We use credit card data to estimate the impact of increasing minimum payments on delinquency, payments, spending, and write-offs. Our identification strategy exploits an unusual institutional feature: Borrowers can use their account to make purchases with both revolving loans (on which minimum payments increased) and term loans (on which there was no change). Payment increases by delinquent borrowers are insufficient to match increasing minimums, resulting in lower cure rates and an increase in write-offs. Affected borrowers migrate away from these accounts by decreasing charges and increasing payments, consequently lowering the interest earned by the bank.

We examine the cross section of stock returns using an original dataset consisting of annual observations on price, dividends, and shares outstanding for nearly all stocks listed on U.K. exchanges between 1870 and 1913, supplemented with additional information about attrition. The only clear pattern in the historical U.K. data is the high returns of extremely small stocks. Among the largest 99.8% of stocks, the historical U.K. data do not display the pattern found in modern U.S. (CRSP) data of excess returns for small stocks or stocks with poor past performance. Unlike CRSP data, stocks that do not pay dividends do not outperform stocks that pay small dividends during this period. However, as in the modern data, there is a weak relation between dividend yield and performance for stocks that pay dividends.