Monetary policy and the added worker effect

We exploit cross-sectional variation in the response of US states to a monetary policy shock to study how the impact of monetary policy varies with the share of married women who work. We find that the economy's response is more muted the lower the share of married women employed before the shock. We argue that a plausible explanation is a shielded demand response by households, insured by the "added worker effect". When women are only weakly attached to the labor market, they can flexibly enter and exit to supplement household income in times of need, providing a powerful insurance mechanism against aggregate shocks. We provide three additional pieces of evidence. First, monetary policy shocks have a stronger effect in states where married women are more firmly attached to the labor market (making fewer transitions in and out). Second, following an increase in the federal funds rate, married women themselves are comparatively more likely to be employed (and to enter employment) in states where the share of married women working pre- shock is low. Third, in contrast to employment, wages of married women fall more in states where married women have worked less, consistent with a differential labor supply response to the shock.