借鉴各国应用案例：在最关键的时候衡量福利

Learning fromcountry applications Contents Acknowledgments This edited volume was prepared by a team from the World Bank Poverty GlobalDepartment consisting of Kimberly Bolch, Maria Eugenia Genoni, and HenryStemmler. The work was conducted under the supervision of Luis Felipe López-Calva (Global Director, Poverty Department) and Gabrielainchauste (PracticeManager, Poverty Department). The contributed chapters were authored by EmilyAiken, Joshua Blumenstock, Erwin Knippenberg, Walker Kosmidou-Bradley, MoritzMeyer, William Seitz, Christina Wieser, Nobuo Yoshida, and Kazusa Yoshimura. This document benefitted from consultations with many members of the PovertyGlobal Department as well as other World Bank teams who led the developmentand implementation of many of the initiatives referenced here. The team is par-ticularly grateful to Alemayehu Ambel, Aziz Atamanov, Oscar Barriga, Paul Corral,Yeon Soo Kim, Erwin Knippenberg, Walker Kosmidou-Bradley, William Seitz, TaraVishwanath, Christina Wieser, and Nobuo Yoshida for serving as chapter reviewersthroughout the drafting and revision process. Additionally, the team would also like to thank Federico Haslop for his researchsupport and Juliana Soares for her support in organizing the production pro-cess. This volume received editing support from Robert Zimmerman and designservices from Carlos Reyes and Gabriel Lora. This volume benefitted from finan-cial support provided by the Umbrella Facility for Poverty and Equity. Introduction in a global context marked by heightened uncertainty, the ability to act on reli-able, up-to-date information is more essential than ever. As emphasized by theWorld Development Report 2021: Data for Better Lives, data is a foundational inputfor development, enabling governments to tailor policies to people’s needs andrespond effectively to shocks. Yet in many countries, the information required toguide decisions on poverty and vulnerability is not available when it is needed most. Traditional household surveys, which underpin official poverty estimates, remainindispensable—but are often conducted too infrequently to inform timely policyaction. This is particularly true in low-income countries and in fragile and conflictaffected situations, where surveys may be implemented with even larger lags dueto financial and operational constraints. In response to this challenge, the World Bank’s Poverty and Equity Global Practice(GP) has been deploying innovative approaches for more timely welfare monitor-ing. These approaches typically work by integrating traditional surveys (“baselinedata”) with alternative high-frequency data sources (“auxiliary data”) and apply-ing a range of modelling approaches (Figurei.1). While most approaches rely ona strategic combination of these three elements, others focus solely on monitor-ing welfare through the direct collection or use of more high-frequency data (forexample, by implementing rapid surveys or using administrative data). While thedevelopment and testing of these approaches has been ongoing for around adecade, it was accelerated by recent crises, such as the COVID-19 pandemic andclimate-related disasters. In recent years, these methods have also increasinglybeen able to leverage more frontier data sources and methodologies, such as bigdata and machine learning. in 2023, the Poverty Global Department launched an initiative to take stock of thisgrowing body of knowledge. What did we know about which real-time monitor-ing (RTM) approaches worked best in different settings? A key milestone in thisagenda was the publication of Measuring Welfare Whenit Matters Most: A typol-ogy of approaches for real-time monitoring.That publication mapped out thebroader landscape of existing RTM approaches, reflecting on relevant use casesand caveats and providing a summary of key methodological resources. The aimwas to guide practitioners in choosing the most context-appropriate tools toanswer their questions. This edited volume was prepared as a complement to that publication, aimed atthose readers interested in learning more about how RTM approaches have beenpractically applied on the ground.Measuring Welfare When it Matters Most:Learningfrom Country Applicationsdelves deeper into selected examples, offering a moredetailed look at how to design and implement high-frequency monitoring systemsin different types of country settings and in response to different types of policyquestions. The chapters walk the reader through these case studies and reflect onmethodological best practices, practical challenges, and lessons learned. This volume includes five chaptersfeaturing country examples from across mul-tiple regions—including Eastern & Southern Africa, Central Asia, and South Asia.Each application focuses on a different type of RTM approach.The first threechapters showcase how different types of “rapid surveys” can be used to col-lect new high-frequency data, exploring examples of how a decentralized modelcan enab