Inflation and Growth: An Inverted-U Relationship

This study explores a novel channel for monetary policy to impact growth and welfare---a cash-in-advance constraint on R&D combined with R&D subsidies by seigniorage tax. In a scale-invariant Schumpeterian growth model, growth is an inverted-U function of the inflation rate. Friedman rule is suboptimal (optimal) when the elasticity of labor supply is low (high). By contrast, the inverted-U relation does not exist when R&D is subsidized by other taxes or in an AK model. Calibration confirms our prediction and finds that the growth and welfare effects of inflation are large. Using panel data for 154 countries during 1970--2014, both non-parametric cubic spline and parametric regressions show that growth is an inverted-U function of the inflation rate in samples with an annual inflation rate below 30%. The cutoff point for inflation to have a zero marginal effect on growth is around 5% in ordinary least squares estimation and 3% in instrumental variables (IV) estimation. We also find that the share of labor employed in R&D---rather than the physical capital investment rate---is an inverted-U function of inflation in IV estimation. Our empirical evidence provides support for our theory.