An international multilevel analysis of product innovation Daniel Lederman Development Economics Research Group, The World Bank, Washington, DC, USA Correspondence: D Lederman, Development Economics Research Group, The World Bank, 1818 H Street NW, Washington, DC 20433, USA. Tel: þ 1 202 473 9015; Fax: þ 1 202 522 1159; E-mail: dlederman@worldbank.org Received: 29 June 2007 Revised: 29 December 2008 Accepted: 19 January 2009 Online publication date: 28 May 2009 Abstract This paper studies multilevel determinants of product innovation by incumbent firms with data from 68 countries, covering over 25,000 firms in eight manufacturing sectors. The author assesses the predictions of interdisciplinary research on firm behavior in different contexts, which have emphasized that product innovation is affected by three sets of factors: the extent of global engagement, information spillovers, and market structure. The empirical model of the probability of observing a product innovation, a probit model, by a firm considers three levels of analysis: firm characteristics, industry characteristics, and the national context. The econometric evidence supports the global engagement and information spillovers hypotheses, but the evidence on the role of market structure is mixed. Regarding global engagement, the evidence suggests that product innovation by incumbent firms is positively correlated with the act of investing in R&D and licensing of foreign technologies, and country-level average import tariffs reduce the probability of product innovation. Regarding information spillovers, the evidence also shows that a country’s patent density is positively correlated with product innovation. In contrast, concerning the market structure hypotheses, country-level indicators of business density and of policy-induced costs of entry are not robustly correlated with product innovation. Journal of International Business Studies (2010) 41, 606–619. doi:10.1057/jibs.2009.30 Keywords: new product development; innovation and R&D; economics of innovation; levels of analysis INTRODUCTION It is so widely recognized that innovation is a key driver of economic growth that it is a cliche ´ to say so. Existing literatures from economics and international business (IB) on innovation and firm performance in different contextual environments can be divided into three distinct hypotheses, which we call the global engagement, the information spillovers, and the market structure hypotheses. The global engagement hypothesis posits that firms engaged in global business activities, through foreign investment, adoption of foreign technologies, or through imports of capital goods, are more likely to undertake innovations than other firms. The information spillovers hypothesis suggests that firms can learn from aggregate accumulated knowledge even if they have not made the research and development investments to produce this knowledge. The implication is that firms that have broader access to commercial knowledge, including by adopting foreign technol- ogies, will tend to have higher propensities to innovate than Journal of International Business Studies (2010) 41, 606–619 & 2010 Academy of International Business All rights reserved 0047-2506 www.jibs.net