How effective is public investment in stimulating output?

Empirical estimates of the public investment multiplier vary widely across studies, raising questions about the conditions under which public investment is most effective in stimulating output. This paper argues that a key source of this heterogeneity is the initial stock of productive public capital. In a neoclassical model, we show theoretically that the magnitude of the public investment multiplier depends critically on the initial stock of productive public capital. The model predicts that the public investment multiplier is high (low) when the stock of public capital is low (high). For a sample of European countries and relying on local projection methods (Jordà, 2005), we validate the theoretical predictions of the neoclassical model and find that the size of the public investment multiplier indeed depends negatively on the initial stock of public capital. These results suggest that fiscal stimulus programs based on public investment spending should be more effective in countries with limited public infrastructure, and less effective in more advanced economies where public capital is abundant.