Credit spread and the transmission of government purchases shocks

I estimate government purchases multipliers using structural VAR analysis, featuring credit spreads and real-time forecast data to purge spending innovations of their anticipated components. The cumulative output multiplier of a temporary rise in government purchases is about 2.0 after year 5. I propose an explanation for these observed effects based on a dynamic stochastic general equilibrium model with financial intermediation. The key to this framework is the differential efficacy of fiscal policy dependent on the tightness of financing constraints. The cumulative multipliers are about or higher than one in regimes when financing constraints on banks bind tightly. In contrast, in times when financing constraints are loose the multipliers are smaller than one. The result can be explained by the crowding-in of private investment following an increase in government consumption in regimes of tighter financing constraints.