Overcoming institutional voids via arbitration

Extending the literature on institutional voids, we introduce theory from law that highlights the ability of firms to choose the laws and enforcement mechanisms that govern their international joint ventures (IJVs). Specifically, firms may overcome institutional voids by borrowing institutions via binding international commercial arbitration (BICA) rather than relying on host-market institutions. Leveraging an institution-based view, we develop a theoretical framework to articulate the conditions under which IJV partners may choose BICA as opposed to domestic courts to overcome institutional voids in host markets.


INTRODUCTION
A leading proposition in the institution-based view suggests that ''institutions directly determine what arrows a firm has in its quiver as it struggles to formulate and implement strategy'' (Ingram & Silverman, 2002: 20). The cross-border setting intensifies these struggles because multinational enterprises (MNEs), especially those interested in joint ventures (JVs) with local firms, may face formal institutions in host markets that are not sufficiently developed (Luo, 2001). In other words, firms may face institutional voids-defined as the failure of market-supporting and contractenforcement institutions to efficiently facilitate exchange between firms (Khanna & Palepu, 1997). Overcoming institutional voids, firms on the one hand can rely on informal institutions in host markets to partially substitute for the lack of effective formal institutions (Ahlstrom & Bruton, 2006;Puffer, McCarthy, & Boisot, 2010). On the other hand, better market-supporting formal institutions may develop in host markets (North, 1990;Peng, 2003). However, are there other ways to overcome institutional voids in host markets? Can international joint venture (IJV) partners opt out of the problematic institutional framework of host markets altogether?
We address these questions by integrating theory from law, a discipline that is not often leveraged in international business (IB) (Cheng, Guo, & Skousen, 2011;Lan & Heracleous, 2010). In the context of market entries via IJVs, the laws underlying the contract and its enforcement are critical (Beamish & Lupton, 2016;Lumineau & Malhotra, 2011;Zhou & Poppo, 2010). The IB literature frames such contracting around an assumption that we identify as ''one law, one court in one country.'' Specifically, the contracts literature underlying IJV research assumes that one domestic law applies to the IJV, and that the contract will be enforced in the domestic courts of the host country. In turn, IJV partners would be forced to cope with any institutional voids of the host country. We contribute to this literature by offering a mechanism for IJV partners to overcome institutional voids through institutional borrowingthe explicit use of institutions from outside the domestic institutional environment. We focus on one such mechanism from legal theory. Specifically, international disputes may be resolved through public ordering via domestic courts or a hybrid of public and private ordering via binding international commercial arbitration (BICA) (Drahozal, 2009). BICA is a legally binding dispute resolution process that substitutes for domestic courts. 1 By introducing BICA, a widely accepted institution that substitutes for domestic courts, we begin to challenge the traditional contracts assumption of ''one law, one court in one country'' underlying the IJV literature.
In other words, IJV partners do not have to rely on domestic courts in the host market for their enforcement mechanisms (Siegel, 2009). Specifically, legal theory suggests that BICA allows foreign MNEs and host firms to specify laws that apply to IJV contracts (beyond host country laws), and tailor their enforcement mechanisms to their needs (Davidson, 2010;Drahozal, 2009). In turn, this may allow some (although not all) IJV partners to embed strong institutions in IJV contract by choosing to include BICA, while performing the contractual duties in host countries with institutional voids.
We focus on the ex ante enforcement decision between a foreign MNE and a host firm engaging in IJV negotiations. 2 This is a critical stage of IJV formation when partners negotiate the remedy for possible enforcement complications (Reuer & Ariño, 2007). Institutional differences between the foreign MNE's home market and the host market often magnify the transaction costs and emphasize the need for mutually agreed-upon enforcement mechanisms (Xu & Shenkar, 2002). Enforcement mechanisms, in turn, have implications for IJV performance, durability, and survival. In part, this is because IJVs have the advantage of strong internal alignment features that are often complemented by external enforcement mechanisms (Beamish & Lupton, 2016;Dhanaraj & Beamish, 2009). For most IJV partners, the inclusion of arbitration clauses in IJV contracts is paramount. Conservative estimates of the ex ante inclusion of arbitration clauses in IJV contracts is around 85% (Drahozal, 2009;Mistelis, 2013). One rationale behind this effect is the shifting importance of the IJV and the motivations of partners (Reuer & Ariño, 2007;Siegel, 2009). Another rationale concerns the uncertainty in the exit process in an IJVexit can take very different forms ranging from liquidation to acquisition of a partner's share (Beamish & Lupton, 2016). Thus concerns over exit processes from the IJV and enforcement may thus be key in the selection of BICA over domestic courts. Yet we lack theory on why IJV partners choose BICA over the default, domestic court enforcement option in their ex ante discussions, thus calling for additional research.
When forming the IJV in an environment of institutional voids, enforcement is crucial (Khanna & Palepu, 2010;Siegel, 2009). Given the importance of the IJV relationship and the ability of the partner firms to exit the IJV, it is fruitful to ask: What are the conditions that will make it beneficial for firms to choose BICA when engaged in IJVs? According to transaction cost economics (TCE), (domestic) arbitration helps reduce transaction costs in many settings, such as contractual incompleteness, specialized governance, and expertise in ''filling the [contractual] gaps'' (Williamson, 1975(Williamson, : 75-77, 1985. These discussions tend to take place within a domestic context, where the assumption of ''one law, one court in one country'' naturally makes sense (Argyres & Mayer, 2007;Lumineau & Oxley, 2012;Poppo & Zenger, 2002;Weber & Mayer, 2011;Zhou et al., 2014). Given that the assumption no longer holds in a crossborder setting, firms engaged in IJVs have a unique opportunity to decide between domestic courts and BICA (Drahozal, 2006;Drahozal & Ware, 2010;Kapeliuk, 2012).
We develop a theoretical framework to argue that IJV partners can overcome institutional voids via BICA because they can borrow institutions. Fundamentally, firms' choice of BICA may support the IJV even when formal institutions in a host market are not effective (Drahozal, 2009;Mistelis, 2013).
Highlighting a theory of borrowing institutions that allows IJV partners to overcome institutional voids, this article endeavors to make three contributions. First, leveraging legal theory, we identify a gap in the assumption of ''one law, one court in one country'' in the contracts literature on cross-border exchanges. We argue that relaxing the assumption of choice of dispute resolution mechanisms (enforcement mechanisms) presents an underexplored pathway for facilitating IJV relationships by reducing cross-border transaction costs. Second, we contribute to the discussion on institutional borrowing as a mechanism for overcoming institutional voids by showing that BICA helps IJV partners overcome institutional voids of enforcement. Given BICA's multilateral institutional framework, we argue that the joint efforts of 152 countries support enforcement outside the scope of domestic courts in one host country. Third, we explore the boundaries that may provide a more nuanced approach to our theory of institutional borrowing with a focus on IJV partners' choice of BICA in light of institutional voids, search and negotiation costs, and partner constraints. Ultimately by focusing on a specific institution (BICA) and the ability to overcome institutional voids, we contribute to the institution-based view (Meyer & Peng, 2016;Peng, Sun, Pinkham, & Chen, 2009) by showing how institutions matter.

BICA AND ENFORCEMENT COSTS
When firms engage in IJVs, they are susceptible to institutional differences (Henisz & Zelner, 2005;Kostova & Zaheer, 1999;Meyer & Peng, 2016;Peng et al., 2009). Institutional differences magnify the effects of inefficiencies in market exchanges and emphasize the need for enforcement mechanisms (Xu & Shenkar, 2002). 3 IJV partners may engage in a cost-benefit analysis balancing informal and formal arrangements in their relationship (Peng, 2003). This analysis is critical for IJVs (Meyer et al., 2009). When the transaction environment is uncertain, more contingency planning is necessary (Williamson, 1985). One stream of research suggests that under high uncertainty firms may have intricate contracts with more contingency planning (Williamson, 1985). Another stream posits the opposite: under high environmental dynamism (Dess & Beard, 1984), firms may be more likely to forego the elaborate planning for long and complex contracts and engage in short-term, underdeveloped contracts as they proceed with some degree of flexibility (Li et al., 2008;Mayer & Argyres, 2004;Zhou et al., 2008). Thus there may be an inherent tension between enforcement uncertainty and the default choice of domestic courts versus BICA.

Enforcement Costs
When negotiating the IJV contract, the foreign MNE and the host firm engage in a forward-looking cost-benefit analysis to help determine the contingencies and specifications (Poppo & Zenger, 2002;Williamson, 1985). One area is potentially high enforcement costs of the IJV contract should a dispute arise. Generally, when IJV partners can significantly reduce the uncertainty surrounding enforcement costs, they are likely to choose the enforcement mode that provides for the most predictability. Choosing BICA over domestic courts may create additional enforcement benefits, such as (1) expertise, (2) time (urgency), (3) confidentiality, and (4) potentially most important, predictability of outcomes (Williamson, 1985: 70-71).
First, the costs of expertise (calling experts and expert testimony) are likely to be high in conflicts between IJV partners. The need to for expert witnesses and expert consulting is a fundamental shortcoming of domestic courts (Williamson, 1985: 70-71). BICA may overcome this shortcoming, because firms may select arbitrators who already possess the necessary expertise (Kapeliuk, 2012). This will save the IJV partners on the typically high expert costs (in addition to search costs associated with locating experts) (Heuman, 2003;Redfern et al., 2009;Williamson, 1985: 70-71). Therefore this is an advantage of BICA (Heuman, 2003;Redfern et al., 2009).
In domestic courts, the foreign MNE partner must try to find local experts or import foreign experts, which may generate travel and coordination expenses. Moreover, trade standards and other areas of expertise vary tremendously among countries, and foreign experts may introduce uncertainty surrounding expert testimony in domestic courts (Berkowitz, Moenius, & Pistor, 2004). Legal theory suggests that experts are less necessary when the underlying product is simple or standardized (Berkowitz et al., 2004;Drahozal, 2009). However, when experts are necessary, the foreign MNE may bear the majority of the enforcement costs in terms of coordination and travel.
BICA allows the IJV partners to choose arbitrators with specific expertise. This may provide more control over the expert costs. For instance, the foreign MNE may mitigate costs by appointing an arbitrator with sufficient expert knowledge (thus combining arbitrator fees with expert fees). While arbitrators are not always industry experts, they tend to be more industry savvy than domestic judges (Benson, 1999;Drahozal, 2009).
Second, IJV partners often seek timely resolution of their conflict. While there is a wide variation in domestic courts for contract enforcement timelines, in 2014 the worldwide average time to a decision was 1 year and 9 months (Ease of Doing Business, 2015). In contrast, the average arbitration institute takes about 8 months through normal channels. IJV partners may also expedite the BICA process. A special category of BICA includes expedited proceedings, which have specific guidelines and a set time table designed for time-sensitive disputes (Heuman, 2003;Redfern et al., 2009). An expedited BICA enjoys all the protections and benefits of BICA, but attaches a time requirement for resolution. The timelines for expedited proceedings can be as short as 3 days (!), as long as 3 months, but on average 45 days. BICA emphasizes speed and accuracy, an emphasis absent in domestic court systems where the expedited process is unavailable (Heuman, 2003;Redfern et al., 2009). For IJV partners where a decision is urgent, the costs associated with delays in the typically slow domestic court systems may be exceedingly high. BICA can alleviate these enforcement costs, and is even more attractive when decisions are urgent (Table 1).
Third, confidentiality is an important issue for conflicts over intellectual property (IP). One of the advantages of IJVs is that firms can access confidential information and knowledge in partner firms (Beamish & Lupton, 2016). However, these arrangements have shortcomings when the IJV partners disagree about how to treat information and knowledge. For instance, the rights protecting trade secrets may be extinguished when confidentiality is broken (Williamson, 1985) and are not recognized in many domestic courts around the world. For this reason, ensuring confidentiality is extremely important to firms due to the high cost of lost IP protection. In addition, confidentiality may carry implications for the liability of foreignness (Zaheer, 1995). Because the losing firm may be publicly shamed, and the public nature of domestic court proceedings may make domestic courts reluctant to rule against host firms, especially when such firms are national icons (Li, 2009).
Fourth, the bulk of enforcement costs lies in the predictability of outcomes. There are at least two dimensions: (1) the amount of the outcome (specified in dollars or performance) and (2) the costs of the resolution process (court, administrative, and lawyer costs) (Stephenson, 2009). Setting the costs for BICA will be one of the first scheduled decisions by the arbitrators or may be assessed by a standing body prior to the formation of the arbitral tribunal (Heuman, 2003;Redfern et al., 2009). The benefits of predictable arbitrator fees and administrative costs greatly reduce uncertainty for the partners.
Since BICA allows the IJV partners to use their existing lawyers (not local lawyers often mandated by the domestic courts), lawyer fees are likely to be less when compared with domestic courts (Heye, 2004;Laprès, 2009;Li, 2009). Overall, BICA may thus be attractive in locations suffering from high enforcement costs. Table 2 presents a direct comparison between the costs (as a percent of the claim) associated with domestic courts and BICA by region. In every region, BICA represents the lower cost option. It is not surprising that IJV partners often endeavor to include BICA clauses in IJV contracts to address the ex ante concerns with enforcement costs. When enforcement costs are high, the choice of BICA may become much more attractive in the IJV. In summary: Proposition 1: As enforcement costs increase, firms forming an IJV are more likely to engage in institutional borrowing by choosing binding international commercial arbitration as opposed to domestic courts in the host country of the MNE.
Going above and beyond Proposition 1, next we develop a contingency model (see Figure 1) to highlight the factors that intensify the relationship discussed in Proposition 1. The positive relationship between enforcement costs and arbitration grounds our model given our emphasis on BICA as an enforcement mechanism that substitutes for domestic courts (Proposition 1). Notably, this is a competing model, and when IJV partners choose BICA they cannot choose domestic courts. We expect that institutional voids in the host market will increase the positive relationship between enforcement costs and the choice of BICA (Proposition 2). In addition, high search and negotiation costs will further increase the positive relationship between enforcement costs and the choice of BICA (Proposition 3). Finally, the development of our institutional borrowing perspective draws on an IJV context. IJVs vary widely, and we have tailored IJV characteristics that may increase the positive relationship between enforcement costs and the choice of BICA over domestic courts. 4

INSTITUTIONAL VOIDS
Institutional voids represent a major concern for IJV partners in host markets with weak institutions (Khanna & Palepu, 1997). Enforcement of the IJV contract will be at the forefront of their concerns in ex ante discussions. Even in strong institutional environments, discussion of the adoption of BICA in IJVs tends to be a priority (Reuer & Ariño, 2007). When enforcement institutions are weak in the host market, the IJV partners are likely to look for adjudication mechanisms elsewhere-in other words, institutional borrowing.
A common institutional void relating to enforcement is the formalization of urgency. While the IJV partners may wish a quick resolution to move forward with their relationship or to exit the IJV, the domestic courts are often mired in expansive court cases and long-time horizon decision processes. Reputation and other non-market costs may be tied-up in these domestic disputes. Legal theory suggests that the length of time to resolve a dispute directly relates to the uncertainty surrounding the outcome. Moreover, host markets that suffer from long timelines tend to have the compounding problem of more (rather than less) filings in the court. For example, in 2002, India amended its civil code to improve the speed for decisions in the high courts. Locations that adopted the procedural rules saw fewer (rather than more) contractual breach claims and increased investment (Chemin, 2012). However, the effects were not uniform in India, emphasizing how formal institutional changes may fail to effectively overcome voids in a host market. Thus where institutional voids around enforcement urgency are present, enforcement costs for the IJV partners may increase, making BICA more attractive.
For example, some IJVs may be formed on access to trade secrets and the advantage in the marketfacing activities in the host market that the trade secrets provide. However, trade secrets are not universally recognized, manifesting in a salient institutional void. When submitting to a domestic court that does not recognize trade secrets, the IJV partners risk extinguishing the trade secret in other locations (not just the host market) should the court choose to reveal the secret. Other issues with confidentiality may relate to non-financial costs such as loss of reputation or even legitimacy. For instance, foreign MNEs tend to be in a weak position to manage and handle media attention in host markets relative to the host firm (Stevens, Xie, & Peng, 2016). The lack of confidentiality may provide host market media with an opportunity to bring the dispute to the public, potentially undermining the reputation of foreign MNEs. However, IJV partners may avoid these costs of enforcement if they engage in a confidential enforcement process such as BICA.
The predictability of costs associated with enforcement of decisions in locations with institutional voids tends to be low. In general, the number of disputes in BICA that arise from locations with weak institutions provides a conservative proxy for the choice of BICA when IJV partners confront rising enforcement costs associated with institutional voids. This makes it difficult to ascertain the extent to which these firms choose BICA in general. However, we can count the outcomes of the choice of BICA related to the location (institutionally strong or weak). Consider Table 3, which reports aggregate statistics about the breakdown of BICA in the major arbitration institutes. Except at Stockholm Chamber of Commerce (SCC) in Sweden, the majority of BICA cases adjudicated in the other locations involve parties from weak institutional environments. Overall, BICA tends to facilitate a large number of dispute resolutions with host firms from weak institutional environments. In summary: Proposition 2: Institutional voids of the host market intensifies institutional borrowing for firms forming an IJV to choose binding international commercial arbitration as opposed to domestic courts in the host country of the MNE. Following Williamson (1985), we recognize crossborder transaction costs may manifest in two categories in addition to the enforcement costs discussed above: (1) search costs and (2) negotiation costs. As a caveat, prior work investigating cross-border transaction costs along these categories emphasized the general costs (but not the legal costs) associated with the IJV relationship (Peng & Ilinitch, 1998). Building on Hennart (1989) and Williamson (1985), we focus on search and negotiation costs in terms of cross-border legal transaction costs. 5

Search Costs
Due diligence may be heightened in cross-border exchanges where information asymmetries are high-this is particularly true in developing economies with weak formal institutions (Peng, 2003). Yet in IJVs one of the advantages is the ability for foreign MNEs to delay final decisions about acquisition while gathering information (Chen & Hennart, 2004). Search costs will arise when (1) choosing legal representation and (2) choosing the place of enforcement because these are forward-looking concerns that partners are likely to consider ex ante. In BICA, unlike domestic courts, firms are free to choose legal representation (Heuman, 2003;Redfern et al., 2009). For instance, in China, foreign lawyers are not allowed to represent the foreign MNE in court proceedings (Laprès, 2009). For the foreign MNE, this increases search costs because it requires the selection and retention of a local (Chinese) lawyer. Ironically, retention agreements for a local (Chinese) lawyer are additional formal contracts-in this case between the foreign MNE and the Chinese lawyer. Search costs may increase because the foreign MNE is unable to safeguard against potential opportunism associated with the retention agreement with the Chinese lawyer, and adding another formal contract will be time consuming and costly (Argyres & Mayer, 2007;Schepker et al., 2014). In contrast, the China International Economic and Trade Arbitration Commission (CIETAC) allows a foreign lawyer to represent the foreign MNE in BICA proceedings in China (Laprès, 2009). As a result, despite China's well-documented weak rule of law, BICA can lower the search costs for partnering with a host firm in China. This is critical because the bulk of FDI into China has taken the form of IJVs, thus underscoring the important but previously under-appreciated role played by BICA (Heye, 2004). In addition, BICA provides multiple forums for enforcement (Heuman, 2003;Redfern et al., 2009). If IJV partner firms are uncertain about BICA enforcement in one country, they can ''shop'' for other enforcement locations. Unlike court decisions that are enforceable only in the jurisdiction of one court system (e.g., Chinese court decisions are not enforceable in France), BICA awards are enforceable in the 152 countries that signed the New York Convention (1958) (see Appendix). For instance, in the BICA concerning the IJV disputes between Danone (France) and Wahaha (China), the winning firm could therefore go to France, China, or any of the other 150 countries to enforce the BICA award. In this way, IJV partners may engage in enforcement forum shopping to reduce enforcement uncertainty and thus overcome a major limitation of IJV contract under the assumption of one court for enforcement.

Negotiation Costs
IJVs also tend to face significant negotiation costs, which BICA may reduce. Here, we will focus on (1) the location of the resolution process, (2) the language of the process, and (3) the concept of liabilities of foreignness as it relates to domestic courts and BICA.
Location-specific issues may arise during a conflict ranging from travel to the potential disadvantages of being in different time zones. These disadvantages stem from distance created by foreignness (Zaheer, 1995). Travel expenses should account for both the mode of travel (such as air and ground) and accommodations necessary while negotiators are away from home. The hassles associated with travel often prove to be a determinant of location selection (Schotter & Beamish, 2013).
Time zones and geographic distance tend to be another major concern (Zaheer, 1995). When extensive travel is necessary, negotiators who have been traveling may require time to recuperate. In addition, negotiation may increase coordination costs within the firm, where the distant representatives may need information, instruction, and

Overcoming institutional voids via arbitration
Brian C Pinkham and Mike W Peng authorization from the home country office that is in a different time zone. With this in mind, the dispute resolution coordination costs may be jointly reduced if both partners a priori choose a dispute resolution location that is geographically midway. This compromise may allow both firms to avoid home court advantages relating to proximity to headquarters (Davidson, 2010;Park, 1998).
Countries and cities may also create supportive BICA policies to attract firms seeking a neutral ground (Drahozal, 2004). Countries such as Sweden are very successful in attracting BICA business. In addition, such competition is filtering down to the city level as the major global cities compete for BICA-related business. For example, in 2013, New York City hosted 1,165 BICA cases, Paris 767, and Miami 128 (Olson, 2014). These major BICA hubs also had high numbers in 2014: London (301 BICA cases), Singapore (259), and Stockholm (203) (Croft, 2014). In other words, certain countries and cities around the world market their institutions to be borrowed, thus helping firms reduce costs.
One advantage of a middle ground is the expressed willingness for both sides to share the burden of legal intervention in the cross-border transaction relationship (Aghion & Holden, 2011;Spier, 1992), which may be a critical element of building and maintaining trust in the relationship (Carson, Madhok, & Wu, 2006;Gulati, 1995). Legal theory suggests that the process of BICA relative to court systems may range from adversarial to constructive. Relative to court proceedings that tend to be adversarial, BICA may be more constructive and more likely to maintain the IJV relationship (Lumineau & Oxley, 2012). Legal intervention is a necessary part of the contracting process that need not be adversarial (Spier, 1992). Hence if the goal of BICA is to clarify contractual points or resolve emerging problems through a neutral third party to continue the IJV relationship, sharing particular costs reinforces the relationship and thus reduces negotiation costs (Drahozal, 2009). Thus a middle ground may help maintain the relationship, which suggests that the anticipated shadow of the future may drive some IJV partners to choose BICA over domestic courts (Eisenberg & Miller, 2007).
The choice of language governing the contract may also increase negotiation costs. Domestic courts will use the official languages of the countries. For the foreign MNE, choosing domestic courts will bear this added cost of language. BICA may provide some relief for the foreign MNE through the choice of language of the proceedings and arbitrator. When IJV partners communicate in different languages, the choice of multilingual arbitrators may provide an advantage unavailable in domestic courts. In addition, arbitrators may ask for clarifications in other languages (Heuman, 2003). The proposed advantages are not costless, however. For the foreign MNE, the costs of language in the domestic court are likely to be high, because language may be a barrier of communication in court, with local lawyers, with communications outside the court during the exchange of documents, and so on. The foreign MNE must also pay for a translator during BICA or find an arbitrator who is multilingual if the parties choose a language other than the foreign MNE's native tongue. Thus due to language concerns, the foreign MNE may prefer BICA.
A middle ground may also have implications for liabilities of foreignness that the foreign MNE experiences in domestic courts in terms of national sentiment. Assuming that domestic courts are completely neutral and a conflict between the foreign MNE and the host firm can be objectively decided, the domestic populace fanned by the domestic media may still be biased against the foreign MNE (Stevens et al., 2016;Zaheer, 1995). Therefore foreign MNEs may prefer BICA because it may mitigate liabilities of foreignness that stem from national sentiment. Overall: Proposition 3: High search and negotiation costs intensify institutional borrowing for firms forming an IJV to choose binding international commercial arbitration as opposed to domestic courts in the host country of the MNE.

THE ROLE OF IJV PARTNERS
The choice of BICA versus domestic courts fundamentally lies within the powers of the IJV partners. We argue that the following partner-related attributes sit at the intersection of transaction costs and the choice of including BICA in a contract: (1) IJV partner size, (2) IJV size, (3) IJV duration, and (4) IJV partner profile.

IJV Partner Size
There are two points related to the size of the partners that are relevant when considering legal resources and enforcement costs. Firm size is a proxy for legal resources, such that larger firms have developed legal capabilities such as in-house counsel and large legal departments (Bagley, 2008;Brenner, 2011). Under the assumption that firms can learn to contract and better leverage their legal resources, larger IJV partners may have a stronger preference for BICA (Argyres & Mayer, 2007;Bagley, 2008;Mayer & Argyres, 2004). Most large firms (foreign and domestic) report a preference for BICA in cross-border deals (Mistelis, 2013). In turn, this suggests that smaller firms may not have the same opportunities as their larger counterparts, making BICA an expensive, unattractive option. Thus small and large firms' preference for BICA or domestic courts may differ.

IJV Size
The relative size and importance of the IJV may also drive the attention partners give to these transaction costs. The magnitude of the transaction costs is likely to increase as the IJV size (value) increases. As the IJV size increases, sunk costs (credible commitment to the IJV) increase, creating alignment incentives internally, but also highlighting the mutual hostage nuance of the IJV partners (Oxley, 1997). As the IJV size increases, both parties will seek strong enforcement mechanisms in anticipation of potential breaches.
This also has implications for IJV ownership, which is likely to influence the costs borne by IJV partners (Chen & Hennart, 2004). The need for protecting ownership position and facilitating changes in ownership as the IJV evolves makes strong enforcement mechanisms necessary (Beamish & Lupton, 2016;Zhou & Poppo, 2010). Generally, as ownership increases, the need for predictable enforcement may increase. Additionally, IJVs with higher potential for change in ownership and exit options may also drive the IJV partners to contemplate BICA to overcome increasing enforcement costs (Beamish & Lupton, 2016;Dhanaraj & Beamish, 2009).

IJV Duration
The duration of IJVs may also influence enforcement costs. On the one hand, when IJVs have a defined end date, enforcement costs tend to be higher (Lumineau & Malhotra, 2011). On the other hand, as IJV duration increases, conflict likelihood decreases and IJV partners are more likely to resolve issues without a neutral third party (Lumineau & Oxley, 2012). In part, IJV partners learn to cooperate (Mayer & Argyres, 2004;Lumineau & Oxley, 2012). Also, as ties between two partners intensify, the need to rely on external mechanisms may decrease (Gulati, 1995). Thus, as IJVs' duration increases IJV partners may not have the same priorities. Critically, this suggests that when the IJV duration is shorter and defined, IJV partners may be more likely to select BICA over domestic courts (Lumineau & Malhotra, 2011).

IJV Partner Profile
High profile host firms are attractive to foreign partners (Shi, Sun, Pinkham, & Peng, 2014). They tend to be in the best position to possess local resources. However, in the event that disputes arise in IJVs, high profile host firms may have tremendous advantages in domestic courts. On the one hand, domestic courts are likely to side with host firms (especially national icons) that have strong domestic reputations. On the other hand, domestic courts may render a decision in favor of foreign MNEs, but then refuse to enforce the decision against host firm assets. In addition, Siegel (2009) documented an interesting effect of asset movement by host firms out of developing economies. In other words, host firms (anticipating conflicts) may move the relevant assets to offshore locations where domestic courts do not have jurisdiction (Siegel, 2009). This means that the domestic courts may have jurisdiction to hear the grievance, but may not be able to enforce the decision because few assets are present to seize under their jurisdiction. This complicates collection on a domestic judgment even if foreign MNEs win. We recognize that this may be more apparent when host firms have ties to the government or have government endorsement, a common mechanism for host governments to ''support'' market transactions in the host market (Khanna & Palepu, 2010). This may pressure IJV partners to commit to the domestic court system. However, open government support for one IJV partner will escalate the need for a neutral enforcement mechanism for the foreign MNE. Locations where the growth of the economy has a strong dependence on government ties and FDI such as China also tend to promote BICA as a matter of policy (Drahozal, 2004;Li, 2009). In this sense, government ties will tend to drive foreign MNEs to push for the selection of BICA in ex ante negotiations. Thus: Proposition 4: Certain IJV attributes-such as large size of the partner firms, large size of the IJV, defined duration, and high partner profile-intensify institutional borrowing for firms forming an IJV to choose binding international commercial arbitration as opposed to domestic courts in the host country of the MNE.

DISCUSSION
Overall, we present a theory of how IJV partner firms may overcome institutional voids by using BICA as opposed to domestic courts. By integrating the legal literature on BICA in IJVs with the institution-based view, we bring a new perspective to the mechanisms that may help firms bridge institutional voids. Given that we know firms have the ability to borrow institutions, we must ask: under what conditions will the firm borrow institutions to overcome institutional voids? To begin to disentangle this complex issue, we identify BICA as a mechanism that helps reduce some cross-border transaction costs. We argue that some foreign MNEs may have a vested interest in avoiding host country courts, and that BICA allows foreign and domestic firms engaged in forming IJV relationships to create a neutral ground should conflicts arise.

Contributions
In our view, at least three contributions to the institutional voids literature emerge. First, we have identified and relaxed the assumption of ''one law, one court in one country'' that permeates the contracts literature underlying IJV research (Argyres & Mayer, 2007;Poppo & Zenger, 2002;Williamson, 1985). By bringing in the legal literature based on the choice of contract law, we enrich our understanding of how firms may use law and enforcement mechanisms beyond the default of only using local institutions. Second, we contribute to the discussion on institutional borrowing by outlining how BICA may help IJV partners bring strong legal institutions into weak legal institutional environments. In turn, this may reduce transaction costs for partner firms and address issues of enforcement when faced with institutional voids in domestic court systems. The adjudication mechanism underlying BICA allows the firms to borrow institutions to overcome weak enforcement institutions in the host market. Third, with a focus on IJVs attributes, we begin to explore the boundaries of our model by developing a more nuanced approach to our theory of institutional borrowing and the ability of IJV partners to overcome institutional voids. We believe this brings us closer to understanding when IJV partners may borrow institutions. By leveraging the specific case of how a specific institution-arbitration-matters, we have extended and enriched the institutionbased view with a focus on BICA (Blevins, Moschieri, Pinkham, & Ragozzino, 2016;Meyer & Peng, 2016;Peng et al., 2008Peng et al., , 2009. BICA has broad implications for IB and management research. Bagley's (2008) argument that experience with legal resources may result in firmspecific capabilities can be extended to expertise in BICA. The uniformity under the New York Convention and predictability under the auspices of institutional bodies of experts suggests that BICA offers an area of concentration for resources (especially legal expertise) that may otherwise be dispersed over many jurisdictions. This may help firms create more sophisticated and complex contracts (Reuer & Ariño, 2007), or alter the path of learning about contracting where we may see firms developing capabilities to use and manage BICA (Argyres & Mayer, 2007;Mayer & Argyres, 2004). In addition, given variations among property rights, the stability and predictability offered in BICA potentially impact licensing and technology contracts (Anand & Khanna, 2000). By offering a substitute for domestic courts, BICA provides a potential mechanism for institutional borrowing, where firms may avoid local instability and market failure.

Limitations and Future Research Directions
Because BICA is a private dispute resolution system, it is difficult to empirically observe the disputes and outcomes. We introduce tightly coupled relationships from the legal literature but there is still a need for more empirical testing to develop causal relationships. We encourage future BICA research to draw on information from established institutions such as the SCC, CIETAC, and ICC. Alternatively, researchers may contact large accounting and law firms with a BICA-focused practice. A qualitative approach (with access to these firms) would greatly enrich our understanding of the borrowing mechanism related to BICA. Specifically, our inclusion of aggregate statistics (Tables 2 and 3) provides a baseline for the use of arbitration, but elaboration by country, region, and industry would provide an enriched account of BICA with further insights.
There is more work to be done surrounding the IJV. One of the challenges in our article is to bring in sufficient, generalizable perspectives. We framed these around the foreign MNE, and included information on the relevance of BICA for MNEs from weak institutional environments (Table 3). Yet the legal literature suggests that IJV partners may have different priorities in terms of (1) enforceability (finality of the decision), (2) choice of law, and (3) location of the arbitration (IFC, 2006;Mistelis, 2013). Thus, a focus on host firms in IJVs may provide further insights on host firms' unique interests.
In addition, the foreign MNE and the host firm are rational partners but only boundedly so. One such boundary on the rationality in the choice of BICA will be awareness. The awareness of BICA is not universal, and many host firms in markets with weak enforcement institutions may have an interest in learning more about BICA. For instance, a survey in Ukraine indicates that many small and medium-sized (domestic) firms were not aware of BICA (IFC, 2006). For this reason, awareness may play out differently depending on whether both (or one of the) firms in the IJV are aware of BICA. It would be natural in this setting, however, to see that host firms (rather than foreign MNEs) have a deeper commitment to domestic courts and therefore need more incentives to agree to BICA (Puffer et al., 2010). One concern may be that the foreign MNE bringing in a clause like BICA that is not well understood by the host firm may undermine the relationship (Beamish & Lupton, 2016). Overall, the foreign MNE and the local firm essentially are engaged in a game of negotiations to construct an IJV, and how this game-in our context with a focus on whether to choose BICA or domestic courts-plays out can be fascinating new ground in future research drawing on game theory.
In addition, a finer-grained approach to partner selection would be interesting and warranted. We assumed that partners engage in a discussion around the inclusion of BICA. There is more work to be done on the longevity of the IJV, as we only address the duration of the IJV above. There are interesting questions that arise when considering the ex ante discussions of open-ended IJV duration and the adoption of BICA. While IJV partners may be together for a long time, they cannot predict with certainty how long the IJV will endure when engaged in ex ante discussions (Lumineau & Malhotra, 2011;Lumineau & Oxley, 2012). This begs the question: how may a cost-based approach change when looking more deeply into the longevity and evolution of the IJV?
In addition, foreign MNEs that require BICA as part of IJV contracts would not fit in the arguments here, since we assume that using BICA is a choice.
Beyond the IJV, we may still consider that different types of governance may have different needs for BICA. For high-risk cross-border acquisitions, for instance, the legal literature anecdotally notes that firms seem to prefer domestic courts (Mistelis, 2013). Yet for other high-risk exchanges such as technology transfer, BICA is often preferred (Drahozal, 2009).
BICA may also provide an interesting setting for trust-based studies of contracts (Malhotra & Murnighan, 2002). Recent research on the role of contract and resolution clauses (such as domestic arbitration clauses) suggests that private forms of dispute resolution have a different effect on the interfirm relationship relative to court proceedings (Harmon, Kim, & Mayer, 2015;Lumineau & Malhotra, 2011). To this end, the trust-based contracts literature has focused on the general, private forum of dispute resolution, an aggregate measure that may include dispute resolution forms such as mediation, arbitration, and conciliation. Given that many firms openly prefer arbitration in cross-border disputes (Drahozal, 2009;Mistelis, 2013), future research can more closely compare court and BICA effects on trust in the interfirm relationship. In addition, foreign MNEs and host partners are likely to vary along contextual (and cultural) competencies that may relate to the ability to identify and manage institutional voids (Khanna & Palepu, 2010;Peng, 2003).
We have engaged in the discussion of a specific (and critical) subset of IJV partnerships between one foreign MNE and one host partner firm. Other notable extensions may include multiple IJV partners and IJV partners that are both foreign to the host market. In these instances, multinationality and coordination costs may demand the use of BICA rather than domestic courts. But, perhaps more salient in these instances would be the dilution of local ties to the host market. The question then becomes whether the choice of BICA in these instances is driven by institutional voids or by the fundamental lack of local embeddedness.
Beyond the IJV context, BICA studies can extend to other conflict settings. First, while we focus on firm-firm IJVs, a comparison of arbitration of firmfirm conflicts within national borders (domestic) and across national borders (international) may be interesting. Second, we do not address firm-state conflicts that may be particularly relevant for firms interested in protecting their investment from nationalization without compensation (Stevens et al., 2016). This is a common occurrence as seen in recent conflicts between multinationals in the oil and gas industry and the governments of Argentina and Venezuela (Pinkham, 2010). Finally, we have not addressed firm-consumer conflicts. Forcing consumers to accept binding private arbitration and to forego their rights for court proceedings may save the firm some transaction costs, but may have grave ethical (and potentially legal) ramifications. How a foreign MNE uses arbitration to deal with host country consumers remains a fascinating area for future research. Overall, while we have concentrated on how BICA can facilitate conflict resolution in IJVs-a relatively focused domain-in a broad sense whether arbitration undermines domestic law by substituting law with secret arbitration remains a worthwhile endeavor for future research.

CONCLUSION
Leveraging and extending the institution-based view, our central argument is that IJV partners may overcome institutional voids by borrowing enforcement mechanism via BICA that offers an alternative to domestic courts. Going above and beyond the typical setting of ''one law, one court in one country'' that characterizes much of the contracts literature, BICA can be used as a fundamental building block for facilitating market exchanges in the face of institutional voids. In conclusion, using BICA reduces cross-border transaction costs because it enables the borrowing of enforcement institutions as an alternative to domestic court systems.

ACKNOWLEDGMENTS
Earlier versions of this article were presented at three Academy of Management Annual Meetings (Chicago, 2009;Montreal, 2010;and San Antonio, 2011). We thank the four reviewers for their comments and direction. We appreciate the hard work and input of the Special Issue editors, and in particular Suzana Rodrigues' editorial guidance and support. We thank Paul Beamish, Hao Chen, Gregory Dess, Seung-Hyun Lee, and Roberto Ragozzino for helpful comments, and Tarun Khanna and Sheila Puffer for detailed discussion of earlier drafts at the JIBS paper development workshop at Harvard Business School. This research was in part supported by the UTD Graduate Fellowship and the Jindal Chair. NOTES 1 Arbitration is ''a non-judicial proceeding in which disputing parties submit their conflict to an impartial person or group of persons for a final and binding resolution instead of to a judicial tribunal and must be invoked by voluntary agreement of the parties … arbitrators may consider rules of contract law, practice, custom, and general principles of equity, as well as personal concepts of justice, public policy, logic and ethics'' (Galanter, 2001: 586). International commercial arbitration provides a neutral arbitrator (Berger, 1993), avoids ''hometown justice'' (Drahozal, 2009;Park, 1998), and results in an award (a final, binding decision) that is enforceable across borders where a domestic court judgment is not (Bühring-Uhle, 1996). 2 We recognize that enforcement is a multifaceted concept, and we use the term ''enforcement'' in three ways. First, we draw on the law literature to narrowly define enforcement as the enforcement of the IJV contract-if the partners use BICA, the underlying assumption in the enforcement of the BICA award is also relevant (Mistelis, 2013). Second, we adopt the term enforcement mechanisms from the transaction costs and institutional voids literatures to describe BICA and domestic court systems (Khanna & Palepu, 2010;Williamson, 1985). Third, we adopt Williamson's definition of enforcement costs to develop our theoretical model and describe the differences in the mechanisms (Williamson, 1975(Williamson, , 1985. 3 This is true of both developed and developing economies. The institutions affecting market exchange in the EU, for instance, are quite different from those of the US (La Porta et al., 1997). Similarly, differences exist between developed and developing economies and among developing economies (Meyer et al., 2009;Peng et al., 2009). 4 We thank the editor and the reviewers for their guidance in developing the conceptual model. 5 Hennart (1989) divided cross-border transaction costs into market and control costs. Market costs are associated with exchange, and control costs are associated with integration of the target exchange partner across borders. When control costs are less than market costs, firms are likely to integrate to draw on internal exchange advantages (Brouthers & Hennart, 2007;Buckley & Casson, 1976). We focus on market costs, but not control costs.