October 13: Financial Inclusion & Consumer Protection

Abstract: Digital credit has helped drive financial inclusion in Kenya over the past decade. However, the expansion of credit access to the newly financially included may pose consumer risks such as multiple borrowing, taking overlapping loans from multiple providers, and credit default. To study these issues, I analyze a novel dataset of digital credit transactions in the Kenyan market collected as part of the Digital Credit Market Inquiry undertaken by the Competition Authority of Kenya in collaboration with IPA. A unique de-identification approach allows me to match consumers pseudonymously across providers, allowing for the study of multiple borrowing behavior and default behavior among these consumers. I match consumers across four providers, finding that about 6% of consumers at these four providers have accounts at multiple, and most of those with multiple accounts also multiple borrow. I analyze late repayment and multiple borrowing for a key provider, to understand how multiple borrowing externalities of two other providers drive their repayment outcomes. I find that while those who multiple borrow have lower default and late repayment rates per loan, that they eventually default at a higher rate. The latter result matches results from IPA’s survey of digital financial service users, where I find that those who have ever multiple borrowed are 14 percentage points more likely to have ever defaulted on a loan. Furthermore, it is consistent with a story where borrowers keep themselves afloat via sequences of loans. I am in process of additional analysis of the data. In particular, this key provider raised their fees midway through my sample, changing consumer incentives for multiple borrowing and repayment. I plan to use event study methods to understand how this price hike impacted both the composition and behavior of borrowers.

Abstract: By relieving individuals from crippling debt, bankruptcy has been shown to increase homeownership, annual earnings, employment, and promote the creation of new businesses. In addition to improving filers’ future financial outcomes, bankruptcy may also improve psychological wellbeing by relieving financial stress. However, despite the benefits that insolvency offers, many low-income debtors lack access to bankruptcy due to costs, technological barriers, and stigma. We conduct, to our knowledge, the first Randomized Controlled Trial (RCT) to examine whether lowering the financial barriers to insolvency improves access to this important financial institution for low-income debtors. We will use our RCT to achieve three aims: 1) Estimate a demand curve for bankruptcy and use this curve to estimate the effects of bankruptcy reforms on economic welfare, 2) determine the effects of low-cost bankruptcy on long-term credit and financial outcomes, 3) test whether bankruptcy alleviates the psychological effects of poverty on mental health, life satisfaction, and moment-to-moment experience and happiness. In partnership with a licensed insolvency trustee (LIT) in Canada, we randomly subsidize potential filers with $1,000 toward bankruptcy or consumer proposal filing fees. After piloting an online version of the experiment, we are switching to an in-person design, where we subsidize a subset of the approximately 500 debtors who use our LIT’s debt services per quarter. A separate treatment arm will offer debtors a surprise subsidy after they have decided to file so that we can separately identify income effects from the economic effect of lowering the financial barrier to access bankruptcy. We also use moment-to-moment experience sampling and sub-clinical questionnaires to track participants’ stream of consciousness and mental health throughout the filing process, to determine whether debt relief improves wellbeing by relieving financial stress. 

Abstract: The effectiveness of loan collection process is at the core of stability and efficiency of credit markets. Yet, most research has focused on loan origination, such as screening of borrowers and on types of credit contracts. Apart from work on bankruptcy resolution in courts, little research has been dedicated to the determinants of loan recovery. In low- and middle-income countries, where credit markets are characterized by weak credit enforcements and limited consumer protection, the ability of loan collectors to recover delinquent loans carries important implications. However, little is known about the value added by the loan collectors, and whether improving the assignment of collectors to late borrowers can enhance the debt recovery process.

In this paper, we make first progress on these questions. We partner with a large bank of consumer credit in Mexico and develop a measure to estimate the ability of loan collectors. We uncover large dispersion in loan collectors’ ability, both within and across types of loan products.  Next, we will document the determinants of these productivity differences, both within and across markets.  Finally, we will use our estimated measures and results to quantify gains achievable through reallocation of late borrowers to more efficient loan collectors within and across types of loan products. We will compare these estimates to the current allocation as well as to the random allocation.

Abstract: Section 304 of the Credit Card Accountability, Responsibility, and Disclosure Act (2009) limited the marketing and sale of credit cards to college students, nudging them away from these high-rate products. While it reduced card use, we document the nudge raised student loan balances by 8.4%, 15% among the less affluent. To assess the benefits of this substitution, we design a survey that reveals prevalent sub-

optimal financial decision-making among students tied to higher card debt. Model- based evidence demonstrates how these results imply the policy raised welfare. A complementary analysis indicates higher grade-point averages and on-time graduation rates resulting from the policy.

Abstract: We evaluate the trade-offs of hard paternalistic financial regulation designed to reduce indebtedness. We do so exploiting a series of government policies tightening credit card minimum payments in Quebec that left minimum payments unchanged in the rest of Canada. We show how lenders respond to this regulation. We estimate the causal effects of these policies on consumers by using a difference-in-differences methodology and consumer credit reporting data covering all Canadian credit cards.

Abstract: We seek to understand switching behavior by studying the market for digital financial services (DFS) in Tanzania -- particularly mobile money -- where users rely on agents to complete transactions such as withdrawing cash from their digital wallets. These agents tend to be concentrated geographically and offer the same basic services, giving consumers the choice of which agent to use. Cross-country evidence from mystery shopping shows there is significant variation in fees and reliability (probability that the transaction will be successfully completed) indicating that consumers do not always use the lowest cost or the most reliable agent. This inaction suggests that other factors may be more important to consumers such as social ties with the agent, or that consumers are unaware of cost and reliability.

Abstract: Digital payment technologies like mobile money are highly popular worldwide. However, their use is generally confined to remittances, with individuals not taking advantage of these services for payments or saving. One potential explanation for this is the need to learn about how to use these services in this manner and build habits. This suggests that helping women start to use mobile money services in new ways could permanently shift how they use these services, and experience greater benefits from them. We test this using a randomized control trial in which 750 female entrepreneurs with microfinance loans in Tanzania were randomly switched to repay their loan using mobile money instead of the usual method of cash. We find that this exogeneous shift in the way that women use mobile money for loan repayment increases their use of mobile money for saving and business payments. Women gain increased control of their finances, experience less pressure to share money with others and have higher levels of empowerment in the household. They save more, and are more able to meet their business goals, though do not experience any changes in their business outcomes. These findings highlight the benefits that encouraging greater use of digital technologies can bring to women.

Abstract: One of the most important technological advances over the last quarter-century has been the global diffusion of mobile phones. Mobile technology enables improved communication capabilities, access to information, utilization of digital financial services, and adoption of internet-based applications. We study the causal effects of the uptake of this multi-faceted technology leveraging an RCT among married women non-phone owners in Malawi (n=1500). Participants randomly assigned to the smartphone conditions were provided handsets, SIM cards, and mobile phone and empowerment training; the phone distribution and training were delivered either to the participants individually (Individuals; n=400) or with the participation of their spouses (Couples; n=400). We compare the effects of smartphones to an unconditional cash transfer ($70), the equivalent value of the handset (Cash; n=400), and a Control (n=300). After nine months, we find that the Individuals and Couples conditions significantly increased women’s mobile phone ownership, mobile connectivity, use of digital financial services, and financial inclusion (as measured behaviorally). Cash transfer recipients tended not to invest in mobile technology, but instead used their grants to support micro-enterprise and also become more active in village banking. However, this produced more consistent and robust economic benefits than the smartphone conditions. Compared to Control, members of the Cash group had significantly higher individual savings, household consumption, and receipt of loans—potentially supporting their micro-enterprise. The smartphone effects were more mixed and varied by training regimen: whereas Couples experienced significant increases in household consumption compared to Control, Individuals realized larger gains in weekly income. Overall, this represents one of the first analyses of the differential impact of financial capital versus mobile technology on household economic growth.   

Abstract: The positive effect of bank account ownership on women’s labor force participation, occupational mobility and savings is well studied in the literature but there is limited evidence on how it impacts women's control over money and its use at home. A bank account serves as an entry point in several countries for access to formal savings and credit and is therefore expected to positively impact her agency in decisions around use of household and personal resources. However, the availability of banking services and costs to access them are complementary determinants of women's decision making ability linked with account ownership. This paper exploits an exogenous account expansion policy in the districts of India to test the effect of household's account ownership on women’s self-reported participation in decisions on resource allocation and autonomous use of cash. I find that faster growth in bank accounts in the absence of supporting banking infrastructure within a district does not translate into improved outcomes for women.  To determine the context in which account ownership improves women's bargaining power, this paper tests individual and district-level measures of bank account use, interacted with cost-saving features of financial inclusion such as distance from bank branch, on her outcomes. It then explores changes in households' consumption and investment patterns as well as women's time use in response to women's bank account ownership using a household panel data. As economies are rapidly improving financial technology and adopting digital finance to reduce poverty, it is critical to understand which tools serve the vulnerable and socially disadvantaged groups. 

Abstract:  At the core of many consumer lending decisions is a credit score: an algorithmic assessment of a customer’s creditworthiness. Traditional credit scores are designed to maximize lender profits, and use machine learning algorithms to predict which customers will repay loans. This paper proposes and tests a different paradigm for consumer lending, in which ‘welfare-sensitive’ credit scores allow the lender to balance expected profits against the expected welfare impacts of specific loans. Using data from a randomized control trial in Nigeria, we show how machine learning algorithms can be trained to predict the welfare impact of lending to a client, and how those welfare scores can be combined with traditional credit scores to characterize a Pareto-efficient tradeoff between welfare and profits. Our main result suggests that, in the Nigerian context, the lender could achieve a welfare-neutral lending operation by sacrificing just 2.5% of profits.

Abstract: The digitization of financial services has enabled tremendous innovation in the provision of credit in low- and middle-income countries (LMICs), which some hail as a transformative development with potential to drive financial inclusion, reduce poverty, and spur economic growth. However, others associate digital credit with a proliferation of misconduct, consumer abuses, and over-indebtedness, which can have severe consequences for the most vulnerable consumers and amplify inequality. The Center for Effective Global Action and Innovations for Poverty Action have produced a new report rejecting both the worst case fears of critics and the positive potential held out by proponents. Rather, the evidence on digital credit thus far points to modest improvements in resilience and subjective wellbeing for the average consumer, but which fall short of being transformative. While studies thus far have not found that recipients are being harmed by digital loans, the proliferation of digital credit creates urgency for addressing consumer protection challenges, and much remains unknown. Despite this, there are reasons for optimism. Digitization has catalyzed rapid growth in financial inclusion, and addressing policy issues now can be instructive for informing management and regulation of digital credit products more broadly, including the potential movement of digital credit into more productive parts of the economy in the horizon.

Abstract: Digital credit that uses non-traditional scoring techniques has already expanded credit access to many new populations.  A point of policy debate is whether digital lenders should be fully integrated into information sharing systems, such as those offered by  credit reference bureaus (CRBs). We study an example of a digital lender in Mexico adopting credit bureau scores into their screening process.  Using unique administrative data, we estimate a regression discontinuity in time around the lender's integration of credit bureau scores. We find that the acquisition of credit scores reduces defaults, with the likelihood of borrowers' repayment increasing by 10-13.