Ramey (2011a) and others argue that increases in government spending associated with wars and military build-ups constitute a good instrument for measuring the macroeconomic effects of fiscal shocks. We argue that this instrument has two important drawbacks: the composition of government...