Multi-party international joint ventures: Multiple post-formation change processes Chris Changwha Chung a, *, Paul W. Beamish b,1 a Korea University Business School, Korea University, Seoul, Republic of Korea b Richard Ivey School of Business, University of Western Ontario, London, Ontario, Canada 1. Introduction An international joint venture (IJV) is a separate legal organization that represents the joint equity holdings of two or more partners, in which the headquarters of at least one partner is located outside the country of the venture operation (Shenkar & Zeira, 1987). Although this widely used definition does not restrict the number of partners, most research only considers conventional two-partner IJVs that are typically comprising one foreign and one local partner (i.e., dyadic IJVs). A substantial number of IJVs however involve three or more partners (i.e., multi-party IJVs). Makino and Beamish (1998) found that 55% of 737 Japanese IJVs they reviewed involved three or more partners. Garcia-Canal, Valdes-Llaneza, and Arino (2003) reported that 49% of a sample of 80 Spanish IJVs involved three or more partners. In Gong, Shenkar, Luo, and Nyaw’s (2007) sample of Chinese IJVs, approximately 30% had three or more partners. Despite the prevalence of multi-party IJVs, researchers have seldom examined the effect of partner numbers in IJVs. When the number of partners is included, these numbers are frequently used as a control variable (Hennart & Zeng, 2002; Makino & Beamish, 1998), or they are simply dichotomized between dyadic and multi- party IJVs (Garcia-Canal et al., 2003). Only a few researchers exclusively focus on multi-party IJVs and use the number of partners as a key construct (e.g., Gong et al., 2007). Even then, the researchers who use the partner number as a key construct typically use this variable as a static predictor on governance choices (Garcia-Canal, 1996; Garcia-Canal et al., 2003) and venture performance (Beamish & Kachra, 2004; Griffith, Hu, & Chen, 1998). The most significant paucity in multi-party IJV research is that it has been limited to a static context. Little attention has been paid to analyzing dynamic post-formation change processes that occur in the middle stages of multi-party IJV evolution. Most IJV research focuses on the early stages of IJV development, including IJV formation. As the IJV research field develops, researchers have shifted their focus from the formation stage of the IJV to its end destinations (e.g., termination); however, these researchers pay little attention to analyzing the middle developmental stages. A new stream of research that focuses on post-formation processes has emerged. However, this research is largely based on deductive theoretical approaches and conceptual models (Das & Teng, 2002; Khanna, Gulati, & Nohria, 1998; Koza & Lewin, 1998; Kumar & Nti, 1998; Ring & Van de Ven, 1994), 2 and includes a few case-based studies (Arino & de la Torre, 1998; Brouthers & Bamossy, 2006; Journal of World Business 47 (2012) 648–663 A R T I C L E I N F O Keywords: International joint ventures Multi-party complexity Multiple structural changes Performance and survival A B S T R A C T Research on multi-party IJVs has been limited to a static context. Little attention has been paid to analyzing dynamic post-formation change processes. This study investigates the evolving influences of multi-party IJV complexity on performance in a dynamic context where the multi-party IJV goes through multiple waves of structural change. Using a static context, some previous studies found support for a negative impact of multi-party complexity on performance, while others did not. Analyzing 2652 multi-party IJVs over a period of 17 years, we attempt to reconcile previous work by investigating whether there is a threshold beyond which the negative impact of multi-party complexity on performance becomes salient. ß 2011 Elsevier Inc. All rights reserved. * Corresponding author. Tel.: +82 2 3290 2624; fax: +82 2 3290 1307. E-mail addresses: chungc@korea.ac.kr (C.C. Chung), pbeamish@ivey.uwo.ca (P.W. Beamish). 1 Tel.: +1 519 661 3237; fax: +1 519 661 3700. 2 To study the three alliance evolution stages of negotiation, implementation, and adjustment, Ring and Van de Ven (1994) developed a process framework for analyzing repeated interactions in each stage. Kumar and Nti (1998) proposed an alliance developmental path theory for examining outcomes following environ- mental or major strategic changes instigated by one or more partners. Koza and Lewin (1998) presented an alliance development and success framework reflecting co-evolving strategies among partner firms; changing competitive, organizational, and institutional environments; and adjustments in alliance goals among managers. Khanna et al. (1998) established a theoretical framework addressing learning alliance dynamics, and used it to study efforts to create alliance structures that result in optimal private and common benefit configurations. Das and Teng (2002) used their alliance process model to analyze co-evolutionary dynamics that reflect the interests of partner firms and to clarify transformations of alliance conditions over different stages. Contents lists available at SciVerse ScienceDirect Journal of World Business jo u r nal h o mep age: w ww.els evier .co m/lo c ate/jwb 1090-9516/$ – see front matter ß 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jwb.2011.08.001