Lockdown and restaurant closures: evidence from large-scale data in China

Abstract

This study examines the impact of the COVID-19 lockdown on China’s restaurant industry, a critical contributor to national GDP and employment. Using a large-scale dataset of 14,488,951 restaurant-year observations covering 5,560,345 unique restaurants across 301 cities (2020–2023), we employ the Cox proportional hazard model to examine how lockdown influences restaurant closure. We find that each additional 12 days of local lockdown increases the closure risk by 12.7%. While most restaurants face elevated risks, those with higher star ratings are more resilient. Chain restaurants, older establishments, higher-priced venues, and those offering unique cuisines or located near commercial hubs and transit stations are less likely to close. In contrast, newer, independent, lower-priced restaurants, especially those offering common cuisines, providing delivery, or located in less accessible areas, are more vulnerable. These findings highlight the uneven impact of lockdowns across restaurant types and locations and point to the key factors that support restaurant resilience during disruptions.