Discretionary Taxes and Public Investment when Government Promises are not Enforceable

We characterize Markov-perfect equilibria in a setting where the absence of government commitment affects the financing of productive public capital. We show that at any date, a government in office only considers intertemporal distortions over two consecutive periods in choosing taxes. We then use our framework to quantify the value of commitment, which we define as that obtained from binding governments to a course of actions that produce the second-best allocations. Because this calculation relies on numerical approximations, we contrast alternative approaches in the literature. We find that both linear quadratic and perturbation methods deliver accurate steady states, but that the former can yield spurious policy implications along the transition. Ultimately, our analysis suggests that very small costs of setting up a commitment technology are enough to prevent its adoption. Furthermore, while households’ decision to forego government commitment may be rational at some initial date, it is nevertheless the case that consumption allocations may differ considerably in the long run. JEL Classification: E61, E62, H11