Zhongying Gan

Job Market Paper:

"Inspecting the Inspectors: The Causes and Consequences of Restaurant Inspector Leniency"
(sole author) [link]
Abstract: This paper studies inspector leniency as a challenge to regulation enforcement in the context of Los Angeles County restaurant hygiene inspections. Inspectors score restaurants numerically, but only letter grades reflecting wide score intervals are mandatorily disclosed. It is documented in the literature that inspectors bump scores up for restaurants at the margin of a higher grade, referred to as inspector leniency. This study shows that inspector leniency has compromised the effectiveness of health grades in signaling restaurants' hygiene conditions, as restaurants whose grades are inflated to A have significantly worse hygiene conditions than restaurants who have earned their A. Moreover, inspector leniency discourages hygiene improvements by preventing a re-inspection that a restaurant would likely request if they did not experience grade inflation. The paper then uncovers a novel motive behind inspector leniency: inspectors avoiding work that does not increase promotion prospects, and proposes changes to the inspector performance evaluation policy as a remedy. The paper follows by revisiting  a well-documented motive, attachment to the restaurants formed through repeated inspections, and proposes more frequent inspector rotation as a remedy. The two drastically different remedies highlight the importance of understanding inspector motives in improving enforcement. Lastly, I evaluate the effects of these two remedies on inspectors' grading and restaurants' sanitation efforts under a theoretical framework.

Published Paper:

"Do Electric Vehicle Charger Locations Respond to the Potential Charging Demands from Multi-Unit Dwellings? Evidence from Los Angeles County," Transport Policy, 138 (2023): 74-93.
(sole author) [link] [Replication Files] [Ken Small Award for Best Student Paper in Transportation Economics at UC Irvine]
Abstract: With the transportation sector being the largest contributor of greenhouse gas emissions in the United States, states are actively planning the deployment of plug-in electric vehicles (PEVs). Mass adoption of PEVs requires attracting potential buyers living in multi-unit dwellings (MUDs). Given the current low adoption rate of PEVs in MUDs and MUD residents’ reliance on public charging, this paper studies whether there is a positive correlation between the number of public L2 chargers and MUD density (measured by total square footage of MUD per capita) across census block groups (CBGs) in LA County. The results show that high MUD-density CBGs and low MUD-density CBGs do not differ much in terms of the number of chargers. The charger-to-PEV-ratio range for MUD residents derived in this paper is below the ideal charger-to-PEV-ratio range in the literature. A direct policy implication is that more charging infrastructure should be made available to MUD residents. This includes public Level 2 chargers near MUDs, which are discussed in this paper, and onsite MUD charging, which is still at an early stage.

Work in Progress:

"Freeway Congestion and Labor Force Participation: Evidence from the Greater Los Angeles Area" (sole author)
Project description: Traffic congestion is a growing problem in major metropolitan areas. This paper aims to deepen the understanding of the economic costs of congestion by estimating the effects of traffic congestion along commute routes on labor force participation at high geographic granularity. The project attempts to answer the following three questions: 1) Is an individual's preference on the workplace location affected by congestion along the commute? More specifically, for a given residence census tract, does the percentage of residents commuting to a certain workplace census tract decrease following an increase in congestion along the commute route? 2) Does congestion affect the extensive margin of labor force participation? More specifically, does the labor force participation rate in a given residence census tract decrease following an increase in average congestion (weighted by job flows) along the commute routes out of the residence census tract? 3) Does congestion affect the characteristics of the job an individual takes? More specifically, how does the distribution of jobs held by individuals residing in a given census tract change in terms of earnings and industry sector following an increase in congestion along the commute routes out of the census tract? To answer these questions, I construct an annual panel dataset from 2013 to 2019, using census-tract-level data from Longitudinal Employer-Household Dynamics Origin-Destination Employment Statistics (LODES), and novel data on congestion along commute routes from tract to tract. The identification relies partly on the temporal variations in congestion. Unfortunately, web mapping services commonly used in the urban and transportation literature, such as Google Maps, are unable to retrieve historical travel time. I overcome this challenge of estimating the historical congestion level by incorporating historical freeway speed data from traffic detectors accessed from the Caltrans Freeway Performance Measurement System (PeMS) into the open-sourced OpenStreetMap.