The extent of the market for early American bank notes

How far did early American bank notes travel? This question has vexed banking historians for decades. Using newly digitized data from 18 Pennsylvania banks and a small-town bank note broker, I find that notes tended to circulate within a 50 mile radius of the issuing bank. Notes rarely circulated 200 miles from their point of origin. When they traveled, notes tended to follow trade, mostly along the regions rivers and canals. 

Forthcoming in  Journal of Economic History (June 2026)

Earlier version appeared as: NBER working paper 31886 (November 2023)

Silverback bankers

Despite a large literature concerning the nature and effects of boards of directors, relatively little is known about the length of board service (tenure) and its consequences for firm behavior. This paper investigates board tenure at US banks in the nineteenth and early twentieth century. The paper first introduces a new data set of director service and characteristics. It then uses those data to explore three issues: (1) how a director’s personal and professional characteristics are related to length of service; (2) whether the replacement of board members is correlated with bank contemporaneous performance; and (3) whether long-term service affects bank profitability and risk taking. I find that occupation and outside board service are strong predictors of tenure; director spells are more likely to end when dividend distributions are lower; and longer tenure increases bank profitability and risk taking. The evidence points to the importance of a previously overlooked characteristic of boards for corporate behavior. 

Working paper (July 2018)

Short-term loans and long-term relationships revisited

Abstract: Commercial banking is an industry in which long-term, nonexclusive relationships between bankers and their borrowers are governed by short-term contracts. This paper documents the nature of bank-borrower relationships in an historical setting. We use a new dataset of 55,000 loans to 10,000 unique borrowers over 35 years in the mid-nineteenth century. Two-way fixed effects models reveal that a one standard deviation increase in the number of previous loans leads to an increase in loan size by 4.72%. A one-year increase in the duration of a relationship leads to an increase in loan size of 5.6%. Relationships have small effects on loan rates, but substantial effects on collateral. Borrowers with that borrower from competing banks receive substantially smaller loans. Propensity score matching reveals that banks learn about borrowers early in the relationship. Finally, the bank severed relationships with about one-fourth of borrowers who defaulted on a loan. The relationship effects on loan size, interest rates, and collateral were also altered after default for those relationships that were not terminated.