Research

I study how public debt shapes fiscal policy transmission. Using local projections for the United States, I find that government spending has smaller effects on output and consumption when households hold more government debt relative to GDP. To understand this finding, I develop a heterogeneous-agent New Keynesian model in which government bonds provide households with safe assets for self-insurance. A larger bond supply raises the equilibrium real return, changes the distribution of liquid assets, and lowers intertemporal marginal propensities to consume (iMPCs), weakening the private-consumption amplification of government spending. The decline in iMPCs arises through a policy-function channel (repricing) and a distributional channel. Repricing captures the effect of changes in real rates and taxes on household consumption-saving rules, while the distributional channel reflects how higher public debt shifts households away from constrained, high-iMPC states. A decomposition shows that each channel explains about half of the decline in iMPCs, and the real rate accounts for almost all of the repricing channel. Fiscal effectiveness therefore depends not only on public debt itself, but also on how it affects household liquidity and equilibrium prices.

Understanding how policies can stabilize household welfare during recessions requires a framework that captures household heterogeneity, unemployment risk, and general-equilibrium labor market dynamics. We study a contractionary demand shock in a Heterogeneous-Agent New-Keynesian model with search-and-matching friction on the labor market (HANK-SAM) and compare the effectiveness of alternative income-stabilization policies. Using a common fiscal envelope, we compare increases in unemployment insurance generosity, targeted transfers to hand-to-mouth households, and labor-income tax cuts. Policy effectiveness is assessed through aggregate household welfare, measured in consumption-equivalent variation units. In an economy calibrated to U.S. data, unemployment insurance and labor-income tax cuts generate larger welfare gains than targeted transfers under realized fiscal costs. We also compare a state-contingent rule with an ad hoc policy path for each instrument, conditional on the downturn having occurred. The two forms of implementations generate similar welfare gains once direct fiscal costs are equalized.

This paper studies how production networks shape household liquid wealth inequality. In a heterogeneous-agent economy with input-output linkages, we characterize the Consumer Leontief Multiplier (CLM) as a sufficient statistic for labor intensity under the model’s production and demand assumptions. A higher CLM lowers labor intensity as the real unit profit rises relative to the real wage. In calibrated counterfactuals, network changes that raise the CLM increase liquid wealth inequality under benchmark dividend incidence. The effect depends on households’ general equilibrium responses and the distribution of dividends: incidence favoring high-skill households amplifies inequality, whereas equal incidence can reverse the effect. Further, we validate these model predictions on the basis of descriptive empirical evidence. Both country-specific calibrations and panel evidence for European economies are consistent with the relationships between production network structure, income allocation, and the household wealth distribution: higher CLM is associated with lower labor intensity and greater concentration of wealth.

In this paper we provide evidence that there are statistical and economically meaningful differences in terms of attitudes towards risk at the aggregate level across countries, as captured by country-specific estimations of the coefficient of relative risk aversion. This has important implications for fiscal policy as it leads to large differences in the output response to the same fiscal policy shock. When calibrating the risk aversion at the country level, using country-specific estimates of the coefficient of relative risk aversion, we find multipliers to the same fiscal consolidation shock to differ as much as between 0.35 and 0.55.