We study the effect of financial relationships on firms' investment decisions and access to external finance. In the early twentieth century, securities underwriters commonly held directorships with American corporations. Section 10 of the Clayton Antitrust Act prohibited bankers from serving on the boards of railroads for which they underwrote securities. We find that following the implementation of Section 10, railroads with strong preexisting relationships with underwriters saw declines in their investment rates, valuations, and leverage, and increases in their costs of external funds. Reassuringly, we do not observe similar effects among industrials and utilities, which were not subject to Section 10. Our results are consistent with underwriters on corporate boards acting as delegated monitors, and highlight the potential for regulations intended to address conflicts of interest to disrupt valuable information flows. (JEL G24, G31, G32, G34, K22, N21, N22)

We study the effects of the Liberty Bond drives of World War I on financial intermediation in the 1920s and beyond. Using panel data on US counties, and an instrument that captures differences in the approaches used to market the bonds, we find that higher Liberty Bond subscription rates led to an increase in investment banks and a contraction in commercial bank assets. We also find that in the late 1930s, individuals residing in states where Liberty Bond subscription rates had been higher were more likely to report owning stocks or bonds. Although they were conducted to support the American effort in World War I, these bond drives reshaped American finance.

We study the role of war bonds and inflation in post-WWII federal elections. After WWII, major bouts of inflation in 1946-48 and 1950-51 depressed the real returns of the bonds sold to households during the war. In a difference-in-differences framework, we find that counties with higher war bond purchases shifted their votes towards the Republican Party in postwar elections. To address the endogeneity of war bond purchases, we use an instrumental variables design, and obtain similar results. Bond ownership raised the saliency of inflation, and the severe rise in prices after the war triggered a backlash against the incumbent Democrats.

We study the effects of a contraction in financial intermediation on nonfinancial firms. The Panic of 1907 originated in the shadow banks of the time, New York’s trust companies. The runs were caused by a shock unrelated to the trust companies’ nonfinancial corporate clients. In the years following the panic, corporations affiliated with the worst-affected trusts made fewer capital investments, paid lower dividends, and suffered lower profitability and higher borrowing costs relative to firms without such connections. The shock to New York’s trust companies accounted for at least 18.4 percent of the decline in corporate investment in the United States in 1908.