We investigate the effects of government expenditure on private consumption when the private sector anticipates the fiscal shocks. In order to capture anticipation of fiscal policy, we develop a new method based on a structural vector autoregression (SVAR). By simulating data from a theoretical model featuring (imperfect) fiscal foresight, we demonstrate the ability of our new approach to correctly capture macroeconomic dynamics. We take advantage of the flexibility of our econometric approach and study those subcomponents of total government spending, which have different macroeconomic effects according to economic theory. Using post-WWII US data, we find that when taking into account anticipation, private consumption significantly decreases in response to a defense expenditure shock, whereas when considering shocks to non-defense spending, consumption increases significantly. A standard SVAR does not produce clear consumption responses, highlighting the importance of anticipation. Our results thus reconcile the different findings of the narrative and SVAR approaches to the study of fiscal policy effects.