Impacts of a Credit Comparison Tool on Costs and Consumer Welfare

Context

Although research has shown that prospective borrowers could benefit from seeking offers from multiple lenders, particularly in markets where prices can vary widely, high search costs often prevent consumers from comparing offers in practice.

In Mexico, household debt nearly doubled from 2000 to 2019, largely driven by increases in consumer debt. This trend is particularly strong among low income consumers, who have reported being less likely than richer individuals to shop between different credit products.

Since 2020, the regulatory division of Mexico’s central bank has been developing a new consumer credit application and comparison platform that will enable consumers to apply for credit once through the platform and have their application evaluated by multiple banks, rather than having to separately apply to each bank. After each bank decides whether to offer credit to the consumer and what terms to offer, the consumer will be able to compare the individualized credit offers from each bank through this platform. Importantly, the offers will be binding as long as the information that the consumer provided is accurate.

Study Design

Using administrative data from Banco de México, this project seeks to determine whether reducing search costs has an effect on the credit card users choose to contract, leading them to access cheaper credit. Researchers will also identify whether banks internalize that consumers are now searching more than in the absence of the platform, and adjust the offered credit card terms, including the APR and non-monetary characteristics.

The first set of interventions are targeted towards users of the platform. Researchers will measure the accuracy of people’s predictions about future card usage. To the extent that some borrowers are not accurate predictors, the project will focus on understanding who mispredicts, and why. In addition, researchers will investigate the impact of the platform on credit terms and consumer welfare. Theoretically, lower search costs should enable more comparison and thus access to lower-cost cards, but these assumptions will be tested using detailed administrative data from Banxico, including the channel borrowers used to sign up for a card.

The project will also examine whether competition affects the terms offered to consumers, including APR and non-monetary characteristics. Because banks’ offers to clients are binding, but they do not know how many other banks are extending offers or what the terms are, they are incentivized to offer competitive rates. Researchers can then use variation in the number of banks receiving a user’s application to tease apart the effects of competition from the quality of applicants.

Results & Policy Lessons

Banco de México has faced challenges getting financial institutions to opt into the cost comparison platform. Only four of thirteen eligible financial institutions have agreed to join the platform; Banco de México has agreed to launch the platform once a fifth institution joins. This has delayed the project, alongside negotiating access to Banco de México’s administrative records, which includes information on credit card users’ contracts such as interest rate, annual fee and credit limit, as well as data on their monthly balances and repayment. The project collected survey data from respondents recruited through Google Ads and has 3,809 respondents. Preliminary data suggests financial literacy improves with experience: the average number of correct answers was 4.6 with a median of 5 out of 8 and those who owned a credit card, particularly those who had owned one for a longer time, got more questions right. The team will compare survey respondents’ predictions about credit use and costs to administrative records to provide insights on why consumers wrongly predict credit use and costs. This will inform the design interventions that address these.