Collaborative Research: A Unified Framework for Optimal Public Debt Management

Project Outcomes Statement

Fiscal and monetary tools are essential levers for macroeconomic stabilization. Much of the existing literature on optimal policy has been conducted under restrictive assumptions about household heterogeneity and financial markets. These assumptions cast doubt on the lessons derived from those analyses. This research conducted under this grant develops tools that address more realistic settings and to uses these tools to derive recommendations for tax, debt, and interest rate management. We summarize the outcomes as contributions in two lines of research.

The first line of research has developed tools to analyze models with rich household heterogeneity and incomplete insurance markets. These features are crucial for understanding the distributional effects of macroeconomic shocks. Analyzing competitive equilibria in such a context is known to be challenging. We tackled this challenge by developing a series of computational methods that are widely applicable, fast, and user-friendly. These methods build on perturbation techniques, and we extend those techniques so that they apply to settings with lots of heterogeneity. As a by-product of our analysis, we also developed a framework to decompose the welfare effects of policies into components that capture gains from redistribution, insurance, and aggregate efficiency. We applied these methods to reassess optimal fiscal and monetary policy over business cycles. Our findings overturn many of the predictions of the existing literature. Gains from using tax and monetary policies to provide insurance for households outweigh concerns of maintaining relatively constant rates of tax and inflation.

The second line of research extends to study optimal policy in settings with rich financial markets. To facilitate the tractable analysis of policy in these contexts, we propose an approach that combines insights from the "sufficient statistics" method used in public finance with small-noise expansions from computational macroeconomics. We apply these methods to study the level and maturity structure of public debt. This application results in simple formulas for optimal debt portfolios, relying on a small number of statistics that can be derived from observable macroeconomic and financial market data. When applying our approach to U.S. government debt, we find that the optimal bond portfolio shares decrease approximately exponentially with maturity, and its maturity structure, while similar to that of the U.S., has a longer duration.