, and that the development costs for the later entrants is incurred with little benefit to consumers, who would have been better off if the late entrants had pursued some other un-met medical need. However, it is a debateable proposition whether me-too drugs have had a negative net effect. There is data showing that competition in on-patent drugs has an effect on prices, and there are also examples where patients have benefitted from having a selection of drugs available with slightly different efficacy and side effect profiles. It is also speculative to say that pursuing other targets would have led to drugs in new therepeutic areas.
Furthermore, while this critique may have been accurate during a specific period in the late 80’s and early 90’s, when lots of companies were heavily pursuing a few notable therapeutic classes, it is not clear whether this is the still the case. In my experience, industry emphasis has shifted more towards developing drugs to meet un-met medical need. This is in some part due to public scolding by the FDA, but also in large part due to the fact that the me-too drugs were not very sucessful. It turns out that there is a huge first-mover advantage in drug marketing, and experience has shown that unless they demonstrate significantly improved efficacy or side effect profiles, late-comers to a given therapeutic class don’t sell very well. Now, the recieved wisdom is that the way to make a lot of money is to be the first entrant in a new class.
In addition to the fact that the problem of me-too drugs may not actually be a problem, and that to the extent it is, it may have somewhat corrected itself, things get pretty thorny when you try to come up with patent or regulatory policies to solve it.
The first problem is that the long time-scale of drug development, the significant possibility of failure in late-stage development, the number of competitors in the R&D market, and the relative paucity of new biological targets with large potential markets create a circumstance where there may be a dozen companies or so which have development programs in a particular area where there is unmet medical need, each of which has no idea who is going to be the first to market or are going to make it at all. As some of these competitors are successful, those who complete R&D later may come to be seen as me-too entrants, but it can’t really be said that was their intention, and it can’t accurately be predicted what research or launched drug is wasteful. Also, to the extent that some of the follow-on companies have already expended a significant fraction of the total development cost of their me-too drug by the time it becomes apparent that it is in fact a me-too drug, the cost-benefit analysis may indicate that completing development of the me-too drug is the best option for everyone, even the consumer. To me, at least, it seems that an inherent problem in regulating me-too activity is that it is impossible to identify the parties you want to be researching something else, and by the time you know who they are, you may be better off just letting them finishing development.
Also, because of the advantages of having at least some alternatives in a given therapeutic class, there is going to be an optimal number of alternative drugs in each class. If you want to regulate me-too drugs, you are going to have to come up with a better allocative strategy than having markets, and I really doubt that there is one.
Timon makes an interesting point about deflation in markets for certain technologies. I used to work for a small biotech which funded its internal drug development program through sales of robotic systems for ultra-high-throughput biological screening and combinatorial synthesis. I can say that while such cost-reducing technologies, as well as data-manipulation techniques used in clinical trials, have a revolutionary effect in their given phases of drug research, there are still large research areas that are time and labor intensive and are not amenable to automation. One of the trends there is to offshore more menial laboratory tasks to places like China.