Research

A Theory of Downward Wage Rigidity

Working Paper, 2025

I develop a model where workers are averse to losses as in the cognitive psychology literature, and need to search in order to find a job. In a frictional market in which both workers and firms determine the terms of labor contracts, nominal downward wage rigidity emerges endogenously as a result of the privately optimal division of gains from trade. The model implies that the response of wages to shocks is asymmetric. In response to a temporary negative productivity shock that is not too large, nominal wages are initially rigid and take some time to catch up. In response to a symmetric positive shock, firms increase nominal wages immediately but let real wages erode over time. Inflation “greases the wheels of the labor market”, in the sense that the inaction region is smaller in a high-inflation environment. The model rationalizes a number of additional empirical regularities: (1) wages of job-switchers are more flexible than wages of job-stayers, but not conditional on employment history, (2) the Phillips curve is nonlinear, and (3) the probability of wage changes is state-dependent. Moreover, a calibration to US microdata yields a good fit to the distribution of nominal wage changes with parameters that are consistent with common estimates. The model prescribes an optimal positive inflation target, and a countercyclical response to shocks.

Download Paper

Behavioral Sticky Prices

Joint with Sergio Rebelo and Pedro Teles. Journal of Monetary Economics, 2025

We develop a model in which households make decisions using a dual-process framework. System 1 relies on fast, intuitive heuristics but is prone to error, while System 2 demands cognitive effort but yields more accurate decisions. Monopolistic firms can influence which system households engage through pricing. This strategic influence creates a novel source of price inertia. The model accounts for the “rockets and feathers” phenomenon (prices rise quickly but fall slowly), explains why firms with unexpectedly high demand often avoid price changes, and why hazard functions are downward sloping. Our model implies that price stability is not optimal.

Download Paper

Sales, Reference Prices, and Consumer Behavior

Joint with Sergio Rebelo, Pedro Teles, Miguel Godinho de Matos, and Pedro Amorim. Work in Progress, 2026

While nominal prices change frequently, many of these changes are due to temporary sales. These sales induce short-lived deviations from a stable reference price. Once a sale ends, prices often revert to the prior reference price, which tends to remain fixed for extended periods. This paper demonstrates that this two-tier pricing structure, consisting of a reference price and a sale price, emerges naturally in a model where households rely on a dual-process cognitive system. In familiar settings, consumers rely on the intuitive but error-prone System 1, whereas unfamiliar situations activate the more deliberate and accurate System 2. Monopolistic firms exploit consumers’ reliance on System 1 by strategically alternating between reference and sale prices in response to cost shocks.

Behavioral Expectations

Draft Coming Soon, 2026

I study a model where firms form inflation expectations using dual-process reasoning. When the state of the economy is familiar, firms employ System 1 - a fast, heuristic mode of thinking that is prone to systematic errors. When the state of the economy is unfamiliar, firms trigger System 2 - a deliberative mode of thinking that is more accurate, but also more costly. The model features a Phillips curve with a state- and history-dependent slope and, due to strategic complementarities in learning, can exhibit multiple equilibria. The optimal policy response to cost-push shocks involves front-loading the cognitive costs associated with System 2 and, for shocks that are not too large, greater price stability than under rationality.