RAPID Choices under Short-Term Threats and Behavioral Response to Social Distancing in the COVID-19 Pandemic

必须遵守社交限制，以遏制新冠肺炎等高传染性疾病的传播。然而，确保遵守严格的卫生干预措施，需要更好地理解风险下的个人决定。在研制出疫苗之前，政策制定者不仅需要为人们找到避免身体接触的最佳激励措施，而且他们还需要制定计划，在不久的将来放松社会距离。本项目中开发的模型中的行为预测有助于在这两种情况下指导决策。具体地说，该项目为几个与政策相关的问题提供了答案，例如：人们正在采取哪些社会行为变化来应对新冠肺炎爆发造成的破坏？谁更有可能遵守健康指南，包括社会距离协议？人们对公共场所(药店、杂货店、餐馆的外卖)的实际接近的安全/风险认知是什么？这些认知如何影响日常选择？如何激励个人保持与社会的距离？新冠肺炎中断是如何导致个人福利损失的，分析师如何得出衡量此类损失的指标？在这个项目中，研究团队采用并显著采用了零售和交通研究中的拥挤研究工具和方法，以分析社会距离行为作为预防对健康威胁的行动。一项创新的基于虚拟现实的在线调查收集了时间敏感的行为数据，并对社交距离进行了选择实验。数据采用微观计量经济学离散-连续选择模型进行建模，其中态度分量采用结构方程，偏好冲击采用重尾误差分布。与标准的细尾分布不同，出现重尾的误差项不仅概括了标准假设，还解决了决策不确定性行为。风险下的柔性决策规则被集成到离散-连续选择模型中，该模型表示完全封锁和部分封锁期间的时间使用调度。最终，这项研究的研究成果为政策制定者提供了指导，指导他们如何最好地实施社会距离和隔离措施，以控制主要流行病，然后如何最好地逐步淘汰这些措施。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Social-interaction restrictions have to be respected to contain the spread of highly contagious diseases such as COVID-19. Ensuring compliance to strict health interventions, however, needs better understanding of individual decisions under risk. Until a vaccine is developed, policy makers not only need to find the best incentives for people to avoid physical proximity, but they also need to create plans for relaxing social distancing in the near future. Behavioral forecasts from the models developed in this project help to guide decisions in both cases. Specifically, this project provides answers to several policy-relevant questions such as: What changes in social behavior are people adopting in response to the disruptions created by the COVID-19 outbreak? Who is more likely to respect health guidelines, including social distancing protocols? What are safety/risk perceptions of physical proximity in public spaces (pharmacies, grocery stores, take-outs at restaurants) and how are these perceptions shaping daily choices? How to create incentives for individuals to sustain social-distancing? How are COVID-19 disruptions causing losses in individual welfare, and how can analysts derive metrics to value such losses? In this project, the research team adopts and significantly adapts crowding research tools and methods in retail and transportation studies to analyze social distancing behaviors as preventive action against threats to health. An innovate virtual-reality-based online survey with choice experiments on social distancing collects time-sensitive behavioral data. The data are modeled using micro-econometric discrete-continuous choice models with structural equations for attitudinal components and heavy-tailed error distributions for preference shocks. Unlike standard thin-tailed distributions, error terms that exhibit heavy tails not only generalize standard assumptions but also address decision-uncertainty behavior. Flexible decision rules under risk are integrated into the discrete-continuous choice models that represent time-use scheduling during total and partial lock-downs. Ultimately, research outcomes from this study provide guidance to policy-makers for how to best implement measures such as social distancing and quarantines in order to control major epidemics, and then how to best phase out these measures.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.