There is now growing recognition amongst scholars of the important role of social capital in improving economic outcomes. In influential work in economics and sociology it is argued that social interactions generate social capital, which in turn may facilitate contracting (Putnam (2000); Burt (2004)). This view is supported by several empirical studies (Knack and Keefer (1997); Guiso, Sapienza, and Zingales (2004); Karlan (2005); Karlan et al. (2009)). Additionally, social proximity can mitigate informational frictions, thereby enabling transactions to take place that are otherwise inhibited by adverse selection and moral hazard problems.

However, there is also a dark-side view of social proximity in distorting markets: in a seminal work, Olson (1982) identifies the emergence of self-serving interest groups that are created to further their own interests largely at the expense of broader economic prosperity. He argues that, after a period of stable growth, countries have a tendency to accumulate rent-extracting institutions that ultimately lead to the decline of nations. In a similar vein, Acemoglu and Robinson (2012) warn us about the role of colluding elites in establishing extractive institutions as a major impediment to economic prosperity.

In an empirical paper, we examine how social connections in elite networks affect the allocation of resources in an economy. Specifically, we investigate credit allocation decisions of banks to firms inside a social network. The network in question is a global service club organization that has its headquarters in the US with individual service club branches operating locally in several countries.

We find that after their owners/CEOs enter a club branch, firms experience an increase in lending from the in-group banks, that is banks whose local branch manager is member of the same club branch, that is 37.20 percentage points higher than from out-group banks. Additionally, the probability of establishing a new relationship with the in-group bank relative to an out-group bank increases by 14.95 percentage points after joining an existing club branch. Similarly, after an existing member of a club branch is elected mayor and becomes head of the local state bank, we observe a significantly higher increase in lending to in-group firms from the local state bank compared to other banks (48.60 percentage points), and a significantly higher probability of establishing a new relationship with the local state bank (16.62 percentage points).

To differentiate between a rent-seeking mechanism and positive effects of connections on credit allocation, we evaluate the relative profitability of connected lending by comparing the return on loans (ROL) that banks generate from in-group vis-a-vis out-group transactions. We find that a given bank generates a 4.37 percentage points lower ROL on in-group loans compared to loans extended to out-group firms in the same city. Comparing lending from in-group and out-group banks to the same firm, we observe a 2.73 percentage points lower ROL for in-group banks. Investigating the drivers of the difference in ROL, we observe that while interest rates and recovery rates are not significantly different for loans made by in-group and out-group banks, in-group banks continue to lend to ailing firms after out-group banks start to withdraw lending. It is this excess continuation of firms as a going concern that generates the lower ROL.

To sharpen our analysis on the underlying mechanism, we exploit cross-sectional differences in bankers’ incentives. An implication of Becker’s (1957) work on discrimination is that competition provides a mean to mitigate taste-based discrimination if such practices are costly to firms. In competitive markets, firms (or in our context banks) that engage in costly discrimination face the risk of being driven out of business. Consistent with this view, we observe that banks engage significantly less in in-group lending in areas where credit market competition is higher, and we find that the wedge between in-group and out-group ROL is smaller in areas with competitive banking markets. Furthermore, we examine how career concerns affect rent-seeking behavior of bankers. While younger bankers are subject to strong career concerns (bad performance may have a significant impact on future promotion and compensation decisions), older bankers, close to retirement, are less concerned about the impact of bad performance on future income. When we compare differences between in-group and out-group lending for young and older bankers in private banks, we find that young bankers engage relatively less in preferential in-group lending and the wedge between in-group and out-group loans is significantly smaller for young bankers. An alternative way to assess the effect of career concerns on rent-seeking behavior is to look at ownership. A significant fraction of banks in our sample are state-owned banks in which bankers’ compensation and promotion are less closely tied to their performance. We find that these banks engage more actively in in-group lending, and the wedge between in-group and out-group ROL is also significantly wider for these banks.

Finally, we examine how firms deploy the extra financing they receive through their membership to the network. This provides an additional opportunity to sharpen the evidence on the mechanisms at work. We observe that firms do not use the extra financing to make new investments, as one would expect if social proximity to the lender relaxed financing constraints. Instead, firms use these funds to increase payments to the shareholders, which in most cases means paying out to the CEO herself, as most of these are relatively small, family-owned firms.

Overall, our findings resonate with existing theories of elite networks as rent extractive coalitions that stifle economic prosperity.

David Schoenherr is Assistant Professor of Economics at Princeton University, Department of Economics, Rainer Haselmann is Professor of Finance, Accounting and Taxation at Goethe University Frankfurt and Vikrant Vig is Professor of Finance, London Business School.