How Has the Effect of Government Spending Changed? Evidence from High-Frequency Identified Shocks Since 1947

This paper investigates the time-varying effects of government spending on the U.S. economy. Using a novel identification strategy that combines text analysis with daily stock market data, I construct a new measure of defense spending news spanning 1947–2019, which serves as an instrument for aggregate government spending. The new measure provides richer variation than existing measures, allowing for more detailed examination of how the effects of government spending evolve over time. Empirical analysis shows that the output multiplier fluctuates substantially over time, ranging from 0.8 to 1.5, largely driven by differences in private investment responses. The key mechanism lies in the intrinsic nature of the spending-its composition and time profile. When a policy announcement signals a persistent and back-loaded expansion (time profile) of procurement spending (composition), private investment rises strongly, leading to larger output multipliers. I further demonstrate that incorporating these spending characteristics is essential for a New Keynesian model to rationalize the empirical findings. The results highlight an important policy implication that the effectiveness of government spending is not static, and policymakers need to consider the underlying nature of spending plans, not just their size.

Procurement Heterogeneity and the Transmission of Government Spending  (with James Cloyne)

Government Procurement and Firm R&D (with James Cloyne, Xian Jiang)

Sectoral Effects of Budget Authority by Agency

Demographics and Housing Cycle (with Jaeyoung Yoo, Jinho Kim, Yungu Cho)

Tariff News and Inflation Expectations (with Daihong Min, Yungu Cho)