The New York Times reviews new data on patents from researchers James Bessen and Mike Meuer, with commentary by our adjunct fellow John Duffy (Mike Meurer and John Duffy will be at this year's Aspen, BTW, if you wish to question them in person).

Jim Bessen and Mike Meurer will present in their new book a neat graph, which shows that over the last ten years or so the cost of litigating patents has ramped up so high that the game is now not worth the candle. That is, the costs are now outweighing the benefits.

This is interesting not only for the conclusion and their policy recommendations, but how they got there. I skim a few of the issues here:

--They arrive at the value of today's patent portfolios using stock market data for publicly traded firms, and also use stock market data to measure the impact of litigation when a case is first filed The data sets from some of Jim Bessen's earlier work were controversial and rather limited, but this new data is less susceptible to that critique. However, it does leave out small firms that are not publicly traded, an important omission as many patent-holders are among these smaller firms.

--Defenders of patents are likely to respond with data showing that the increase in patent litigation has not been all that dramatic. In some respects, it has not been (Peter Detkin and Jim Bessen debated this here in an earlier blog). But the key figure--which no one has, including Bessen & Meurer-- is settlement data. Most suits are settled. They do not get to litigation. This is the data that we really need. Bessen & Meurer's stock market data is a very rough proxy for what the market thinks of the impact of suits being filed on economic value, though, it is better than counting lawsuits.

--My take: Even if the Bessen/Meurer measures of total patent value are too low, their data pointing to a litigation problem is well worth paying attention to. Because of the way that patent litigation works, with a presumption of patent validity that requires clear and convincing evidence to overturn, and a lack of loser pays, the incentives to bring weak suits are strong, and the incentives to defend rather than settle are weak. There are many unknowns in patent reform, but I don't think we can go too far wrong with proposals that tackle the litigation end first.