Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
The aggregate spending multiplier crucially depends on the sectoral origin of government purchases. To establish this result, we characterize analytically the response of aggregate output to sector-specific government spending shocks in a tractable production-network economy, showing how it maps into various characteristics of the shocked sector. The response is larger when government spending originates in sectors with a relatively small contribution to private final demand, low markup, high labor intensity, and in those located downstream in the supply chain. We confirm these predictions and evaluate their quantitative relevance within a calibrated multi-sector model of the U.S. economy that embeds several dimensions of sectoral heterogeneity. Leveraging this model, we illustrate how differences in the sectoral composition of purchases across U.S. government levels lead to large variation in the spending multiplier. The latter ranges from 0.47 for federal defense spending, which is relatively concentrated in upstream capital-intensive manufacturing, to 0.82 for state and local spending, which is mainly oriented towards downstream labor-intensive services. Finally, we exploit heterogeneity in the sectoral composition of military spending across U.S. states to provide empirical evidence supporting our theoretical predictions.