Working Papers

Durables and Size-dependence in the Marginal Propensity to Spend (with Martin Beraja)

Abstract: Stimulus checks have become an increasingly important policy tool in recent U.S. recessions. How does the households' marginal propensity to spend (MPX) vary as checks become larger? To quantify this size-dependence in the MPX, we augment a canonical model of durable spending by introducing a smooth adjustment hazard. We discipline this hazard by matching a rich set of micro moments. We find that the MPX declines slowly with the size of checks. In contrast, the MPX is flatter in a purely state-dependent model of durables, and declines sharply in a two-asset model of non-durables. Finally, we embed our spending model into an open-economy heterogeneous-agent New-Keynesian model. In a typical recession, a large check of $2,000 increases output by 25 cents per dollar, compared to 37 cents for a $300 check. Large checks thus remain effective but extrapolating from the response out of small checks overestimates their impact.

Inefficient Automation (with Martin Beraja) (accepted at The Review of Economic Studies)

Abstract: How should the government respond to automation? We study this question in a heterogeneousagent model that takes worker displacement seriously. We recognize that displaced workers face two frictions in practice: reallocation is slow and borrowing is limited. We first show that these frictions result in inefficient automation. Firms are effectively too patient when they automate, and (partly) overlook the time it takes for workers to reallocate and for the benefits of automation to materialize. We then analyze a second best problem where the government can tax automation but lacks redistributive tools to fully overcome borrowing frictions. The equilibrium is (constrained) inefficient — automation and reallocation impose pecuniary externalities on workers. The government finds it optimal to tax automation while labor reallocates, even when it has no preference for redistribution. Using a quantitative version of our model, we find significant welfare gains from slowing down automation.

Investment Dynamics and Cyclical Redistribution

Abstract: Demand for durable goods and residential investment is strongly pro-cyclical. Workers employed in durable industries are imperfectly insured against these fluctuations, leading to distributional consequences during booms and busts. This paper studies the interaction between the cyclicality of durable demand and redistribution of labor income. I explore this feedback loop within a heterogeneous agent New Keynesian (HANK) model with multiple sectors and lumpy durable adjustment. Income redistribution emerges endogenously when labor mobility between sectors is limited. Crucially, lumpy adjustment at the micro level generates non-linearities at the macro level: the average marginal propensity to spend (MPC) on durables varies with the size of income shocks. As a result, redistribution of labor incomes has aggregate effects. I find that the interaction between cyclical investment and redistribution amplifies the aggregate response of durable investment during booms and dampens it during recessions. The lumpy nature of durable adjustment entirely accounts for this non-linear effect.