This paper presents a method to decompose the causal effect of government defense spending into: (i) a local (or direct) effect, and (ii) a spillover (or indirect) effect. Each effect is measured as a multiplier: the unit change in output of a one unit change in government spending. We apply this method to study the effect of U.S. defense spending on output using regional panel data. We estimate a positive local multiplier and a negative spillover multiplier. By construction, the sum of the local and spillover multipliers provides an estimate of the aggregate multiplier. The aggregate multiplier is close to zero and precisely estimated. We show that enlisting disaggregate data improves the precision of aggregate effect estimates, relative to using aggregate time series alone. Our paper provides a template for researchers to conduct inference about local, spillover and aggregate causal effects in a unified framework.