讲座主题：Law Enforcement and Roadway Safety: Street-Block-Level Causal Evidence and Optimal Resource Allocation in Seattle
讲座嘉宾：严佳（华盛顿州立大学）
讲座摘要：While existing work documents a causal link between traffic enforcement and road safety at broad geographic scales (e.g., states and counties), its localized impact within cities—and how it varies across neighborhoods and street environments—remains underexplored. This seminar presents a street-block-level study of the heterogeneous causal relationship between law enforcement activity—officer presence and traffic stops—and traffic accidents. Using quarterly data for 49,020 local street segments in Seattle from 2021–2022, we assemble an Urban Big Data platform that integrates GIS road networks, crash records, police activity, and point-of-interest (POI) and neighborhood characteristics.
Our empirical design is guided by a framework of sequential utility maximization and spatial sorting: heterogeneous drivers choose routes based on amenities and travel costs and then choose driving intensity as a function of localized enforcement risk and individual preferences. Because police deployment responds endogenously to local conditions, we exploit variation in non-traffic crime workload (e.g., trespass, theft, burglary) to predict officer presence and construct a deployment-based control function. Under a triangular structure in which deployment is upstream of stops, and conditional on the deployment shock, remaining variation in stop intensity is treated as exogenous for crashes.
To recover high-dimensional, spatially varying effects without imposing global linearity, we combine cross-fitted nuisance estimation with generalized random forests to estimate conditional average treatment effects (CATE). The results provide granular evidence on where enforcement most strongly reduces accidents, offering a data-driven framework for targeted, neighborhood-level traffic safety policy and police resource allocation.
讲座时间：2026年3月16日（星期一），上午9：00-11：00
线上会议：腾讯会议（会议号：801-856-033）
嘉宾简介：严佳教授现为华盛顿州立大学经济学院终身教授，研究领域为应用微观经济学和应用计量经济学。他的学术论文发表于Econometrica, Journal of Econometrics，American Economic Journal: Economic Policy, Journal of Public Economics, Journal of Urban Economics, Journal of Environmental Economics and Management, Journal of Law and Economics, International Journal of Industrial Organization, Decision Support Systems, Transportation Research Part A and B 等经济学、信息管理与交通工程顶尖期刊，并出版了多部著作。严佳教授曾获2009年美国交通论坛最佳论文奖、2011年国际交通经济学会最佳论文奖等多项奖项。严佳教授还曾担任美国ChoiceStream公司算法科学家，开发应用于Yahoo和美国在线（AOL）的新闻、电影和音乐推荐引擎，具有丰富的大数据分析行业经验。
此次学术活动获“中央高校基本科研业务费资助”（supported by “the Fundamental Research Funds for the Central Universities”），特此致谢。
撰稿：王昱洁 排版：王昱洁
初审：孙宝文 复审：王红梅
