Dine in or Takeout? Trends on Restaurant Service Demand amid the COVID-19 Pandemic.

This article focuses on analyzing the impact of the COVID-19 pandemic on restaurant service demand, particularly the shift between dine-in and takeout modes, using foot traffic data from the Washington metropolitan area. It develops a gamma mixture model (GMM) applied to grouped dwelling-time intervals from SafeGraph mobile phone data to decompose aggregate restaurant visits into dine-in and takeout volumes, addressing challenges such as unlabeled visit modes and low-traffic variability through a regularized expectation-maximization (EM) algorithm. The study finds that limited-service and budget restaurants recovered faster due to their adaptability to takeout channels, while full-service and premium restaurants showed more robust long-term demand for dine-in services. Regional differences emerged, with exurban areas trending more toward takeout, whereas urban areas exhibited less modal shift. The findings highlight ongoing structural transformations in the restaurant industry driven by pandemic-related behavioral changes and the expansion of online food delivery platforms, underscoring implications for employment, business strategies, and policy regulation.