The impact of joint venture status on the longevity of Japanese stakes in US manufacturing affiliates

A number of well-known studies report that between one-third and two-thirds of international joint ventures eventually break up. While many generalizations, explanations, and prescriptions have been based on these statistics, their meaning is unclear. First, all foreign affiliates are subject to normal business risk, and to the risk that they will be divested by parents for strategic or financial reasons. Are these risks higher for joint ventures than for wholly-owned subsidiaries? Second, joint ventures may be shorter lived not because of their joint venture status, but because affiliates which are joint ventured have other characteristics that make them more likely to exit. To know whether joint ventures are shorter lived, one must control for the other factors that affect the longevity of affiliates, whether wholly-owned or joint ventured. Third, many joint ventures contracts contain clauses that allow partners to sell their stakes to one another at specific intervals. Because they make exit easier, joint ventures should have shorter lives, but these shorter lives should only be due to selloffs, not to liquidations. It is therefore important to see whether the supposedly higher termination rate of joint ventures stakes is due to a higher rate of selloffs or to a higher rate of liquidations. In this paper we (1) compare the longevity of stakes in joint ventures versus those in wholly-owned subsidiaries (2) while controlling for other factors that affect the longevity of such stakes and (3) while distinguishing between two types of exit, those through sale and those through liquidation. While past authors have addressed these three issues in piecemeal fashion, we believe we are the first to address them simultaneously, We analyze the factors that affect the longevity of 355 Japanese stakes in U.S. manufacturing affiliates. Controlling for all the factors that affect exit rates, we find that Japanese parents are more likely to terminate their stakes in U.S. joint ventures than in wholly-owned subsidiaries. This higher termination rate of joint venture stakes is explained by a higher probability of selling them, but not of liquidating them. Most of the other factors that have been found significant in explaining gross divestment of foreign affiliates do in fact only affect exits through sales, but not exits through liquidations. Hence it is true that joint ventures have shorter lives, but dangerous to interpret this finding as necessarily meaning that they are more likely to "fail".