Household Finance Essays

The core of the study is a randomized field experiment conducted with a large credit card company in Brazil. This involved a substantial sample size of 19,690 clients. The experiment cleverly manipulated three key aspects of the credit card offers: the offered interest rate (ranging from 3.99% to 11.89% per month), the degree of interest rate disclosure (prominent versus hidden in a footnote), and the featured payment plan (while all clients were offered four plans, one was highlighted in the advertisement). This multifaceted design allowed researchers to assess the independent and combined effects of these factors on credit demand. The use of randomization is crucial for establishing causal relationships between the manipulated variables and consumer responses.

The experimental results revealed a surprising level of consumer sophistication in this Brazilian population. Most borrowers demonstrated significant rate sensitivity—their borrowing decisions were influenced by the interest rate—regardless of whether the interest rate information was prominently disclosed or buried in the fine print. This challenges the prevailing assumption that consumers consistently overlook crucial details like interest rates, particularly when presented with low monthly payment figures. The study indicates that consumers are surprisingly adept at decoding information, even when lenders attempt to obfuscate it through marketing techniques.

While the average effect of prominent interest rate disclosure on take-up rates was small and statistically insignificant, a crucial subpopulation showed a different response: high-risk borrowers. This group exhibited rate sensitivity only when the interest rate was prominently displayed; they were insensitive to the rate when it was concealed. This finding suggests that high-risk borrowers, likely less attentive to contract details, are significantly more influenced by the salience of information. However, even for this group, the study found that the financial consequences of improved information disclosure were modest. This nuanced finding reveals a critical vulnerability of a specific segment within the consumer credit market.

The study also examined the impact of offering various payment plans and the manipulation of 'nudges' (highlighting a particular plan in the advertising). Borrowers exhibited responsiveness to nudges favoring longer-term plans. However, this sensitivity decreased when the financial stakes were higher. This suggests that while consumers can be influenced by subtle cues, these influences are mitigated when the cost of a decision becomes more significant. Furthermore, the company's attempt to attract more customers to payment plans by prominently displaying options with lower monthly payments (and thus longer maturities) was unsuccessful, further supporting the conclusion of sophisticated consumer behavior in assessing the information.

In conclusion, the research suggests that lenders' attempts to exploit limited consumer attention through advertising strategies that conceal interest rates and emphasize low monthly payments are largely ineffective, at least on average. Brazilian credit card holders display notable sensitivity to interest rates even when the information is not prominently presented. Although high-risk borrowers represent a notable exception, with their decision-making significantly impacted by the visibility of interest rates, the overall impact of improved disclosure remains modest. The study underscores the surprising level of consumer sophistication and the limitations of relying solely on obfuscation to influence consumer credit choices in this market.

This chapter delves into the impact of peer effects on financial decisions. The authors distinguish between two key mechanisms driving these effects: social learning and social utility. Social learning refers to the influence on an individual's decision-making based on observing the choices of their peers. This implies that individuals learn from the observed actions of others, potentially inferring information about the quality or desirability of a financial product. In contrast, social utility describes a situation where a peer's possession of an asset directly affects the utility or satisfaction an individual derives from owning the same asset. This might stem from factors like conformity, keeping up with the Joneses, or the desire for joint consumption. The study aims to disentangle these two effects to understand their relative contributions to overall peer influence on financial choices.

To isolate the effects of social learning and social utility, the researchers employed a field experiment with a financial brokerage. The innovative aspect of their design lies in the use of a lottery to determine whether individuals who chose to purchase an asset actually acquired it. This manipulation allowed them to decouple the effect of a peer's decision to purchase (revealing information) from the effect of a peer's possession of the asset. By randomly informing one member of a peer pair about the other's investment decision and then employing this lottery mechanism to determine possession, the study was able to assess the independent impact of information (social learning) and the effect of the peer's actual possession of the asset (social utility) on the other peer's investment decisions. This experimental setup provided a cleaner measure of each channel's influence compared to previous research.

Recognizing potential heterogeneity in investor behavior, the researchers explored how financial sophistication influences the impact of peer effects. They hypothesized that more financially sophisticated investors would be less susceptible to social learning, relying more on their own assessments of asset quality. Conversely, less sophisticated investors were expected to be more heavily influenced by their peers' choices. Similarly, financially sophisticated investors might exert a stronger influence on their peers due to their better knowledge and assessment of financial products. Investigating these variations allowed the study to test for a more nuanced understanding of social learning and its interaction with individual differences in financial expertise, furthering the study's scope beyond a simple average effect.

The authors acknowledge potential alternative explanations for their findings and take steps to address them. They consider the possibility of side payments between peers, particularly within family units, influencing the results. Analysis indicates that such effects are not significantly impacting the study's conclusions. The authors also address concerns that knowing a peer's desire to purchase an asset might give an indication of that peer’s portfolio or future investment intentions, potentially confounding social learning effects. The researchers address this by noting that the effects of social learning are minimal for clients with technical occupations which counters this hypothesis. Finally, the study acknowledges that it focuses solely on a specific type of social learning and excludes other potentially relevant forms which exist in financial contexts, representing a limitation of the research.

This section examines the common practice of 'borrowing high and lending low' (BHLL) behavior in Brazil. The study finds that a substantial portion of clients, approximately 70%, maintain an overdrawn checking account for at least one day each year while simultaneously holding liquid assets in savings accounts. This behavior is particularly noteworthy given the significantly high interest rate spread in Brazil, estimated to be around 150% annually. Despite this large spread, the overall financial cost incurred by consumers due to this behavior is relatively modest, averaging less than 0.5% of their annual earnings, suggesting that while the behavior exists, its financial impact is limited for many individuals. The researchers then explore potential rationales behind this seeming contradiction.

The study explores potential rational explanations for why consumers might engage in BHLL behavior, even when the cost of doing so is relatively high. One reason could be concern about the bank reducing their overdraft limit if their debt is paid off, decreasing available liquidity. This is a relevant consideration as overdraft terms are renewed monthly, and banks have the authority to adjust limits. Another potential rationale involves strategic default. Since savings accounts aren't used as collateral for overdraft debt, consumers could strategically default on the overdraft while retaining access to their savings. The authors contrast this with other situations where perfect substitutes for means of payment may not exist and conclude that transferring funds from savings accounts to cover overdrafts in this specific setting reduces the chance of incurring fees and higher interest rates.

To investigate the effect of interest rates on BHLL behavior, the study compares the practices of bank clients and bank employees. Bank employees benefit from a significantly lower overdraft interest rate (approximately 4% per month) compared to clients (8.33% per month). While bank employees, on average, have higher incomes and credit limits, which could influence their financial choices, the authors control for these factors in their analysis to mitigate potential biases. This comparison allows for a more direct examination of how differences in interest rates influence the prevalence and cost of BHLL behavior. Notably, even with lower interest rates, bank employees exhibited higher prevalence and cost of BHLL behavior than clients, although the cost is still relatively low compared to annual income.

The study's results indicate that BHLL behavior is prevalent even in a high-interest-rate environment such as Brazil, where the spread between borrowing and saving rates is substantially higher than in the US. Around 70% of clients engage in BHLL behavior, incurring avoidable charges averaging approximately R$100 (US$60) annually. This cost represents less than 0.5% of their average yearly earnings. While bank employees faced significantly lower overdraft rates, their BHLL costs and frequency were still substantially higher compared to clients. This suggests that high interest rates, while acting as a deterrent, are not sufficient to completely eliminate or reduce the incidence of such behavior. The study's findings demonstrate that while the behavior is present, its overall financial consequences are relatively limited, even when considering the significant interest rate disparity.