In 2017, the DCO awarded 13 grants to studies addressing one or more of the following research questions:
#1 What are the short-and long-run impacts (both positive and negative) of digital credit on consumers in emerging markets?
#2 Is there heterogeneity in impacts along borrower characteristics (i.e. financial literacy, time preferences, income, and/or gender)?
#3 How can non-traditional credit-scoring algorithms, regulations, and other consumer protection measures be designed to minimize default, over-indebtedness, leakage, fraud and other risks to consumers?
#2 Is there heterogeneity in impacts along borrower characteristics (i.e. financial literacy, time preferences, income, and/or gender)?
#3 How can non-traditional credit-scoring algorithms, regulations, and other consumer protection measures be designed to minimize default, over-indebtedness, leakage, fraud and other risks to consumers?
Anchor Studies
Estimating the Effect of Digital Credit: An Experiment with Airtel Malawi
Lead Researcher: Jon Robinson (DCO Scientific Director, UC Santa Cruz)
Research Team: Pascaline Dupas
Country: Malawi
Partner: Airtel Malawi
This study aims to understand the impacts of Airtel's digital credit product Kutchova on borrowers, and test whether providing information on loan terms affects demand for digital loans.
Lead Researcher: Jon Robinson (DCO Scientific Director, UC Santa Cruz)
Research Team: Pascaline Dupas
Country: Malawi
Partner: Airtel Malawi
This study aims to understand the impacts of Airtel's digital credit product Kutchova on borrowers, and test whether providing information on loan terms affects demand for digital loans.
Building Better Credit Scoring Algorithms
Lead Researcher: Joshua Blumenstock (DCO Scientific Director, UC Berkeley)
Research Team: Daniel Bjorkegren
Country: Multiple
In this study, researchers are developing and testing new algorithms for constructing credit scores in developing countries. Primary focuses are on 1.) identifying borrowers who are most likely to benefit from a loan and 2.) designing algorithms that are "manipulation-proof" and transparent to borrowers.
Lead Researcher: Joshua Blumenstock (DCO Scientific Director, UC Berkeley)
Research Team: Daniel Bjorkegren
Country: Multiple
In this study, researchers are developing and testing new algorithms for constructing credit scores in developing countries. Primary focuses are on 1.) identifying borrowers who are most likely to benefit from a loan and 2.) designing algorithms that are "manipulation-proof" and transparent to borrowers.
Full-Scale Studies
Reducing Default and Improving Resilience on Digital Credit
Lead Researcher: Tavneet Suri (MIT)
Country: Kenya
Partner: Tala
This randomized control trial aims to test a new lending product that would give borrowers who are close to default the opportunity to extend their loan terms and, in fact, take out additional small amounts of capital to help them deal with any unexpected negative events that they may have experienced. The motivation is that borrowers who have experienced an unexpected negative event may need some extra time and capital in the short term that they can later repay fully. This study will test that hypothesis by working closely with the digital lender Tala to offer extended terms to select customers who default on their loans.
Lead Researcher: Tavneet Suri (MIT)
Country: Kenya
Partner: Tala
This randomized control trial aims to test a new lending product that would give borrowers who are close to default the opportunity to extend their loan terms and, in fact, take out additional small amounts of capital to help them deal with any unexpected negative events that they may have experienced. The motivation is that borrowers who have experienced an unexpected negative event may need some extra time and capital in the short term that they can later repay fully. This study will test that hypothesis by working closely with the digital lender Tala to offer extended terms to select customers who default on their loans.
Gender-Differentiated Credit Scoring Algorithms Using Call Detail Records and Machine Learning*
Lead Researcher: Sean Higgins (UC Berkeley)
Research Team: Paul Gertler and Laura Chioda
Country: Dominican Republic
Partner: ALNAP
Low-income women disproportionately lack access to credit, often because they lack credit histories, property rights, and formal earnings. This study aims to understand whether gender-differentiated credit scoring models using nontraditional data can increase women’s access to formal credit, whether women rejected by standard credit scoring models would benefit from credit access, and how these benefits compare to the benefits of credit access for women selected by standard models. In partnership with the largest telecommunications company and one of the largest banks in the Dominican Republic, the research team will conduct a study that combines novel algorithm development with an impact evaluation of credit allocated based on this novel algorithm.
*This study is a continuation of the pilot project funded in Spring 2017
Lead Researcher: Sean Higgins (UC Berkeley)
Research Team: Paul Gertler and Laura Chioda
Country: Dominican Republic
Partner: ALNAP
Low-income women disproportionately lack access to credit, often because they lack credit histories, property rights, and formal earnings. This study aims to understand whether gender-differentiated credit scoring models using nontraditional data can increase women’s access to formal credit, whether women rejected by standard credit scoring models would benefit from credit access, and how these benefits compare to the benefits of credit access for women selected by standard models. In partnership with the largest telecommunications company and one of the largest banks in the Dominican Republic, the research team will conduct a study that combines novel algorithm development with an impact evaluation of credit allocated based on this novel algorithm.
*This study is a continuation of the pilot project funded in Spring 2017
Mobile Financial Services and CDR-Based Credit Scores: A Gateway to Financial Inclusion for Unbanked Haitians?*
Lead Researcher: Travis Lybbert (UC Davis)
Research Team: Oscar Barriga-Cabanillas and Joshua Blumenstock (collaborator)
Country: Haiti
Partner: Digicel
Nano-loans may provide the working poor a valuable source of short-term credit, but simultaneously raise concerns about debt cycles and default. In collaboration with the main telecom provider, Digicel, this study aims to refine the algorithm used to predict credit worthiness of nano-loan borrowers and to understand impacts of nano-loans on these borrowers. Specifically, the research teams plans to estimate the effects of nano-loans on consumption patterns, short-term food security, debt portfolios and the total cost of debt. This study will also test how repayment schedules synchronized to predicted income cycles and simple financial literacy training shape these effects.
*This study is a continuation of the pilot project funded in Spring 2017
Lead Researcher: Travis Lybbert (UC Davis)
Research Team: Oscar Barriga-Cabanillas and Joshua Blumenstock (collaborator)
Country: Haiti
Partner: Digicel
Nano-loans may provide the working poor a valuable source of short-term credit, but simultaneously raise concerns about debt cycles and default. In collaboration with the main telecom provider, Digicel, this study aims to refine the algorithm used to predict credit worthiness of nano-loan borrowers and to understand impacts of nano-loans on these borrowers. Specifically, the research teams plans to estimate the effects of nano-loans on consumption patterns, short-term food security, debt portfolios and the total cost of debt. This study will also test how repayment schedules synchronized to predicted income cycles and simple financial literacy training shape these effects.
*This study is a continuation of the pilot project funded in Spring 2017
Digital Credit for Farmers in Ghana: A Randomized Field Trial
Lead Researcher: Christopher Udry (Northwestern)
Research Team: Monica Lambon-Quayefio
Country: Ghana
Partner: Farmerline
Digital technologies offer the potential to overcome many of the barriers that constrain credit provision in the developing world. This study will explore whether and how Paytime, an innovative digital credit system that provides loans to farmers via mobile phones, can increase investment and influence financial behavior, yields, and profits. In addition to estimating Paytime’s overall effectiveness at filling credit gaps, this study will explore how specific design features affect default rates, as well as investment patterns and other financial behaviors.
Lead Researcher: Christopher Udry (Northwestern)
Research Team: Monica Lambon-Quayefio
Country: Ghana
Partner: Farmerline
Digital technologies offer the potential to overcome many of the barriers that constrain credit provision in the developing world. This study will explore whether and how Paytime, an innovative digital credit system that provides loans to farmers via mobile phones, can increase investment and influence financial behavior, yields, and profits. In addition to estimating Paytime’s overall effectiveness at filling credit gaps, this study will explore how specific design features affect default rates, as well as investment patterns and other financial behaviors.
Slowing Down Digital Credit
Lead Researcher: Alfredo Burlando (University of Oregon)
Research Team: Silvia Prina and Mike Kuhn
Country: Mexico
Partner: EFL Global
Online banking and automated credit scoring technologies are making credit available at a faster rate, often within minutes of completing a loan application. This randomized experiment will test the impact of waiting periods on digital credit. This study will measure the causal impact of experiencing a delay on default rates, repayment rates and indebtedness, the selection effect of delays in credit disbursement, and the total effect of a policy enforcing delays.
Lead Researcher: Alfredo Burlando (University of Oregon)
Research Team: Silvia Prina and Mike Kuhn
Country: Mexico
Partner: EFL Global
Online banking and automated credit scoring technologies are making credit available at a faster rate, often within minutes of completing a loan application. This randomized experiment will test the impact of waiting periods on digital credit. This study will measure the causal impact of experiencing a delay on default rates, repayment rates and indebtedness, the selection effect of delays in credit disbursement, and the total effect of a policy enforcing delays.
Pilot Studies
Digital Credit for Agriculture
Lead Researcher: Sarah Janzen (Montana State University)
Research Team: Nicholas Magnan and Conner Mullally
Country: Kenya
Partner: Acre Africa and TransUnion
This study aims to understand the short and long-term impacts of digital credit tailored to the needs of smallholder Kenyan agricultural producers. Acre Africa and TransUnion will collaboratively design a non-traditional credit-scoring algorithm appropriate for assessing the risk of lending for small-scale agricultural production. To identify the impact of the improved credit score on loan uptake, the research team will randomly notify smallholder farmers of their eligibility to apply to receive a digital agricultural loan and encourage them to do so. This notification and encouragement will help identify the effects of receiving a digital loan.
Lead Researcher: Sarah Janzen (Montana State University)
Research Team: Nicholas Magnan and Conner Mullally
Country: Kenya
Partner: Acre Africa and TransUnion
This study aims to understand the short and long-term impacts of digital credit tailored to the needs of smallholder Kenyan agricultural producers. Acre Africa and TransUnion will collaboratively design a non-traditional credit-scoring algorithm appropriate for assessing the risk of lending for small-scale agricultural production. To identify the impact of the improved credit score on loan uptake, the research team will randomly notify smallholder farmers of their eligibility to apply to receive a digital agricultural loan and encourage them to do so. This notification and encouragement will help identify the effects of receiving a digital loan.
Milking it for all it's Worth: Digital Credit and Payments in the Dairy Sector
Lead Researcher: Alfredo Burlando (University of Oregon)
Research Team: Silvia Prina and Jessica Goldberg
Country: Uganda
Partner: MTN Uganda and MM4P
While access to digital credit and digital payments could be beneficial to farmers in rural areas, their actual impact is unknown. This randomized control trial will assess whether additional access to payments, aside from digital credit, boosts take-up of Mo-Kash (a digital credit product offered by MTN). In addition, the study aims to provide insights on the impacts of digital credit and digital payments on farmers' production, income, assets, investments, and consumption smoothing ability, as well as product prices.
Lead Researcher: Alfredo Burlando (University of Oregon)
Research Team: Silvia Prina and Jessica Goldberg
Country: Uganda
Partner: MTN Uganda and MM4P
While access to digital credit and digital payments could be beneficial to farmers in rural areas, their actual impact is unknown. This randomized control trial will assess whether additional access to payments, aside from digital credit, boosts take-up of Mo-Kash (a digital credit product offered by MTN). In addition, the study aims to provide insights on the impacts of digital credit and digital payments on farmers' production, income, assets, investments, and consumption smoothing ability, as well as product prices.
Digital Credit Linked to Digital Payments: Impact on Small Merchants in India*
Lead Researchers: Ethan Ligon (UC Berkeley) and Ketki Sheth (UC Merced)
Research Team: Carly Trachtman
Country: India
Partner: Catalyst
Following the 2016 demonetization in India, there has been a significant push by the government of India to promote the use of digital finance to reduce corruption and tax evasion. This has resulted in a promising landscape for the promotion of digital credit, thought to be particularly useful for small merchants who lack access to formal credit markets. One potential barrier to digital credit for small merchants is the required adoption of digital payment mechanisms that provide the transaction history required by lenders. This study will use a combination of observational data and experiment with small inducements to adoption to identify demand for digital devices and associate mobile payment schemes.
*This study is a continuation of the pilot project funded in Spring 2017
Lead Researchers: Ethan Ligon (UC Berkeley) and Ketki Sheth (UC Merced)
Research Team: Carly Trachtman
Country: India
Partner: Catalyst
Following the 2016 demonetization in India, there has been a significant push by the government of India to promote the use of digital finance to reduce corruption and tax evasion. This has resulted in a promising landscape for the promotion of digital credit, thought to be particularly useful for small merchants who lack access to formal credit markets. One potential barrier to digital credit for small merchants is the required adoption of digital payment mechanisms that provide the transaction history required by lenders. This study will use a combination of observational data and experiment with small inducements to adoption to identify demand for digital devices and associate mobile payment schemes.
*This study is a continuation of the pilot project funded in Spring 2017
Access to Digital Credit and its Spillover Effects in China
Lead Researcher: Zenan Wang (UC Berkeley)
Research Team: Elisabeth Sadoulet, Tao Sherry Kong, and Bingkai Chen
Country: China
Partner: Institute of Social Science Survey, Peking University
This pilot study will estimate both the impacts of receiving digital credit on loan-recipients and its spillovers on households who did not apply for loans. Additionally, experiments will be designed to further understand why credit constrained non-applicants opt out of applying for loans.
Lead Researcher: Zenan Wang (UC Berkeley)
Research Team: Elisabeth Sadoulet, Tao Sherry Kong, and Bingkai Chen
Country: China
Partner: Institute of Social Science Survey, Peking University
This pilot study will estimate both the impacts of receiving digital credit on loan-recipients and its spillovers on households who did not apply for loans. Additionally, experiments will be designed to further understand why credit constrained non-applicants opt out of applying for loans.
Consumer Protection Oversights in the Chinese Social Credit System
Lead Researcher: Shazeda Ahmed (UC Berkeley)
Research Team: Jael Makagon and Deirdre Mulligan
Country: China
In this qualitative pilot study, unbanked users of China's social credit system will be interviewed in order to understand their perceptions of the privacy, security, and other consumer protection risks they face in using such loosely regulated financial services.
Lead Researcher: Shazeda Ahmed (UC Berkeley)
Research Team: Jael Makagon and Deirdre Mulligan
Country: China
In this qualitative pilot study, unbanked users of China's social credit system will be interviewed in order to understand their perceptions of the privacy, security, and other consumer protection risks they face in using such loosely regulated financial services.