This paper conducts an empirical investigation of the effects of temporary versus persistent fiscal policy shocks. Using data from the US I show that short lived fiscal expansions have a positive effect on output and consumption; while persistent fiscal shocks generate negative effects on consumption and - to a lesser extent – on output. Persistent fiscal expansions are associated with an increase in precautionary savings, collapse in consumers' confidence and an increase the yield curve's term premium. Consistently with consumption smoothing, short-lived fiscal expansions generate a temporary deficit in the current account, while persistent fiscal shocks leave the external balance unaffected. I find evidence of nonlinearity in the effects of temporary and persistent fiscal shocks according to (i) the level of public debt and (ii) the state of the business cycle. Persistent fiscal shocks have larger negative effects on consumption and output if they take place at high levels of public debt, possibly due to the higher costs of fiscal stabilization. The state of the business cycle is also very relevant. Persistent fiscal shocks generate negative multipliers in times of economic boom, but these negative multipliers disappear in periods of recession when credit constraints are more likely to bind. On the other hand, temporary fiscal shocks have positive effects both in times of expansion and in times of recession, but with the multipliers being way larger in the latter case. Differently with what hypothesized by the earlier literature on "non-Keynesian" scale effects, I find little evidence that these effects are asymmetric depending on the size of the shocks.