This research examines the development of market structures in the high-risk medical device sector in the US, specifically with regards to innovatory activities and their concentration among firms under the influence of regulatory changes. It further examines how actual market concentration is related to innovation concentration.

Innovatory activity counts, based on device and change approvals at the US Food and Drug Administration, and review time-derived innovatory significance measures in medical device markets are processed into innovation concentration measures for 16 medical specialty categories. The innovation concentration measure used is conceptually akin to the Herfindahl-Hirschman Index. The data cover three decades from 1978 to 2007 and are in a first step analysed in panel data regression models for the effect of changes to regulation on innovation concentration. The empirical results demonstrate a positive two year-lagged relation between the introduction of regulatory change and concentration of innovatory activities among firms. The results suggest that as some businesses find it more difficult to deal with regulatory change than others, innovatory activity is temporarily limited to less firms until all firms and the regulatory agency have adjusted to the regulatory change.

In the second strand of analysis it is shown that market concentration is related to itself across years, but also related to innovation concentration, entrepreneurial activities and regulatory change. The relationship to regulatory change is less clear-cut as with regards to innovation concentration, as there is a one-year lagged positive and a three-year lagged negative effect. However, it seems feasible that the introduction of regulatory change has an immediate effect in that it increases market concentration, but ultimately leads to a reduction.

The analyses of innovation and market concentration are combined in a statistical assessment of both dimensions in a seemingly unrelated regression model. It is investigated whether market structure based on innovatory activities corresponds to the ‘real’ market structure, and whether increased 2

innovatory activity by new entrant firms, as well as small existing firms led to decreasing concentration and implicitly more competition in medical device markets in terms of revenue market shares, or whether large incumbent firms have been able to defend their market positions and new, small firm innovation mainly takes place in niche medical specialties and markets.

The paper conceptualises the use of a concentration measure for innovatory activities as an additional aggregate measure of market structure. The different model results support the findings of the research and thus demonstrate that the relationship of regulatory change, innovation- and market concentration is robust to different model specifications.