[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-118218-en":3,"doc-seo-118218-105":30,"detail-sidebar-cat-0-en-105":95},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},118218,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Targeting Social Protection Programs with Machine Learning and Digital Data - Thesis/Dissertation","Social protection programs are essential for assisting poor populations, yet governments and humanitarian agencies often cannot cover everyone in need. Accurate benefit targeting depends on reliable poverty information, which is commonly scarce, outdated, or incomplete in low-income settings. This dissertation develops and evaluates machine-learning approaches using satellite imagery and mobile phone data to target aid where traditional survey data are unavailable. Studies across Togo, Afghanistan, and Bangladesh compare machine-learning targeting against geography or occupation-based eligibility and against conventional survey measurement.","UC Berkeley  \nUC Berkeley Electronic Theses and Dissertations  \nTitle  \nTargeting Social Protection Programs with Machine Learning and Digital Data  \nPermalink  \n[https://escholarship.org/uc/item/0cj4c266](https://escholarship.org/uc/item/0cj4c266)  \nAuthor  \nAiken, Emily  \nPublication Date  \n2024  \nPeer reviewed|Thesis/dissertation  \n[eScholarship.org](eScholarship.org) Powered by the California Digital Library  \nUniversity of California  \nTargeting Social Protection Programs with Machine Learning and Digital Data  \nby  \nEmily Aiken  \nA dissertation submitted in partial satisfaction of the requirements for the degree of  \nDoctor of Philosophy  \nin  \nInformation Science  \nin the  \nGraduate Division  \nof the  \nUniversity of California, Berkeley  \nCommittee in charge:  \nProfessor Joshua Blumenstock, Chair  \nProfessor Jennifer Chayes  \nProfessor Hany Farid  \nPofessor Solomon Hsiang  \nProfessor Edward Miguel  \nSpring 2024  \nTargeting Social Protection Programs with Machine Learning and Digital Data  \nCopyright 2024  \nby  \nEmily Aiken  \n1  \nAbstract  \nTargeting Social Protection Programs with Machine Learning and Digital Data  \nby  \nEmily Aiken  \nDoctor of Philosophy in Information Science  \nUniversity of California, Berkeley  \nProfessor Joshua Blumenstock, Chair  \nSocial protection programs are essential to assisting the poor, but governments and humanitarian agencies are rarely resourced to provide aid to all those in need, so accurate targeting of benefits is critical. In developed economies, targeting decisions typically rely on administrative income data or broad survey-based social registries. In lowincome countries, however, poverty information is rarely reliable, comprehensive, or up-to-date. Novel sources of digital data — from mobile phones and satellites, in particular—are well suited to fill this gap: they are predictive of wealth in low-income contexts and ubiquitously collected. The research studies in this dissertation design and evaluate new methods for targeting aid in low-resource contexts using machine learning, satellite imagery, and mobile phone data, and evaluate these methods in large, real-world interventions. Across social protection programs in Togo, Afghanistan, and Bangladesh, the studies in this dissertation show that targeting methods based on machine learning and digital data sources identify poor households more accurately than methods based on categorical eligibility criteria like geography or occupation, but typically less accurately than traditional survey-based poverty measurement approaches. These results highlight the potential for digital data and machine learning to improve the targeting of humanitarian aid, particularly when traditional poverty data are unavailable or out-of-date and in settings where conflict, environmental conditions, or health concerns render primary data collection infeasible. These studies also provide empirical evidence on the limitations and risks of digital and algorithmic targeting approaches, including privacy, transparency, fairness, and digital exclusion.  \ni  \nContents  \nContents i  \nAcknowledgements iii  \n1 Introduction 1  \n1.1 Traditional poverty targeting approaches in low-income contexts ...... 2  \n1.2 Evaluating poverty targeting approaches .................... 4  \n1.3 Measuring poverty with machine learning and digital data .......... 6  \n1.4 Uses of digital data for targeting social protection programs ......... 8  \n1.5 Contributions of this dissertation ......................... 8  \n2 Targeting aid with machine learning and digital data in Togo 11  \n2.1 Introduction and context .............................. 12  \n2.2 Methods ....................................... 16  \n2.3 Results ........................................ 32  \n2.4 Discussion ...................................... 47  \n3 Ultra-poverty targeting with machine learning and phone data in Afghanistan 52  \n3.1 Introduction and context .............................. 52  \n3.2 Methods .............","cbCailUlmnNnbf64","https://ap.wps.com/l/cbCailUlmnNnbf64","pdf",22990943,1,205,"English","en",105,"# Introduction\n## Traditional poverty targeting approaches in low-income contexts\n## Evaluating poverty targeting approaches\n## Measuring poverty with machine learning and digital data\n## Uses of digital data for targeting social protection programs\n## Contributions of this dissertation\n# Targeting aid with machine learning and digital data in Togo\n## Introduction and context\n## Methods\n## Results\n## Discussion\n# Ultra-poverty targeting with machine learning and phone data in Afghanistan\n## Introduction and context\n## Methods\n## Results\n## Discussion\n# Comparing community-based and phone-based targeting in Bangladesh\n## Introduction and context\n## Methods\n## Results\n## Discussion\n# Measuring cash transfer impacts with surveys versus digital traces\n## Introduction and context\n## The GiveDirectly-Novissi program: Design and data\n## Program impacts estimated using survey data\n## Program impacts estimated using mobile phone data\n## Discussion\n# Discussion\n## Directions for future work\n## Conclusion","[{\"question\":\"Why is targeting social protection programs considered critical in low-income settings?\",\"answer\":\"Because governments and humanitarian agencies are rarely resourced to provide aid to all those in need, accurate targeting of benefits is essential. Low-income countries also often lack reliable, comprehensive, and up-to-date poverty information.\"},{\"question\":\"What digital data sources does the dissertation use for poverty and targeting decisions?\",\"answer\":\"It uses novel digital data sources, particularly satellite imagery and mobile phone data. These sources are described as predictive of wealth in low-income contexts and are ubiquitously collected.\"},{\"question\":\"How do machine-learning targeting methods compare with categorical eligibility criteria and traditional surveys?\",\"answer\":\"Across Togo, Afghanistan, and Bangladesh, machine-learning and digital-data-based methods identify poor households more accurately than categorical eligibility criteria such as geography or occupation. However, they are typically less accurate than traditional survey-based poverty measurement approaches.\"},{\"question\":\"What limitations and risks are evaluated for digital and algorithmic targeting?\",\"answer\":\"The dissertation provides empirical evidence on limitations and risks, including privacy, transparency, fairness, and digital exclusion. These concerns are highlighted as important when deploying digital-algorithmic targeting in humanitarian settings.\"}]","Targeting Social Protection Programs with Machine Learning and Digital Data - Thesis/Dissertation | PDF",1785682320,517,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":90,"head_meta":92,"extra_data":94,"updated_unix":28},"targeting-social-protection-programs-with-machine-learning-and-digital-data-thesisdissertation","",{"@graph":36,"@context":89},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/targeting-social-protection-programs-with-machine-learning-and-digital-data-thesisdissertation/118218/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81,85],{"name":72,"@type":73,"acceptedAnswer":74},"Why is targeting social protection programs considered critical in low-income settings?","Question",{"text":75,"@type":76},"Because governments and humanitarian agencies are rarely resourced to provide aid to all those in need, accurate targeting of benefits is essential. Low-income countries also often lack reliable, comprehensive, and up-to-date poverty information.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What digital data sources does the dissertation use for poverty and targeting decisions?",{"text":80,"@type":76},"It uses novel digital data sources, particularly satellite imagery and mobile phone data. These sources are described as predictive of wealth in low-income contexts and are ubiquitously collected.",{"name":82,"@type":73,"acceptedAnswer":83},"How do machine-learning targeting methods compare with categorical eligibility criteria and traditional surveys?",{"text":84,"@type":76},"Across Togo, Afghanistan, and Bangladesh, machine-learning and digital-data-based methods identify poor households more accurately than categorical eligibility criteria such as geography or occupation. However, they are typically less accurate than traditional survey-based poverty measurement approaches.",{"name":86,"@type":73,"acceptedAnswer":87},"What limitations and risks are evaluated for digital and algorithmic targeting?",{"text":88,"@type":76},"The dissertation provides empirical evidence on limitations and risks, including privacy, transparency, fairness, and digital exclusion. These concerns are highlighted as important when deploying digital-algorithmic targeting in humanitarian settings.","https://schema.org",{"og:url":52,"og:type":91,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":93,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":96},[97,101,105,109,114,119,124,127,132,135,139],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":106,"show_sort_weight":107,"slug":108},"Exam",70,"exam",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},5,"Comic",60,"comic",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},6,"Technology",50,"technology",{"id":120,"doc_module":4,"doc_module_name":46,"category_name":121,"show_sort_weight":122,"slug":123},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":125,"slug":126},30,"research-report",{"id":128,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":130,"slug":131},9,"Religion & Spirituality",20,"religion-spirituality",{"id":130,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":130,"slug":134},"World Cup","world-cup",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":136,"slug":138},10,"Lifestyle","lifestyle",{"id":140,"doc_module":4,"doc_module_name":46,"category_name":141,"show_sort_weight":110,"slug":142},19,"General","general"]