Abstract This paper revisits the traditional hierarchy of macroeconomic stabilization tools outside of the zero lower bound. In the benchmark New-Keynesian model monetary policy is always preferred to fiscal policy because fiscal policy has relative costs without delivering any relative benefits. I explore the robustness of this standard “monetary supremacy” result in three steps.
First, motivated by empirical work documenting large marginal propensities to consume out of predictable income changes, I introduce hand-to-mouth consumers into the benchmark model. Such non-optimizing agents amplify the stimulative effects of tax and spending changes. However, I find that even in a model with large fiscal multipliers, there remains no role for fiscal policy outside of the zero lower bound. Although fiscal policy is potent in this scenario, it nevertheless diverts real resources, while monetary policy remains able to implement the first best allocation without imposing any economic distortions.
Second, I explore four other environments where fiscal policy has benefits relative to monetary policy. These include introducing uncertainty about the effectiveness of monetary and fiscal policy (appealing to the logic that it is better to use two tools with independent errors than only one tool alone), adding adjustment costs to give monetary policy a longer implementation lag than fiscal policy (matching the data on impulse responses to policy shocks), introducing heterogeneous agents with varying welfare weights (recognizing that fiscal policy can more easily target assistance), and allowing for fiscal policy to act as a signaling mechanism in a game of incomplete information and coordination failures.
Finally, I turn to the question of relative costs, appealing to the finance literature which argues that low interest rates may instigate or propagate asset bubbles, promote maturity mismatch and excessive leverage, and redistribute across agents. I propose a path for examining optimal countercyclical policy in environments where monetary policy might be forced to trade off gains in price stability and full employment against the potential costs of financial instability.
This paper empirically and theoretically analyzes the effect of debt reductions that reduce long-term but not short-term obligations. Isolating the effect of future obligations allows us to test alternative explanations for borrower default decisions and to analyze the consumption response to mortgage principal reduction for underwater borrowers. Our empirical analysis uses regression discontinuity and difference-in-differences research designs on de-identified bank account and credit bureau records from participants in the US. government’s Home Affordable Modification Program. We find that mortgage principal reductions worth an average of $70,000 have no impact on default or consumption for borrowers who remain underwater. Our results are sufficiently precise to rule out economically meaningful effects. We develop a quantitative life-cycle model that clarifies that borrowers’ short-term constraints govern their response to long-term debt obligations. When defaulting imposes utility costs in the short-term, default is driven by cash-flow shocks such as unemployment rather than by future debt burdens. When principal reductions do not push borrowers sufficiently above water so as to relax collateral constraints, consumption is unaffected because borrowers are unable to monetize increased housing wealth. Collateral constraints drive a wedge between an underwater borrower’s marginal propensity to consume out of cash and their marginal propensity to consume out of housing wealth. Our results help explain why policies that lowered current mortgage payments were more effective than principal reductions at stemming foreclosures and increasing demand during the Great Recession.
We study the spending of unemployment insurance (UI) recipients using de-identified bank account data. Nondurable spending falls by 6% at the onset of unemployment, is stable during UI receipt, and then falls by an additional 13% at benefit exhaustion. Our finding that spending responds to a large and predictable income drop is at odds with predictions from rational models, but can be matched by behavioral models with hand-to-mouth or inattentive consumers. Depressed spending after exhaustion implies that the consumption-smoothing gains from extending UI benefits are at least three times as big as the welfare gains from raising UI benefit levels.