Research

Working Papers

Colluding against Environmental Regulation (with Cuicui Chen, Jing Li, and Mathias Reynaert). 

Revise and Resubmit at the Review of Economic Studies. 

We study collusion among firms against imperfectly monitored environmental regulation. Firms increase variable profits by violating regulation and reduce expected noncompliance penalties by violating jointly. We consider a case of three German automakers colluding to reduce the effectiveness of emission control technology. By estimating a structural model of the European automobile industry from 2007 to 2018, we find that the collusion lowers expected noncompliance penalties substantially and increases buyer and producer surplus. Welfare decreases by EUR 0.73–2.51 billion because of increased pollution. We show how environmental policy design and antitrust play complementary roles in preventing noncompliance. 

Market Structure and the Distributional Implications of Product Bans  (with Fernando Luco and Eve Colson-Sihra)

Product bans affect market competition and possibly the demand for the whole product category by changing consumers' perception of product quality. We analyze the ban of Nestlé's Maggi, the largest producer of instant noodles in India, for alleged non-compliance with health safety standards. Maggi returned to the stores six months later when The High Court at Bombay ruled the ban unjustified. We document substantial heterogeneity in consumers' responses to the ban across socioeconomic categories. We also show the presence of negative spillovers to competitors, which led them to decrease their sales in the first months of the ban. A structural model of consumer and firm behavior allows us to identify demand and supply responses to the ban separately. Counterfactual simulations show a substantial impact of the ban on consumer surplus, with a 33.4 percent reduction on average in the two years that followed the ban relative to the pre-ban mean. Twenty percent of the overall effect was due to spillovers of the ban to the entire instant noodles category, while the rest resulted from changes in market structure. Turning to distributional impacts, the overall effect was 10.5 percentage points larger in high-income markets than in low-income ones, while demand spillovers were 2 percentage points larger in low-income markets than in high-income markets. 

The Impact of Recommendation Systems on Experts: Evidence from Physicians (with Juan Pablo Atal, Alejandra Benítez, and Martín González-Cabello)

How do experts respond to computerized recommendation systems? We study a workers' compensation insurance program covering 40% of the Chilean population, where physicians decide whether a medical visit is work-related and therefore is fully covered by law. We quantify the effects of a computerized recommendation system on physicians' coverage decisions. The alert improves physicians' decisions as measured by decision reversals by a reviewers' committee. We show that the alert system changed physicians' preferences for coverage. Our results help to understand how experts process system recommendations and provide lessons to assess their desirability and how to design them.

Physician Experience and Learning by Doing in a New Hospital (with Pablo Varas)

This paper studies the effects of worker experience on firm productivity analyzing learning by doing among physicians in the emergency department of a new hospital. We measure productivity using total times in the emergency room, medical images, and referrals to specialists. We find large causal effects of physician experience on productivity using other physicians allocation as a source of exogenous variation on current physician assignment. Higher physician experience cause lower visit times, less specialist referrals and less utilization of X-rays. This effect comes from experience acquired both within the emergency department and from other experience, proxied by physician age. We also find that more recent experience has a larger effect than more distant one.  Finally, we study how physician productivity relates to organizational learning, and explore the temporal trade-offs of different physician allocation policies.

Social and Economic Activity Predicts Daily Hospital Emergency Visits (with Juan Pablo Atal and Patricio Domínguez) 

We show a strong  effect of social and economic activity on emergency room utilization using daily data from healthcare facilities in Chile and novel data on activity. Our results imply that the crisis-induced changes in mobility explain a large portion of the 50 percent drop in non-respiratory emergency room visits in the country during the COVID-19 pandemic. We find no corresponding increase in mortality during our study period. Our results provide novel evidence on the relationship between healthcare utilization and the economy, and suggest that mobility restrictions during health crises can benefit public health by freeing up healthcare resources.

Gradually Rebuilding a Relationship: The Emergence of Collusion in Retail Pharmacies in Chile 

This paper studies the emergence of collusion among the three main retail pharmacy chains in Chile. The firms were able to collude on a more profitable equilibrium raising prices of hundreds of drugs gradually over time. Survival models show that collusion was more likely to occur first on products in which the pharmacies were more differentiated. This result is consistent with trust building, as collusion on differentiated products is safer due to smaller losses should the collusive scheme collapse. The decrease in the pharmacies’ monitoring activity of the collusive price increases provides further support of initial mistrust.

Collusive Price Leadership in Retail Pharmacies in Chile

I analyze price leadership among retail pharmacy chains in Chile during a case of price war and collusion on hundreds of drugs. The pharmacies reached higher prices by means of staggered price increases that were mostly led by the smallest chain. I find that both the order of move and the time it took the follower firms to raise prices after the leader's increase are explained by the largest firm’s size in that market. I portray collusive leadership as a screening mechanism that allows the leader to truthfully reveal its type, the implications of which are consistent with my empirical findings.

Work in Progress

1. The Incentives of Patients and Providers in Out-of-Network Emergency Visits  (with Juan Pablo Atal)

2. Information Flows in Healthcare Teams: Evidence from a Field Experiment (with Juan Pablo Atal, Naomi Friedman-Sokuler, and Hannah (Trachtman) Luk-Zilberman)