In the United States and other advanced industrial societies, many economic activities share two features. First, private parties (such as individuals or firms) perform research and development (R&D) of projects, often at considerable expense. Second, political agencies have the ability to veto the entry or use of these projects. This combination of R&D and regulatory deliberation, termed approval regulation, is a significant component of large segments of modern economies. For example, it characterizes the pre-market activity of the pharmaceutical industry, which saw U.S. sales of over $160 billion in 2003, or approximately 1.5% of GDP. This project has two scientific objectives. First, it develops a set of theoretical models. This is essential because there are presently no theories that describe the approval regulation process. Over the past few decades, political scientists and economists have developed models that describe each of the above components-R&D and regulatory decision-making-piecemeal. However, models of R&D do not consider the influence of a strategic regulator, and models of regulatory approval do not consider the strategic development of products for submission. To generate predictions about firm and regulator strategies, it is therefore essential that a theory of approval regulation integrate the two approaches. The second objective is to test the predictions of the theoretical models. These studies focus on the U.S. pharmaceutical market. One of the main challenges will be that of building the appropriate data sets. As the integration of R&D and regulatory decision-making would suggest, this requires collecting data on Food and Drug Administration's (FDA) regulatory decisions, as well as firms' pharmaceutical R&D projects (including those that were never submitted for review). Moreover, some of the critical variables are difficult to measure. Specific issues to be addressed include the incidence of regulatory error and delay, the effects of firm characteristics (such as size) on R&D and regulatory strategy, and the effects of recent FDA reforms.
The project builds on preliminary work conducted at Columbia and Harvard universities over the past two years. This work has resulted in a number of promising preliminary results, both theoretical and empirical. A basic theoretical model has been developed that treats the interaction between a firm and a regulator as a game of incomplete information. The model predicts some counter-intuitive relationships between firm size and regulatory error. These predictions are largely confirmed by the available data. However, more work is needed to address Type II errors (which are inherently more difficult to measure), as well as to incorporate multi-firm competition and a more realistic regulatory review process.
Broader Value: The results of this project will be of interest to both academics and policy specialists. In addition to filling an important gap in the scholarly literature on regulation and bureaucratic politics, the project promises to provide timely and policy-relevant knowledge about a highly controversial issue area. For example, recent controversies over the FDA's handling of COX-2 inhibitors and other products have highlighted the tensions between Type I errors (i.e., approving "bad" products), and Type II errors (i.e., rejecting "good" products). An understanding of the incentives faced by submitting firms and the FDA is crucial for striking a balance between both types of error. The results will also be useful for evaluating the effects of possible policy changes, such as the imposition of submission fees, or the implementation of an independent post-approval monitoring agency. Finally, the data set used in these studies will be a resource for other analysts interested in the topic.
