The financial history of the United States offers an opportunity to investigate this issue. Economic historians have argued that the commercial banks of 19th‐century New England, which led American industrialization, functioned like investment clubs, lending to insiders and channeling capital to the corporations founded by their directors. The personal connections thus established between the banks and nonfinancial corporations of the region likely helped firms mitigate the frictions resulting from asymmetric information. In the context of a financial panic, the firms that were affiliated with commercial banks should have had better access to finance, so long as the panic did not originate among the region’s commercial banks.
My research analyzes the effects of the Panic of 1873 and the subsequent macroeconomic slump on the nonfinancial corporations of Massachusetts. Unique among all American states, Massachusetts imposed extremely strict disclosure requirements on its nonfinancial corporations in the late 19th century that make it possible to observe affiliations with commercial banks, as well as annual accounting information. Using this data, I document the director interlocks between the state’s corporations and commercial banks and investigate whether the firms that were affiliated with banks before the crisis fared better over subsequent years. The Panic of 1873 originated among railroad financiers in New York and was initially concentrated among private banks and brokerage houses. It did not directly impair New England’s commercial banks, which were in sound condition and were unconnected to the events that led to its outbreak. If bankfirm affiliations helped moderate the problems that resulted from asymmetric information, then firms with a commercial bank director on their board should have had better access to finance in the years following the panic.
The results of my empirical analysis indicate that corporations with bank affiliations did indeed suffer less in the wake of the crisis. Around 59 percent of all nonfinancial corporations in Massachusetts had a bank director on their board in 1872. Those firms survived the recession of the 1870s at higher rates and, among the surviving firms, those with bank affiliations grew at faster rates and saw their credit ratings deteriorate to a lesser extent. Consistent with having bankerdirectors help resolve problems related to asymmetric information, these effects were strongest among younger firms and those with lower shares of fixed assets on their balance sheets. The effects were also quantitatively significant: counterfactual estimates suggest that in the absence of bank affiliations, the total assets of all nonfinancial corporations in Massachusetts that existed in 1872 would have been 35 percent lower in 1881.
The presence of a bank director on a corporation’s board could have benefitted firms through several channels. Nineteenth‐century bank directors held discretion over the allocation of credit from their institution, so a directorship with a nonfinancial corporation should have facilitated greater access to credit by the corporation. But an affiliation with a bank may also have functioned as a signal of the quality of the corporation, and a banker‐director may also have contributed financial expertise to its management. In order to investigate the importance of credit relationships in the results, I study the effects of the presence of a bank cashier on firms’ boards. Nineteenth‐century cashiers were the public faces of their banks and oversaw their operations and maintained their accounts. But because bank cashiers normally could not influence the flow of credit the way bank directors could, their presence on a firm’s board likely added financial expertise and signaled quality without directly facilitating greater access to credit. Their presence on firms’ boards produced no discernable effect.
A source of concern regarding the empirical results is that they may reflect the selection of particular types of firms into affiliations with banks. The empirical analysis controls for time‐invariant unobserved characteristics such as the quality of a firm’s management, but if firms that were more resilient to a shock were more likely to be affiliated with banks, then this would imply that the firms without bank affiliations may not constitute an appropriate control group for the firms with bank affiliations.
I use several approaches to address that issue. First, I control for a range of firm characteristics that were likely related to resiliency, such as leverage and size, as well as the average personal wealth of the directors. Second, I estimate specifications in which I include a direct measure of the risk the firm would fail: the credit rating of the firm itself. And, finally, I use inverse propensity scores to re‐weight the treated and control firms so that observable characteristics of both resemble those of the population. In each case the effect of a bank affiliation remains, although of a diminished magnitude.
NOTE
This research brief is based on Eric Hilt, “Banks, Insider Connections, and Industrialization in New England: Evidence from the Panic of 1873,” NBER Working Paper no. 24792, July 2018, http://www.nber.org/papers/w24792.
