Our paper investigates the sectoral effects of government spending shocks and highlights the role of labor mobility. Our VAR evidence for sixteen OECD countries reveals that a shock to government consumption by 1% of GDP increases non-traded value added by 0.7% of GDP and generates a decline in traded value added. The value added share of non-tradables rises by 0.35% of GDP, thus implying that the reallocation of resources accounts for 50% of the sectoral fiscal multiplier. Consistently, our estimates show that the non-traded sector is highly intensive in the government spending shock and experiences a labor inflow. The shift of hours worked toward the non-traded sector is, however, subject to mobility costs which vary across countries. When we explore quantitatively the sectoral effects of a shock to government consumption that is highly intensive in non-traded goods, we find that the model can replicate the magnitude of the rise in the share of non-tradables we document empirically once we allow for both labor mobility and capital installation costs. Financial openness also matters as it further biases the demand shock toward non-tradables. To account for the cross-country dispersion in the responses of sectoral shares we estimate empirically, we have to let the degree of labor mobility vary across countries.