What are the similarities between an ethnic group and a religious group?

Cultural similarity has been found to foster mutual trust, reduce transaction costs, and facilitate economic exchange (Guiso et al., 2009). Therefore, national cultural similarity promotes cross-border trade (Guiso et al., 2009) and cross-border mergers and acquisitions (M&As) (Ahern et al., 2015). In addition, investors react more positively to cross-border merger announcements by firms with similar national cultures than those by firms with different national cultures (Ahern et al., 2015). The importance of national cultural similarity has been documented not only in cross-border trade and acquisitions but also in cross-border strategic alliances (Nielsen, 2007, Shenkar, 2001, Tihanyi et al., 2005). Because national cultural similarity helps with trust building, reduces transaction costs, and expedites information flow and organizational learning between firms, partner firms from countries with similar national cultures are more likely to establish strategic alliances and the performance of such strategic alliances tends to be better (Nielsen, 2007, Tihanyi et al., 2005). Therefore, it is economically rational for firms to establish cross-border strategic alliances with other firms sharing similar national cultures. Investors should react positively to such alliance announcements.

Although strategic alliances between partner firms can be associated with lower transaction costs and economic efficiency, intergroup bias research in psychology suggests that cultural similarity can give rise to in-group favoritism that can bias firm alliance decision making. Intergroup bias regards the tendency of people to evaluate the in-group and its members more favorably than the out-group and its members (Hewstone et al., 2002, Mullen et al., 1992). The fundamental reason underlying intergroup bias is that the knowledge and emotional value of membership in a group are critical to one's sense of self (Tajfel and Turner, 1979) and in-group favoritism and out-group derogation can enhance individual self-esteem and sense of self (Hewstone et al., 2002). Cultural similarity is associated with shared social cues and provides an emotional bond for people sharing similar cultural backgrounds (Chen et al., 2009). Therefore, cultural similarity can lead to in-group favoritism — in-group members extend preference and favor to each other (Hewstone et al., 2002). Although in-group favoritism can foster in-group trust and facilitate communication between alliance partners, it can become a hurdle for firms to establish alliances with other firms that have dissimilar cultures but can potentially be better alliance partners, resulting in bias alliance formation decisions.

To explore the existence of in-group favoritism associated with cultural similarity, we examine the influence of cross-regional cultural similarity in the U.S. on strategic alliance formation and combined alliance announcements. We choose the U.S. to study the impact of within-country cultural heterogeneity on strategic alliance formation and announcement returns because the U.S. is one of the most culturally and ethnically diverse nations in the world. As a federal republic, each state has inherited unique cultural traditions and has formed its own political systems over time. From the Deep South and the Bible Belt, to the Rustbelt and the Stroke Belt, the U.S. has long been divided into distinct regions with different cultures (Rentfrow et al., 2013). Further, we examine the implications of cross-regional cultural similarity in the same country instead of cross-border cultural similarity because cross-border alliance formation decisions can be driven by both cultural factors and formal institutional factors. Thus, domestic strategic alliances provide a cleaner context to examine what role cultural similarity plays in economic exchange.

We use religious similarity and ethnic similarity to proxy cultural heterogeneity across U.S. states because of the crucial role of religion and ethnicity in defining individual and group identities in the U.S. (Higginbotham and Andersen, 2012, Putnam and Campbell, 2010). According to a 2009 Gallop survey, 56% of Americans report that religion is very important in their lives, 63% claim that they are a member of a church or a synagogue, and 55% say that they attend a church or a synagogue once a month. A large body of sociological literature documents the widespread effects of religion on agents' economic behaviors (Lehrer, 2004). Thus, religion has become an integral part of society and individual identity (Putnam and Campbell, 2010). As an immigrant country, the U.S. has also attested to the importance of ethnicity in influencing social, economic, and political lives (Higginbotham and Andersen, 2012). Ethnic background is an important determinant of cultural values because each ethnic group is defined by its unique heritage, norms, values, attitudes, and behaviors (Phinney, 1996). Therefore, ethnic identity has been shown to affect intergroup conflict and cooperation (Ting-Toomey et al., 2000).

Given that religion and ethnicity are two important identity cues that distinguish in-group members from out-group members, firms locating in regions that are similar in religious beliefs and ethnic backgrounds are more likely to establish alliances because these firms may extend in-group favoritism to each other. Such a prediction is similar to the economic rationale underlying most studies on the role of culture in cross-border strategic alliances. However, intergroup bias research proposes that in-group favoritism based on religion and ethnicity can bias organizational decision making and partner-selection processes (Hewstone et al., 2002, McPherson et al., 2001) and deter firms from establishing alliances with partners that generate the most economic benefits — representing a form of agency costs. Thus, investors will respond less favorably to alliance announcements between firms located in regions1 that have similar religions and ethnicities because of in-group favoritism.

To test our arguments, we apply a “gravity” model of international trade to interstate strategic alliance activities. Other researchers have applied gravity models to predict how geographic distance affects cross-border relation intensity (Frankel and Romer, 1999). We use gravity models to examine the influence of religious and ethnic similarities on the volume of interstate strategic alliances. Using domestic strategic alliance data from 1997 to 2013, we find strong evidence that religious and ethnic similarities increase the volume of interstate strategic alliances. The size of the effect is worth mentioning as well. With a change from the 25th percentile to the 75th percentile in religious similarity, the number of strategic alliances will increase by 4%. The same change in ethnic similarity will lead to a 19% increase in the number of strategic alliances.

More important, we find a negative relationship between religious and ethnic similarities and combined abnormal returns upon alliance announcements. This suggests that investors may perceive that in-group favoritism based on religious and ethnic similarities could have biased firm alliance formation decisions and thereby may respond negatively. In multivariate regressions accounting for sample selection bias, we find that a change from the 25th to the 75th percentile in religious similarity would lead to a 1.04% drop in the combined accumulative abnormal return (CAR). The same change in ethnic similarity will lead to a 0.46% drop in the combined CAR. These results indicate that religious similarity and ethnic similarity impose substantial costs for strategic alliance partners. However, the negative relationships between religious and ethnic similarities and combined CARs are weaker for research and development (R&D) alliances. Two reasons may explain the negative moderating effect of R&D alliances. First, R&D alliances are associated with high-level uncertainties and transaction hazards (Li et al., 2008) and cultural similarity can deter opportunism among alliance partners (Parkhe, 1993). Second, R&D alliances are concerned with high levels of tacit knowledge, and cultural similarity between partner firms may facilitate the transfer of tacit knowledge among partners. Therefore, investors are more likely to believe that economic rationale instead of in-group favoritism drives alliance formation decisions.

To identify the causal effect of interstate religious and ethnic similarities on alliance activities requires an exogenous shock to religious and ethnic similarities across states. Such an event is quite unlikely. Thus, to control for alternative explanations, we use instrumental variable regressions to address potential endogeneity and reverse causality. We instrument for religious and ethnic similarities using climate differences between two states. Specifically, we use dyadic state-level climate differences in terms of average monthly total temperature and average monthly total precipitation in 1931 to instrument for religious and ethnic similarities. The instrumental variables that we choose pass the test of relevance and exogeneity and our results are robust to instrumental variable regressions. In addition to using the volume of strategic alliances as a dependent variable, we also investigate the influence of religious and ethnic similarities on the volume of interstate joint ventures and M&As. We find that religious similarity and ethnic similarity both bear statistically significant and positive relationships with the number of interstate joint ventures as well as the number and value of interstate M&As, lending further support to our arguments.

The paper is related to several strands of the literature. First, we find that cross-regional religious similarity and ethnic similarity are negatively associated with combined abnormal returns upon alliance announcements, which is aligned with the prediction from intergroup bias research. Such a finding indicates that cultural similarity may give rise to suboptimal decision making. Although religious and ethnic similarities increase mutual trust and facilitate communication and coordination, they can give rise to in-group favoritism that can potentially be detrimental to effective alliance decision making. By documenting the negative influence of religious and ethnic similarities on combined alliance announcement returns, we illustrate the potential dark side of cultural similarity and provide a new insight into how investors evaluate economic exchange between partners sharing similar cultural backgrounds.

Second, this study shows that within-country cultural similarity also plays an important role in shaping economic exchange. The relationship between cultural values and economic exchange has attracted great scholarly attention (Guiso et al., 2006). To date, research has focused on the role of culture in cross-border economic exchanges and generally assumes that each country has homogenous cultural values. Yet, large variations in cultural values exist across different regions in the same country — each neighborhood, city, or state has its unique cultural values that play a critical role in shaping their residents' decisions and behaviors (Rentfrow, 2014). Whereas most of the literature explores the influence of national cultures on cross-border economic activities (Giannetti and Yafeh, 2012, Kogut and Singh, 1988), findings from this study suggest that cross-regional cultural similarity also exerts a significant influence on intra-country economic exchange, at least in the form of interstate strategic alliances and alliance announcement returns. In our supplementary analyses, we also find that religious similarity and ethnic similarity between two states affect interstate joint ventures and M&As in the U.S. These results collectively indicate that cultural similarity not only can bias cross-border economic exchange but also can influence intra-border economic activities.

Third, our work is related to the strategic alliance literature. A lot of scholarly attention has been paid to examining what drives strategic alliance formation (Park and Ungson, 1997, Pfeffer and Nowak, 1976, Robinson, 2008). Although culture has been identified as an important factor influencing cross-border strategic alliances (Kogut and Singh, 1988, Pothukuchi et al., 2002, Weber and Camerer, 2003), this study, to the best of our knowledge, provides the first large-scale evidence that intra-country cultural heterogeneity has a substantial influence on domestic strategic alliance and joint venture activities.