My work on narratives documents how individuals adopt models of the world and proposes theories that can rationalize the observed patterns.
We study the dynamics of model adoption during the Covid-19 pandemic. We show that individuals switch beliefs about the effectiveness of preventive behaviors following changes in perceived risk. The adoption of models promoting preventive behaviors is procyclical and model switching is influenced by exposure to conflicting information. We explain the data using a heterogeneous-agent model of competing narratives in which agents exhibit motivated beliefs. Adopting misspecified models increases infection rates, highlighting the importance of promoting accurate beliefs to guide behavior.
My theoretical work on large games focuses on characterizing equilibrium and providing robust comparative statics when agents have heterogeneous incentives.
I study large games played by heterogeneous agents whose payoffs depend on the aggregate action such as Cournot competition, public goods and externalities, search models and coordination games. I prove the equivalence between the global games selection and potential maximization in games with strategic complementarities, and characterize the selected equilibrium as the solution to maximizing the ex-ante expected payoffs of a player with pessimistic beliefs. I also derive novel comparative statics results that can be used to analyze both parameter changes and the impact of heterogeneity.
outdated version! Some of the results in the paper are now special cases of the more general results in the paper "Large Aggregative Games with Heterogeneous Players"
My work in this area focuses on modeling contract enforcement under asymmetric information, with an emphasis on credit contracts, and on understanding the effect of technological progress in credit markets.
We study the negative feedback loop between the aggregate default rate and the efficacy of enforcement in a model of debt-financed entrepreneurial activity. The novel feature of our model is that enforcement capacity is accumulated ex ante and thus subject to depletion ex post. In the model default decisions by entrepreneurs are strategic complements, leading to multiple equilibria. We propose a global game selection to overcome equilibrium indeterminacy and show how shocks that deplete enforcement capacity can lead to a spike in the aggregate default rate and trigger credit rationing.
(previously circulated under the title: Credit Enforcement Cycles)
We investigate the role of information technology (IT) in the collection of delinquent consumer debt. We argue that the widespread adoption of IT by the debt collection industry in the 1990s contributed to the observed expansion of risky lending to consumers, such as unsecured credit card lending. We stress the role of private information about a delinquent borrower’s solvency. IT is used to focus collection efforts on those delinquent borrowers who are likely to repay.
AEA Research Highlight: The Quiet Revolution in Debt Collections (by Tim Hyde)
About 20% of credit card debt is transferred annually to new cards through low-interest balance transfer offers. We propose a dynamic model of repricing that accounts for the observed level of balance transfers and unveils a novel source of macroeconomic fragility in consumer credit markets. We quantify the potential of this channel to account for the observed deleveraging in the credit card market during the 2007-09 financial crisis.
My research explores the role of information and beliefs in shaping prices in financial and speculative markets and the link between information and risk preferences.
Individuals likely face uncertainty about underlying risks in insurance and financial markets. To characterize demand for financial products, it is therefore necessary to understand preferences toward both known and uncertain risks. We survey a representative sample of US households to estimate such preferences. We find that probability weighting is much more heterogeneous in the risk domain than in the uncertainty domain. While individuals overweigh small probabilities and underweigh large probabilities across domains, probability distortion is lower among individuals with higher financial literacy and cognitive ability. We suggest how to account for unobserved uncertainty when estimating risk preferences from observational data.
Technological advances allow insurers to be better informed about underlying risks than consumers. We evaluate the impact of these information frictions on welfare by combining demand elicitation surveys with insurance claim data. We find an ‘uncertainty premium’ -i.e., consumers are willing to pay more for insurance when risks are uncertain. We show that the uncertainty premium is negatively correlated with risk aversion, leading to a selection effect: individuals who purchase insurance are not necessarily the most risk averse. We show that the resulting misallocation of insurance leads to large welfare losses.
We study belief heterogeneity in asset markets and show it leads to the favorite-longshot bias (FLB). We show how equilibrium maps onto a demand system that can be identified and estimated using market-level data. The FLB is explained by a two-type population: agents with correct beliefs and noise traders with significant belief dispersion. Variation in public information shows that belief heterogeneity outperforms preference-based explanations of the FLB.
I study a model of speculative markets populated by sophisticated and naive traders. Three pricing regimes arise in equilibrium: perfect pricing, partial mispricing and complete mispricing. Perfect pricing obtains when sophisticated traders are predominant in the population. When the fraction of naive traders is not too low mispricing arises and exhibits a systematic pattern of overpricing low value assets and underpricing high value assets.
I show that in bilateral bargaining with asymmetric information no equilibrium with positive probability of trade exists in any trade environment with pure common values. I also provide a counterexample in which trade happens with probability one showing that no trade results do not hold when there are more than two traders.
My empirical work in this area explores some implications of contract regulation in the Spanish labor market, which is characterized by very high unemployment rates and low job mobility.
We examine the poverty implications of temporary employment. Our findings suggest that fixed-term contracts are linked to a greater poverty exposure among women and older men relative to open-ended contracts. This greater poverty exposure can last several years due to feedback effects operating via job instability or via the transition to work statuses characterized by higher risk of falling into poverty.
We show that wage growth among employees with indefinite work contracts largely occurs via job mobility, whereas fixed-term workers gain via job mobility as well as on-the-job. Consequently, job stayers with fixed-term contracts narrow their wage gap over time with respect to similar workers with open-ended contracts. Yet, this effect is solely driven by fixed-term workers with 6-months contracts that are able to keep their jobs beyond their initial contract period.
Short descriptive piece about fixed-term employment and poverty incidence in Spain.