The Data-Pooling Problem

American innovation policy as expressed through intellectual property law contains a curious gap: it encourages individual research investments, but does little to facilitate cooperation among inventors, which is often a necessary precondition for innovation. This Article provides an in–depth analysis of a policy problem that relates to this gap: increasingly, public and private innovation investments depend upon the willingness of private firms and institutions to cooperatively pool industrial, commercial, and scientific data. Data holders often have powerful disincentives to cooperate with one another, however. As a result, important research that the federal government has sought to encourage through intellectual property policy and through other targeted investments is being held back. This Article addresses this issue by offering three contributions—one theoretical, one empirical, and one prescriptive. The theoretical contribution builds upon legal, economic, and public choice literature to explain why pooling data is relevant to innovation policy and why the level of data sharing in some settings may be suboptimal. This discussion offers a conceptual framework for scholars and policymakers to examine how data–pooling relates to innovation policy goals. This leads to the second contribution: an ethnographic study of private efforts to pool data in an important field of research. This Article focuses on the field of cancer treatment because it is one of the most active areas where efforts to pool data have recently coalesced. Interviews with lawyers, executives, and scientists working at the vanguard of “Big Data” projects in the field of cancer treatment offer a detailed view of how, precisely, data–pooling problems can hinder technological progress. The study’s most significant finding is that impediments to the pooling of patient treatment and clinical trial data are diverse, nuanced, and not reducible to collective action problems that are already well understood by legal scholars and economists, such as the free–rider dilemma. These findings lead to the third key contribution: a set of targeted policy suggestions designed to facilitate data pooling through regulatory action, amendments to federal healthcare legislation, and tax incentives. These prescriptive measures are tailored to address the sharing of health–related data, but they capture an approach that could be applied in other settings where technological progress depends upon data–pooling