Using a simple stochastic growth model that nests both exogenous and endogenous growth, this paper shows that the growth rate should be mean stationary if growth is exogenous and difference stationary if growth is endogenous and any variable affecting investment is difference stationary. Permanent changes in the share of output devoted to government consumption should permanently affect the growth rate if and only if growth is endogenous. I test these implications and find no evidence that growth is endogenous. Furthermore, even if growth is endogenous, the evidence indicates that its degree of endogeneity is likely to be small.
