Jocelyn La Fleur

A Dynamic Game of EV Infrastructure Buildout

Abstract: I estimate a dynamic, structural econometric game of the EV charging station infrastructure buildout in the United States. Charge point operators (CPOs) are able to make decisions such as deciding which spatial market to enter, how many charging stations and types of plugs to install. I use over a year of real time data on EV drivers charging station habits collected from over 30,000 charging stations in the United States. The model is intended for policy makers and electrical grid operators to run simulations of policy counterfactuals integral towards understanding how infrastructure is built, maintained, and degrades over time.

Draft available soon here.

Consumer Sentiment and EV Charging

Abstract:

Negative electric vehicle (EV) charging sentiment has been one of the most significant hurdles for widespread adoption of EVs in the United States. Numerous reports of excessive wait times for charging, broken chargers, unexpected fluxes in charging prices, being blocked by fellow EV drivers, and the presence of non-electric automobiles when attempting to charge are just a few of the many pain points that EV drivers have encountered while charging. Using within-station variation and spatiotemporal differences across locations in a dynamic panel framework, I empirically estimate the effect of consumer sentiment on visits to charging stations, leveraging a unique dataset of over 1.3 million reviews from more than 76,000 U.S. EV charging stations. I find that when sentiment is categorized across charge point operators (CPOs), Tesla has the highest positive sentiment compared with popular CPOs such as ChargePoint and Blink. Furthermore, consumer sentiment overall is a strong and statistically significant predictor of the aggregate consumers visiting a charging station, with every additional visit to the charging station leading to a measurable decline in average sentiment for that station within a year-quarter. I also explore the political economy of EV charging station sentiment by analyzing the relationship between partisan bias and consumer sentiment and how this ultimately affects a consumer’s decision on where to charge. I find a complementary and intriguing result that while Democrat leaning counties have higher adoption of EVs overall, they are more likely to express negative sentiment in their charging station reviews.

Quantifying the Effect of Spatial Spillover Sentiment on the Utilization of U.S Charging Stations

with Beia Spiller & Ben Leard

Abstract:

Electric vehicle drivers often rely on charging station reviews to decide where to charge, weighing proximity against perceived station quality. With access to rating apps, drivers can now compare nearby stations in real time, allowing for reviews to create spatial spillovers. Sentiment about surrounding stations can now shape utilization at a focal station, even when the focal station’s own characteristics remain unchanged. This study estimates a spatial spillover model to isolate and quantify whether average consumer sentiment within a surrounding radius affects weekly utilization at individual U.S. charging stations. Using more than 1.3 million PlugShare reviews linked to utilization and reliability data for Level 3 charging stations, we construct a spatially weighted sentiment index and use station reliability as an instrument to estimate the effect of sentiment on station use. Preliminary results suggest that lower sentiment at surrounding stations is associated with higher utilization at the focal station, consistent with consumers substituting away from nearby stations viewed less favorably. These findings highlight the role of crowd-sourced information platforms in shaping charging behavior and infrastructure performance.

Network Effects and Price Dispersion in the U.S. Solar Panel Installation Industry

with Kenneth Gillingham & Naim Darghouth

Abstract:

A key aspect of equilibrium price dispersion of homogeneous goods is the heterogeneity of consumer information on the prices offered within the market. Consumers can become informed on price information within their respective market by facing search costs to retrieve information on the prices currently being offered by firms in the market. One potential method for consumer’s to retrieve information on the price distribution within a market is from interacting with their networks, however the relationship between network effects and price dispersion has not been explored greatly in empirical industrial organization literature. This paper seeks to elucidate the reason for the severe price dispersion currently occurring in the U.S. residential solar PV system installation industry, an industry that, simultaneously, benefits from positive network effects increasing adoption rates of installations. In this paper, we ask whether clusters of installations grouped together by spatiotemporal similarity face higher price dispersion if they have less network effects compared to clusters with significant network effects. We also ask whether network effects are a significant contributing factor to price dispersion in solar PV system clusters as opposed to other economic mechanisms such as a supply effect. Using a rich dataset containing more than 2 million solar panel installations installed in the U.S. from 2000 to 2021, we estimate the role that network effects play in price dispersion in the U.S. solar PV market. We estimate a reduced form model to determine whether network effects still remain significant in the presence of a firm’s supply effect. We find that network effects are statistically significant even in the presence of firm’s supply effects. Network effects are also stronger when we have many consumers tightly clustered around the central consumer within a cluster, so everyone benefits from having lower prices in solar. Finally, we find consumers further away from this central consumer face higher prices than those who are closer to the central consumer. Through this paper, we provide a new empirical method for capturing and analyzing network effects for industries that significantly benefit from network effects for production adoption. This is also the first study to provide empirical results of the role that network effects play, in price dispersion, corroborating the current evidence that only exists in theoretical economics literature.

Gender pay gaps in the young adult labor force: Prejudice-based discrimination or misreading of the observed to offered wage relationship

with Michael J Camasso & Radha Jagannathan (2024)

Oxford Economic Papers, 76(4), pages 1168-1188

DOI

Modeling the employment decisions of young men and women in nine European countries: An application of random utility theory and revealed preference

with Radha Jagannathan, Michael J Camasso & Simona Monteleone (2024)

Economic Analysis and Policy, 82(C), pages 233-247

DOI