Control of international joint ventures

Prior research has found that many international joint ventures (IJVs) fail. Some authors have suggested that control problems are one of the primary causes of IJV failures. However, little IJV-related research has focused on control issues. This paper reports the results of a largely exploratory study of the control practices employed by partners involved in three arguably successful IJVs. The study found some control-system similarities among the three IJVs, but it also found signi®cant di€erences related to the use of dispute-settling mechanisms, control focus (broad vs. narrow), and control tightness. The paper describes some of the factors that seem to cause the di€erences and, thus, provides a start for a contingency theory of IJV control systems. # 2000 Elsevier Science Ltd. All rights reserved. As markets have become more globalized and competitive, it has become increasingly dicult for any single company to excel, and thus be competitive, in all aspects of a given business. Few companies have all the requisite expertise and specialized knowledge in-house. To compete, many companies have solicited the assistance of other ®rms. Many ®rms have ``outsourced'' some non-core activities, such as computer operations, training, employee bene®ts, or internal audit. And many ®rms have chosen to enter into shared, cooperative agreements with ``partner'' ®rms. Joint ventures, which are sometimes called ``equity joint ventures,'' are one prominent form of cooperative agreement. Joint ventures (JVs) are separate entities owned by two or more partners. A JV is considered international either when at least one parent is headquartered outside of the venture's country of operation or when the JV has a signi®cant level of operations in more than one country. International joint ventures (IJVs) have become more prevalent and more important as the business world has become more globalized and competitive (e.g. Christelow, 1987; Croese, 1989; D'Aveni, 1994). Anderson (1990) observed that more JVs and cooperative arrangements were announced in the period 1981±1990 than in all previous years combined. Sherman (1992) calculated that JVs between US companies and international partners, speci®cally, had been growing by 27% per year since 1985. Even competitors that formerly did not talk to one another are entering into JVs (Templin, 1995). And, importantly, empirical evidence has shown that announcement of and involvement 0361-3682/00/$ see front matter # 2000 Elsevier Science Ltd. All rights reserved. PI I : S0361-3682(99 )00057-4 Accounting, Organizations and Society 25 (2000) 579±607 www.elsevier.com/locate/aos * Corresponding author. 1 The cooperation can also be achieved by other means than equity JVs. A wide variety of contractual arrangements, often referred to as ``non-equity joint ventures,'' can be used to foster cooperation between partners. Examples of non-equity JVs are licensing, distribution and supply agreements, or technical assistance and management contracts. in IJVs has, in general, a positive impact on ®rm value (Hu, Chen & Shieh, 1992; Lee & Wyatt, 1990; Morck & Yeung, 1991). Management of JVs is dicult, however. Morris (1998), for example, noted that JVs require special cooperation from the partners because no one party has total control and that accountants, in particular, must have ``an array of specialized knowledge and skills to function well'' in JV situations (p. xiii). Consistent with the diculty of managing JVs, many JVs encounter performance problems and fail. Gordon Redding, director of the University of Hong Kong Business School, estimated that ``About 50% of joint ventures fail'' (Young, 1994, p. 35). Kogut (1988) showed that 32% of one sample of JVs failed within their ®rst 10 years. And in a sample of over 5000 subsidiaries of 180 large US multinationals, Gomes-Casseres (1987) found that while 16% of wholly owned subsidiaries were ``unstable,'' this percentage was higher Ð 31% Ð for JVs. Instability meant that the entity was liquidated, sold, or taken over by one or more of the original owners. Park and Russo (1996) found that the failure rate is nonmonotonic, rising to a maximum during the ®rst 2 years and then declining. JVs formed between direct competitors appear to have a shorter duration, while JVs formed by partners who have additional ties between them have a longer successful life. Good data about international JV (IJV) failure rates, speci®cally, do not exist. Estimates of unsatisfactory IJV performance have ranged from 37 to over 70% (Geringer & Hebert, 1991). IJV failure rates are probably higher than are those for domestic JVs because IJVs generally face greater challenges. For example, many IJV partners must monitor operations in settings with which they have little familiarity (e.g. markets, distribution systems, legal systems); they must often cope with signi®cant geographical separation and time differences; and they must bridge cultural boundaries (e.g. Brandt, 1990; Brown, Rugman & Verbeke, 1989). Much is yet to be learned about IJVs. As Geringer and Hebert (1989, p. 250) observed, ``Our understanding of international joint venture management lags behind the demands of practice.'' Young (1994, p. 35) concluded, ``Since little published material exists on how to succeed at joint ventures, there is no tried and trusted formula.'' And Foster and Young (1997) stated that ``Research journals have barely scratched the surface in this area [of business globalization topics like joint ventures and technology licensing arrangements].'' Prior writings suggest strongly, though, that many IJVs su€er from poor control practices: Many IJV partners have trouble coping with a partner's con ̄icting interests, an inevitable loss of operating autonomy, di€ering cultures and/or management styles. And a lack of trust between partners sometimes leads to more complex and, hence, destructively slow decision-making processes. This study was focused on IJV control practices which, Geringer and Hebert (1989) concluded, is one of the prime determinants of IJV success or failure. For purposes of this study, having good control over an IJV means that it is highly likely that each of the partners' objectives for the IJV will be met (de®nition adapted from Merchant, 1998). If any partner perceives that an IJV is out-of-control, then it is likely to terminate its involvement in the IJV and bring about the IJV's failure. Little research has been directed at the questions related to what controls are and should be used in IJVs and what causes them to fail. For example, Geringer and Hebert (1989) wrote, ``The issue of control has received only fragmented and unsystematic attention in the JV literature'' (p. 237), and ``managers have received minimal guidance about when and how to use [the various control options], as well as about the potential trade-o€s between alternative control options'' (p. 250). This study was addressed at this void. The central research question was: How do partners exercise e€ective control over their IJVs? Follow-on questions were: How are IJV control systems similar? How are they di€erent?What causes the di€erences? Since IJVs have attracted no attention from accounting researchers, two more speci®c, accounting-related questions were also chosen to guide the research: (1) How important are accounting performance measures in controlling an IJV? (2) What accounting measurement issues do IJV partners face, and how are they handled? 580 T.L.C.M. Groot, K.A. Merchant / Accounting, Organizations and Society 25 (2000) 579±607