A for Effort? Using the Crowd to Identify Moral Hazard in New York City Restaurant Hygiene Inspections

From an upset stomach to a life-threatening foodborne illness, getting sick is all too common after eating in restaurants. Although health inspection programs for restaurants are designed to protect consumers, such inspections typically occur sporadically, allowing restaurant hygiene to remain unknown for diners. At the same time, online reviews for restaurants provide a valuable source of information about the current status and quality of a restaurant. In this paper, we use the text contained in these reviews of restaurants to effectively identify cases of hygiene violations in restaurants, even after the restaurants have been inspected. Using data about restaurant hygiene in New York City from 2010 through 2016 and the associated set of online reviews for the same set of restaurants from Yelp, we use supervised machine learning techniques to develop a hygiene dictionary specifically crafted to identify hygiene-related problems. With this dictionary, we report systematic instances of moral hazard, wherein restaurants with positive hygiene inspection scores are seen to regress in their hygiene maintenance within 90 days of receiving the inspection scores. Based on these results, we provide strategies for how cities and policymakers may design effective restaurant inspection programs.