Publication: Program Targeting with Machine Learning and Mobile Phone Data: Evidence from an Anti-Poverty Intervention in Afghanistan
Loading...
Date
2022-12
ISSN
Published
2022-12
Editor(s)
Abstract
Can mobile phone data improve program targeting By combining rich survey data from the baseline of a “big push” anti-poverty program in Afghanistan implemented in 2016 with detailed mobile phone logs from program beneficiaries, this paper studies the extent to which machine learning methods can accurately differentiate ultra-poor households eligible for program benefits from ineligible households. The paper shows that machine learning methods leveraging mobile phone data can identify ultra-poor households nearly as accurately as survey-based measures of consumption and wealth; and that combining survey-based measures with mobile phone data produces classifications more accurate than those based on a single data source.
Link to Data Set
Citation
“Aiken, Emily L.; Bedoya, Guadalupe; Blumenstock, Joshua E.; Coville, Aidan. 2022. Program Targeting with Machine Learning and Mobile Phone Data: Evidence from an Anti-Poverty Intervention in Afghanistan. Policy Research Working Papers;10252. © World Bank. http://hdl.handle.net/10986/38491 License: CC BY 3.0 IGO.”
Associated URLs
Associated content
Other publications in this report series
Publication The Future of Poverty(Washington, DC: World Bank, 2025-07-15)Climate change is increasingly acknowledged as a critical issue with far-reaching socioeconomic implications that extend well beyond environmental concerns. Among the most pressing challenges is its impact on global poverty. This paper projects the potential impacts of unmitigated climate change on global poverty rates between 2023 and 2050. Building on a study that provided a detailed analysis of how temperature changes affect economic productivity, this paper integrates those findings with binned data from 217 countries, sourced from the World Bank’s Poverty and Inequality Platform. By simulating poverty rates and the number of poor under two climate change scenarios, the paper uncovers some alarming trends. One of the primary findings is that the number of people living in extreme poverty worldwide could be nearly doubled due to climate change. In all scenarios, Sub-Saharan Africa is projected to bear the brunt, contributing the largest number of poor people, with estimates ranging between 40.5 million and 73.5 million by 2050. Another significant finding is the disproportionate impact of inequality on poverty. Even small increases in inequality can lead to substantial rises in poverty levels. For instance, if every country’s Gini coefficient increases by just 1 percent between 2022 and 2050, an additional 8.8 million people could be pushed below the international poverty line by 2050. In a more extreme scenario, where every country’s Gini coefficient increases by 10 percent between 2022 and 2050, the number of people falling into poverty could rise by an additional 148.8 million relative to the baseline scenario. These findings underscore the urgent need for comprehensive climate policies that not only mitigate environmental impacts but also address socioeconomic vulnerabilities.Publication Exports, Labor Markets, and the Environment(Washington, DC: World Bank, 2025-07-14)What is the environmental impact of exports? Focusing on 2000–20, this paper combines customs, administrative, and census microdata to estimate employment elasticities with respect to exports. The findings show that municipalities that faced increased exports experienced faster growth in formal employment. The elasticities were 0.25 on impact, peaked at 0.4, and remained positive and significant even 10 years after the shock, pointing to a long and protracted labor market adjustment. In the long run, informal employment responds negatively to export shocks. Using a granular taxonomy for economic activities based on their environmental impact, the paper documents that environmentally risky activities have a larger share of employment than environmentally sustainable ones, and that the relationship between these activities and exports is nuanced. Over the short run, environmentally risky employment responds more strongly to exports relative to environmentally sustainable employment. However, over the long run, this pattern reverses, as the impact of exports on environmentally sustainable employment is more persistent.Publication The Macroeconomic Implications of Climate Change Impacts and Adaptation Options(Washington, DC: World Bank, 2025-05-29)Estimating the macroeconomic implications of climate change impacts and adaptation options is a topic of intense research. This paper presents a framework in the World Bank's macrostructural model to assess climate-related damages. This approach has been used in many Country Climate and Development Reports, a World Bank diagnostic that identifies priorities to ensure continued development in spite of climate change and climate policy objectives. The methodology captures a set of impact channels through which climate change affects the economy by (1) connecting a set of biophysical models to the macroeconomic model and (2) exploring a set of development and climate scenarios. The paper summarizes the results for five countries, highlighting the sources and magnitudes of their vulnerability --- with estimated gross domestic product losses in 2050 exceeding 10 percent of gross domestic product in some countries and scenarios, although only a small set of impact channels is included. The paper also presents estimates of the macroeconomic gains from sector-level adaptation interventions, considering their upfront costs and avoided climate impacts and finding significant net gross domestic product gains from adaptation opportunities identified in the Country Climate and Development Reports. Finally, the paper discusses the limits of current modeling approaches, and their complementarity with empirical approaches based on historical data series. The integrated modeling approach proposed in this paper can inform policymakers as they make proactive decisions on climate change adaptation and resilience.Publication The Asymmetric Bank Distress Amplifier of Recessions(Washington, DC: World Bank, 2025-07-11)One defining feature of financial crises, evident in U.S. and international data, is asymmetric bank distress—concentrated losses on a subset of banks. This paper proposes a model in which shocks to borrowers’ productivity dispersion lead to asymmetric bank losses. The framework exhibits a “bank distress amplifier,” exacerbating economic downturns by causing costly bank failures and raising uncertainty about the solvency of banks, thereby pushing banks to deleverage. Quantitative analysis shows that the bank distress amplifier doubles investment decline and increases the spread by 2.5 times during the Great Recession compared to a standard financial accelerator model. The mechanism helps explain how a seemingly small shock can sometimes trigger a large crisis.Publication Impact of Heat Waves on Learning Outcomes and the Role of Conditional Cash Transfers(Washington, DC: World Bank, 2025-07-14)This paper evaluates the impact of higher temperatures on learning outcomes in Peru. The results suggest that 1 degree above 20°C is equivalent to 7 and 6 percent of a standard deviation of what a student learns in a year for math and reading tests, respectively. These results hold true when the main specification is changed, splitting the sample, collapsing the data at school level, and using other climate specifications. The paper aims to improve understanding of how to deal with the impacts of climate change on learning outcomes in developing countries. The evidence suggests that conditional cash transfer programs can mitigate the negative effects of higher temperatures on students’ learning outcomes in math and reading.
Journal
Journal Volume
Journal Issue
Collections
Related items
Showing items related by metadata.
Publication Applying Machine Learning and Geolocation Techniques to Social Media Data (Twitter) to Develop a Resource for Urban Planning(World Bank, Washington, DC, 2020-12)With all the recent attention focused on big data, it is easy to overlook that basic vital statistics remain difficult to obtain in most of the world. This project set out to test whether an openly available dataset (Twitter) could be transformed into a resource for urban planning and development. The hypothesis is tested by creating road traffic crash location data, which are scarce in most resource-poor environments but essential for addressing the number one cause of mortality for children over age five and young adults. The research project scraped 874,588 traffic-related tweets in Nairobi, Kenya, applied a machine learning model to capture the occurrence of a crash, and developed an improved geoparsing algorithm to identify its location. The project geolocated 32,991 crash reports in Twitter for 2012-20 and clustered them into 22,872 unique crashes to produce one of the first crash maps for Nairobi. A motorcycle delivery service was dispatched in real-time to verify a subset of crashes, showing 92 percent accuracy. Using a spatial clustering algorithm, portions of the road network (less than 1 percent) were identified where 50 percent of the geolocated crashes occurred. Even with limitations in the representativeness of the data, the results can provide urban planners useful information to target road safety improvements where resources are limited.Publication Learning from the Experiments That Never Happened : Lessons from Trying to Conduct Randomized Evaluations of Matching Grant Programs in Africa(World Bank, Washington, DC, 2012-12)Matching grants are one of the most common policy instruments used by developing country governments to try to foster technological upgrading, innovation, exports, use of business development services and other activities leading to firm growth. However, since they involve subsidizing firms, the risk is that they could crowd out private investment, subsidizing activities that firms were planning to undertake anyway, or lead to pure private gains, rather than generating the public gains that justify government intervention. As a result, rigorous evaluation of the effects of such programs is important. The authors attempted to implement randomized experiments to evaluate the impact of seven matching grant programs offered in six African countries, but in each case were unable to complete an experimental evaluation. One critique of randomized experiments is publication bias, whereby only those experiments with "interesting" results get published. The hope is to mitigate this bias by learning from the experiments that never happened. This paper describes the three main proximate reasons for lack of implementation: continued project delays, politicians not willing to allow random assignment, and low program take-up; and then delves into the underlying causes of these occurring. Political economy, overly stringent eligibility criteria that do not take account of where value-added may be highest, a lack of attention to detail in "last mile" issues, incentives facing project implementation staff, and the way impact evaluations are funded, and all help explain the failure of randomization. Lessons are drawn from these experiences for both the implementation and the possible evaluation of future projects.Publication The Enduring Impacts of a Big Push during Multiple Crises(Washington, DC: World Bank, 2023-11-07)How do proven strategies to improve the economic conditions of ultra-poor households hold up against the increasing severity and co-incidence of economic, security, and climate shocks Five years after receiving an economic livelihoods package, and shortly prior to the 2021 regime change, “ultra-poor” women in Afghanistan continued to have significantly higher levels of consumption, assets, market work participation, financial inclusion, children’s school enrollment, and women’s psychological well-being and empowerment, relative to the control group. Households boost resilience by diversifying productive activities and the program improves equality by reducing the gaps between ultra-poor and non-ultra- poor households across multiple dimensions. The results illustrate how an increasingly popular approach to improve the conditions of the very poor through a one-off “big push” intervention can strengthen household resilience through multiple shocks in one of the most fragile settings worldwide.Publication No Household Left Behind(World Bank, Washington, DC, 2019-06)The share of people living in extreme poverty fell from 36 percent in 1990 to 10 percent in 2015 but has continued to increase in many fragile and conflict-affected areas where half of the extreme poor are expected to reside by 2030. These areas are also where the least evidence exists on how to tackle poverty. This paper investigates whether the Targeting the Ultra Poor program can lift households out of poverty in a fragile context: Afghanistan. In 80 villages in Balkh province, 1,219 of the poorest households were randomly assigned to a treatment or control group. Women in treatment households received a one-off "big-push" package, including a transfer of livestock assets, cash consumption stipend, skills training, and coaching. One year after the program ended -- two years after assets were transferred -- significant and large impacts are found across all the primary pre-specified outcomes: consumption, assets, psychological well-being, total time spent working, financial inclusion, and women's empowerment. Per capita consumption increases by 30 percent (USD 24 purchasing power parity, USD 7 nominal per month) with respect to the control group, and the share of households below the national poverty line decreases from 82 percent in the control group to 62 percent in the treatment group. Using modest assumptions about consumption impacts, the intervention has an estimated internal rate of return of 26 percent, excluding non-monetized improvements in psychological well-being, women's empowerment, and children's health and education. These findings suggest that "big-push" interventions can dramatically reduce poverty in fragile and conflict-affected regions.Publication Challenges and Opportunities of Mobile Phone-Based Data Collection : Evidence from South Sudan(World Bank, Washington, DC, 2013-01)The proliferation of mobile phones in developing countries has generated a wave of interest in collecting high-frequency socioeconomic surveys using this technology. This paper considers lessons from one such survey effort in a difficult environment -- the South Sudan Experimental Phone Survey, which gathered data on living conditions, access to services, and citizen attitudes via monthly interviews by phones provided to respondents. Non-response, particularly in later rounds of the survey, was a substantial problem, largely due to erratic functioning of the mobile network. However, selection due to non-response does not appear to have markedly affected survey results. Response rates were much higher for respondents who owned their own phones. Both compensation provided to respondents in the form of airtime and the type of phone (solar-charged or traditional) were varied experimentally. The type of phone was uncorrelated with response rates and, contrary to expectation, attrition was slightly higher for those receiving the higher level of compensation. The South Sudan Experimental Phone Survey experience suggests that mobile phones can be a viable means of data collection for some purposes, that calling people on their own phones is preferred to handing out phones, and that careful attention should be given to the potential for selective non-response.
Users also downloaded
Showing related downloaded files
Publication World Development Report 2024(Washington, DC: World Bank, 2024-08-01)Middle-income countries are in a race against time. Many of them have done well since the 1990s to escape low-income levels and eradicate extreme poverty, leading to the perception that the last three decades have been great for development. But the ambition of the more than 100 economies with incomes per capita between US$1,100 and US$14,000 is to reach high-income status within the next generation. When assessed against this goal, their record is discouraging. Since the 1970s, income per capita in the median middle-income country has stagnated at less than a tenth of the US level. With aging populations, growing protectionism, and escalating pressures to speed up the energy transition, today’s middle-income economies face ever more daunting odds. To become advanced economies despite the growing headwinds, they will have to make miracles. Drawing on the development experience and advances in economic analysis since the 1950s, World Development Report 2024 identifies pathways for developing economies to avoid the “middle-income trap.” It points to the need for not one but two transitions for those at the middle-income level: the first from investment to infusion and the second from infusion to innovation. Governments in lower-middle-income countries must drop the habit of repeating the same investment-driven strategies and work instead to infuse modern technologies and successful business processes from around the world into their economies. This requires reshaping large swaths of those economies into globally competitive suppliers of goods and services. Upper-middle-income countries that have mastered infusion can accelerate the shift to innovation—not just borrowing ideas from the global frontiers of technology but also beginning to push the frontiers outward. This requires restructuring enterprise, work, and energy use once again, with an even greater emphasis on economic freedom, social mobility, and political contestability. Neither transition is automatic. The handful of economies that made speedy transitions from middle- to high-income status have encouraged enterprise by disciplining powerful incumbents, developed talent by rewarding merit, and capitalized on crises to alter policies and institutions that no longer suit the purposes they were once designed to serve. Today’s middle-income countries will have to do the same.Publication Western Balkans 6 Country Climate and Development Report(Washington, DC: World Bank Group, 2024-07-16)This Regional Western Balkans Countries Climate and Development Report (CCDR) stands out in several ways. In a region that often lacks cohesive regional alliances, this report emphasizes how the challenges faced across countries are often common and interconnected, and, importantly, that climate action requires coordination on multiple fronts. Simultaneously, it illustrates the differences across countries, places, and people that require targeted strategies and interventions. This report demonstrates how shocks and stressors re intensifying and how investments in adaptation could bring significant benefits in the form of avoided losses, accelerated economic potential, and amplified social and economic spillovers. Given the region’s high emission and energy intensity and the limitations of its current fossil fuel-based development model, the report articulates a path to greener and more resilient growth, a path that is more consistent with the aspiration of accession to the EU. The report finds that the net zero transition can be undertaken without compromising the economic potential of the Western Balkans and that it could lead to higher growth than under the Reference Scenario (RS) with appropriate structural reforms.Publication World Development Report 2023: Migrants, Refugees, and Societies(Washington, DC : World Bank, 2023-04-25)Migration is a development challenge. About 184 million people—2.3 percent of the world’s population—live outside of their country of nationality. Almost half of them are in low- and middle-income countries. But what lies ahead? As the world struggles to cope with global economic imbalances, diverging demographic trends, and climate change, migration will become a necessity in the decades to come for countries at all levels of income. If managed well, migration can be a force for prosperity and can help achieve the United Nations’ Sustainable Development Goals. World Development Report 2023 proposes an innovative approach to maximize the development impacts of cross-border movements on both destination and origin countries and on migrants and refugees themselves. The framework it offers, drawn from labor economics and international law, rests on a “Match and Motive Matrix” that focuses on two factors: how closely migrants’ skills and attributes match the needs of destination countries and what motives underlie their movements. This approach enables policy makers to distinguish between different types of movements and to design migration policies for each. International cooperation will be critical to the effective management of migration.Publication World Development Report 2018(Washington, DC: World Bank, 2018)Every year, the World Bank's World Development Report takes on a topic of central importance to global development. The 2018 Report, Learning to Realize Education's Promise, is the first ever devoted entirely to education. Now is an excellent time for it: education has long been critical for human welfare, but is even more so in a time of rapid economic change. The Report explores four main themes. First, education's promise: Education is a powerful instrument for eradicating poverty and promoting shared prosperity, but fulfilling its potential requires better policies - both within and outside the education system. Second, the learning crisis: Despite gains in education access, recent learning assessments show that many young people around the world, especially from poor families, are leaving school unequipped with even the most foundational skills they need for life. At the same time, internationally comparable learning assessments show that skills in many middle-income countries lag far behind what those countries aspire to. Third, promising interventions to improve learning: Research from areas such as brain science, pedagogical innovations, or school management have identified interventions that promote learning by ensuring that learners are prepared, that teachers are skilled as well as motivated, and that other inputs support the teacher-learner relationship. Fourth, learning at scale: Achieving learning throughout an education system will require more than just scaling up effective interventions. Change requires overcoming technical and political barriers by deploying salient metrics for mobilizing actors and tracking progress, building coalitions for learning, and being adaptive when implementing programs.Publication World Development Report 2004(World Bank, 2003)Too often, services fail poor people in access, in quality, and in affordability. But the fact that there are striking examples where basic services such as water, sanitation, health, education, and electricity do work for poor people means that governments and citizens can do a better job of providing them. Learning from success and understanding the sources of failure, this year’s World Development Report, argues that services can be improved by putting poor people at the center of service provision. How? By enabling the poor to monitor and discipline service providers, by amplifying their voice in policymaking, and by strengthening the incentives for providers to serve the poor. Freedom from illness and freedom from illiteracy are two of the most important ways poor people can escape from poverty. To achieve these goals, economic growth and financial resources are of course necessary, but they are not enough. The World Development Report provides a practical framework for making the services that contribute to human development work for poor people. With this framework, citizens, governments, and donors can take action and accelerate progress toward the common objective of poverty reduction, as specified in the Millennium Development Goals.