This paper analyzes the size of government spending multiplier in two policy mix cases: Active Monetary/Passive Fiscal policy (regime M) in the first instance and Passive Monetary/Active Fiscal policy (regime F), in the sense of Leeper (1991), in the second. I develop a New-Keynesian model where preferences are subject to external deep habits and where some households do not have access to financial markets. I show that these two specifications allow for the crowding in of private consumption in both regimes. However, the private investment falls in regime M while it rises in regime F as a response to a government spending shock. In addition, I show that the impact multiplier increases with the degree of deep habits in regime M, while it decreases in regime F. In this framework, in a low nominal interest rate environment, the government spending multiplier is not too large as vast studies show. However, I find that the global effectiveness of government spending is larger in regime F than in regime M, even though the impact multiplier is greater than unity in both regimes.