Bessen and Maskin have a lovely paper about sequential innovation. I have reviewed this over at Najecon. There are two key ideas in this paper about why patenting may lead to less rather than more innovation. First, innovators will generally have more information about the value of their invention than existing patent holders whose licenses they will need to build their own project. This prevents efficient licensing by existing patent holders. Second, competition is not likely to dissipate all profitability from a new invention - this is a point that Michele and I have emphasized. Michele and I have also pointed out how the need to license many different patents further inhibits innovation.

The relevance of sequential innovation is brought home by the patenting of the human gene. Consider Jensen and Murray's empirical investigation. Money quote:

Our results reveal that nearly 20% of human genes are explicitly claimed as U.S. IP. This represents 4382 of the 23,688 of genes in the NCBI's gene database at the time of writing (see figure, right). These genes are claimed in 4270 patents within 3050 patent families (28). Although this number is low compared with prior reports, a distinction should be made between sequences that are explicitly claimed and those that are merely disclosed, which outnumber claimed sequences roughly 10:1. The 4270 patents are owned by 1156 different assignees (with no adjustments for mergers and acquisition activity, subsidiaries, or spelling variations). Roughly 63% are assigned to private firms (see figure, above). Of the top ten gene patent assignees, nine are U.S.-based, including the University of California, Isis Pharmaceuticals, the former SmithKline Beecham, and Human Genome Sciences. The top patent assignee is Incyte Pharmaceuticals/Incyte Genomics, whose IP rights cover 2000 human genes, mainly for use as probes on DNA microarrays.

Imagine if you will the not unlikely case of a new pharmaceutical product that requires rights to the entire human gene.