| Title | Counterfactual Policy Optimization: Intersections of Predictive/Experimental/Theoretical Approaches |
|---|---|
| Date | 2018年12月5日（水）、12月12日（水）両日とも13:00-16:40 13:00-16:40, December 5 and 12, 2018 |
| Venue | 東京大学経済学研究科・経済学研究棟３階・第３教室にて開催 [地図] Lecture Room No. 3, Economics Research Building, University of Tokyo [MAP] |
| Lecturer | 成田悠輔（Yusuke Narita, Yale University and CREPE) |
| Abstract | What methods should we use for optimizing policy interventions and predicting the counterfactual performance? This course aims for students to understand and combine different types of methods: 1) "Theoretical" or "structural" approach 2) "(Quasi-)experimental," "atheoretical," or "reduced-form" approach 3) "Predictive," "machine-learning," or "Rashomon" approach 4) Old and new attempts to mix (1)(2)(3) for better evaluations and predictions I will explain not only how methods work in theory but also how they are used in empirical applications in a variety of fields and topics (business, development, education, finance, health, industrial organization, labor, marketing, medicine, psychology, public finance; see the bold questions in the following Roadmap). Participants will finish the course equipped with a workman’s familiarity with many key tools of empirical economics and policy analysis, and—hopefully—a good understanding of how best to combine state-of-the-art methods into answering the questions you want to answer. |
| Information | 講義は英語で行われます（Lecture in English） |