Entrepreneurs and Social Capital in China

INTRODUCTION

Does social capital operate differently in China? A long and vibrant literature on the concept of guanxi suggests not only that social capital might have a different character in China but also that it might prove more valuable there to employees and entrepreneurs alike. Drawing on unusually high quality data on Chinese executives, Burt and Burzynska (2017) explore the question of whether the same structural configurations of relationships appear associated with success in China as have been found in the West. Their short answer is yes. As in the West, the founders and managers of larger companies have relationships characterized by a larger proportion of ‘structural holes’ – where the focal individual represents the only connection between his or her contacts (in other words, a brokering relationship). Closure – having multiple ties in common with another individual – meanwhile appears to foster trust, just as it does elsewhere. One could therefore view the results as strong evidence for the universality of a structural approach to social capital, one based on the patterns of relationships between individuals (e.g., Burt, 1992; Coleman, 1988; Podolny, 1993).

But one could also see the glass of their results as half empty – that China really does operate differently. Consider, for example, Figure 3 in Burt and Burzynska (2017). Although the relationships between success and structural holes appear similar, note that the two panels differ in the ranges of their horizontal axes. Managers in China appear far more constrained on average than their Western counterparts (i.e. they have fewer brokering relationships). Within the overlapping range of the data, moreover, the relationship between success and structural holes appears much steeper in China. Social capital, in the form of brokerage, therefore, appears both less common and far more valuable in the East.

Interestingly, almost all of the value of brokerage but none of the relationship between closure and trust comes from the older relationships, those named as important to earlier organizational events on the Chinese survey. One can see this fact for brokerage by comparing Burt and Burzynska's Table 1 to their Table 3 and for trust in the right panel of their Figure 4. For these Chinese managers, the ideal configuration of social relationships appears to come from establishing a diverse set of contacts early in their careers but then building almost clique-like connections later on. Perhaps these differences in the frequency and value of brokering relationships and in the shifting nature of social capital over time represent the signature fingerprint of guanxi in the Chinese data.

But then again one might also question whether one should observe guanxi in social structure. Perhaps guanxi pertains not to the structure of relationships but rather to the nature of exchange at the dyadic level. The discussion of rings of guanxi sounds more like sets of rules for interaction than an essay on social topology (Luo, 2011; Luo & Cheng, 2015; cf. Simmel, 1950).

Regardless of whether one sees the glass as half full or half empty, however, the results reported here raise a number of important issues for future research, some methodological, some conceptual.

EVENT CONTACTS

Research on social structure has advanced considerably on the basis of the kinds of surveys implemented here, using name generators to elicit information about relationships. The best of these surveys, moreover, have further solicited information about the relationships between the respondent's contacts, allowing the researcher to identify cases of brokerage and closure. The survey instrument used here offers an interesting technique for pushing this methodology even further by using significant events in the life of the company as anchors for additional name generators.

This event-based methodology has some attractive features. Whereas one would normally worry that respondents might not accurately recall when they first met someone or might forget those with whom they had lost contact, important events undoubtedly etch unusually deep impressions in the memory. People probably recall well who had been actively involved in them. These events also anchor the relationships in time, allowing for an appropriate sequencing of when they had been important.

This approach also helps to address the issue of causality. Anytime one examines the relationship between social structure and performance, concern exists that, rather than being causal, social relationships instead reflect the characteristics of the actors (Manski, 1993; Stuart & Sorenson, 2007). Central figures, for example, may have large numbers of contacts because they have magnetic personalities or resources useful to others. If these individuals perform unusually well, should one attribute that fact to their centrality, or to their personal qualities and resources? Using event-based name generators helps to alleviate this issue. Because these events occurred in the past, they allow the relationships observed to precede the outcomes, largely eliminating the possibility of reverse causality.

But the use of event contacts also raises questions. Who are these people? Were they active members of the respondent's network at the time of the event, a lost tie, or do they belong to a large pool of latent ties, activated in a time of need? In either case, both the relational characteristics and the structural properties of these contacts might require a somewhat different interpretation from that of contemporaneous contacts.

Consider first the relational characteristics. These individuals had been important at the time of the event for a reason, one that might have been related to the nature of the relationship between the respondent and the contact. For example, many of the formational events for startups represent mini-crises for the firm and its founders. During these periods of vulnerability, entrepreneurs probably turn to trusted parties. Because the name generator would then have effectively selected on trust, it should perhaps not surprise us that trust has almost no relationship to closure among these managers (see Figure 4B). One must therefore interpret these relational characteristics with care.

In terms of the structural properties, note that the name generator approach typically establishes the extent of brokerage and closure by asking the respondent whether his or her contacts have connections to each other. When using events as prompts, however, the names generated come from different points in time. The survey therefore effectively asks whether past contacts know current ones. But one would ideally want to know whether the event contact had been in a brokering relationship or in one characterized by closure with respect to the other relationships the respondent had at the time of the event.

The nature of the bias in measurement introduced by collapsing the time dimension, moreover, could go either way. If the respondents lost mutual connections, then these event contacts might appear less constrained than they really had been. But, to the extent that structural holes have a tendency to close over time (Ryall & Sorenson, 2007), these relationships may appear to exhibit too much closure when identified retrospectively.

One means of partially addressing these issues would involve using event-based name generators even more extensively, for example for social events as well as organizational ones. By finding important social events that occurred close in time to the organizational events and by generating the names of those connected to those events, the researcher could establish both some additional relationships that had been active at the same time and would select relationships on a different set of characteristics. Doing so, therefore, could allow both for calibration of the selection bias involved in using the events as name generators and for establishing the direction of any bias associated with pooling relationships over time.

ENTREPRENEURIAL SOCIAL CAPITAL

Although Burt and Burzynska (2017) present their results as analyzing the correlates of success and trust among entrepreneurs in China, the sampling design for the survey explicitly excluded small firms in their first few years of operation. The average firm in the sample had been operating for a dozen years. To the extent that brokerage has a positive relationship with firm size, that correlation therefore may stem from the value of these relationships not just in the early days of the business but also after it has become fairly well established.

The information reported in their articles does, however, provide some insight into the relationships that may have mattered most to Chinese entrepreneurs in the earliest days of their firms. Table 7, for example, reports the characteristics of those contacts named as important to founding versus those named in response to the other name generators. Relative to later relationships, founding contacts appeared more embedded, closer (in the sense that they involved more regular communication), more established, and more frequently family. If anything then, strong, embedded ties rather than brokering relationships appeared most important at the time of founding.

If true, moreover, that pattern would accord well with those observed in the West. Although having a diverse set of information sources may help entrepreneurs to spot opportunities (Burt, 2004; Renzulli et al., 2000), founding a firm also requires the mobilization of a variety of resources: raising capital, recruiting employees, and enlisting the support of suppliers and distribution channels. Because access to these scarce resources has strong opportunity costs – in the sense that funding or supporting one entrepreneur means that the resource holder cannot support another – closure often facilitates the trust needed for this resource exchange (Løvas & Sorenson, 2008).

Consistent with this expectation, studies of specific aspects of the process of founding a firm have repeatedly documented the importance of closure and strong ties. Would-be financial backers, for example, worry that the entrepreneur might not spend their money wisely. Trusted third parties, therefore, have been found to play a critical role in connecting entrepreneurs to the venture capitalists who finance them (Shane & Stuart, 2002; Wuebker et al., 2015), a result that has also been replicated in China (Batjargal & Liu, 2004; Batjargal, 2007).

An even more difficult problem faces entrepreneurs when trying to hire early employees. Whereas investors can diversify away some of the risk involved in any particular venture, employees generally must choose a single employer, putting all of their eggs in a single basket. They therefore must have a great deal of confidence and trust in the entrepreneur to leave a secure position at an established firm to join a startup. It therefore does not seem surprising that family and close friends frequently serve as the first employees hired by founders (Ruef, 2010).

Because of the importance of resource mobilization to the founding process, the most successful entrepreneurs in the West have been those with deep connections in a community. Dahl and Sorenson (2012), for example, found that the most embedded entrepreneurs in Denmark founded the firms that survived the longest and that proved most profitable. The importance of strong, embedded ties to entrepreneurs in China would appear completely consistent with these results from the West.

But this discussion also highlights the many different ways in which social capital may matter to entrepreneurs, from opportunity identification to firm formation to building a client base. Different patterns of relationships – different forms of social capital – may prove valuable at different stages and in different activities. Aven and Hillman (forthcoming) argue that the most effective entrepreneurial teams actually have members who specialize in particular structural roles, with some being boundary-spanners and others being deeply embedded in communities. Moving forward further on these issues will probably require both more information about the full set of individuals involved with the startup, what Ruef (2010) calls the entrepreneurial group, and a shift from the overall performance of the firm to evaluations of its effectiveness on the component behaviors that contribute to success.