This post is the second in a series. To read the first post, click here.
This post is the second in a series summarizing recent BPI research that uses a new module of the Consumer Financial Protection Bureau’s (CFPB) Making Ends Meet Survey (MEMS) to study fraud and scams from the consumer perspective.[1] The previous post discussed the incidence, types and reporting propensity of fraud and scams. This post focuses on the part of the research that relates consumers’ characteristics to their likelihood of experiencing fraud or a scam, their financial exposure or loss and their propensity to report an incident.
Several lessons emerge. First, higher-income individuals are less likely to experience fraud or a scam. Higher-income individuals who do are less likely to be substantially exposed financially and to report the incident. In contrast, income volatility, which has been tied to financial fragility, is associated with more victimization and greater financial exposure or loss.[2] Age and education are less strongly associated with the likelihood of experiencing fraud, greater financial exposure or reporting.
Second, personal traits beyond observable demographics, like patience and propensity for risky financial behavior, are salient risk factors. Consumers who are less patient and those who gamble are significantly more likely to experience fraud. Moreover, while less patience is linked to lower reporting propensity, gambling is associated with a higher likelihood of reporting an incident.
Third, lower levels of financial well-being are associated with a higher risk of experiencing fraud and greater financial exposure or loss from it. Consumers who are “just getting by,” who are “afraid [their] money won’t last” or who feel they “will never have the things [they] want” are at significantly higher risk of experiencing fraud and more likely to suffer substantial financial exposure or loss when they do.
These findings are consistent with a receptivity channel: financial dissatisfaction may increase susceptibility to fraud and scams promising financial improvement. They also reveal systematic differences in consumers’ propensity to report an incident. This may be of concern to policymakers who rely on reporting to assess how fraud and scams affect different types of people.
Income, Age and Education as Determinants of Fraud Risk
The research summarized in this post uses Wave 6 of the MEMS to tie consumers’ likelihood of experiencing fraud or a scam, financial exposure and reporting to their demographic and personal characteristics.[3]
Figure 1: Fraud, Age, Income and Education
Figure1 plots how consumers’ income, education and age relate to: (a) the likelihood consumers experience fraud, (b) the likelihood fraud or a scam is carried out using a non-card bank product, (c) whether it led to significant exposure or loss[4] and (d) the likelihood it was reported. The bars in the figure show the estimated (relative) relationship.[5] The black whiskers report the 95 percent confidence interval of that estimate.[6]
As shown in the top panel of the figure, higher income is associated with a lower likelihood of experiencing fraud or a scam. As compared to consumers earning less than $100k per year, those earning between $100k and $250k are nearly 10 percentage points less likely to experience fraud or a scam. Those earning over $250k are nearly 20 percentage points less likely. Higher-income individuals are also less financially exposed, and, though not statistically significant, less likely to report their experiences.
Income variability is associated with higher victimization risk.[7] Consumers whose monthly income varied over the past year are about five percentage points more likely to experience fraud or a scam, including those involving a non-card bank product. Unlike income levels, variability does not systematically predict more financial exposure, nor does it associate with significantly different reporting propensity.
Age and education do not clearly predict fraud or scam experiences overall. However, the data suggest college educated consumers are less likely to experience fraud or a scam that uses a non-card bank product, like a certified check or wire transfer, as opposed to a credit or debit card. Further, relative to consumers under 50 years old, those between 50 and 70 are less likely to report fraud or a scam.
Beyond Demographics: Time Preferences and Risk Taking
The MEMS collects respondent information that goes beyond traditionally observable demographics. Two traits critical for financial decision making are: (1) patience, or time discounting, and (2) appetite for risk. The MEMS captures the former using a well-established elicitation method.[8] The latter is captured by asking consumers if they participated in gambling activities over the past year, whether online, in person at a casino or both.
Using a multivariate regression framework, the research examines whether these preference traits contain additional explanatory power beyond, or conditional on, the demographic characteristics discussed above, i.e., income, age and education. The top panel of Figure 2 plots these associations for time preferences. The bottom of the figure plots them for gambling behavior.
Figure 2: Beyond Demographics: Fraud, Patience and Risk Preferences
As shown in Figure 2, consumers with low patience, or those who heavily discount the future, are significantly more likely to experience fraud. They are also less likely to report it, though this estimate is not statistically significant. Gambling is another risk factor. Consumers who have both gambled online and visited a casino are over 10 percentage points more likely to experience fraud or a scam. However, unlike low-patience consumers, they are significantly more likely to report an incident. In all, these traits systematically predict both risk of victimization and visibility in data based on reports.
Fraud, Perceptions and the Financial Well-Being Scale
A central objective of the MEMS is to construct the financial well-being scale, a survey-based measure of consumers’ subjective financial security. The scale is built from questions covering four main concepts: control over routine finances, capacity to absorb financial shocks, progress toward financial goals and financial freedom of choice, i.e., the flexibility to make life-improving decisions.[9]
Figure 3 relates fraud and scam outcomes to questions that form this well-being scale. Like the analysis of consumer preferences (Figure 2), it is based on a multivariate analysis and plots the additional explanatory power of four key financial well-being questions, conditional on observed demographics. Specifically, respondents are asked how well the statements apply to them.[10]
The top panel of Figure 3 relates fraud outcomes to consumers’ confidence in making financial decisions. Relative to consumers who are not confident making financial decisions, those with complete confidence are over 10 percentage points less likely to experience fraud or a scam. They are also more likely to report the most recent incident.
The bottom panels of Figure 3 tie fraud and scam outcomes to consumers’ perceptions of capacity to absorb financial shocks, progress towards financial goals and financial freedom. These perceptions are collected by asking respondents whether they agree with each statement listed in the figure. The statements are negatively framed, whereby agreement indicates lower financial well-being. Across these measures, the results show consistent patterns.
Figure 3: Fraud, Attitudes and Perceptions
From the figure, those who agree that they are “just getting by” are significantly more likely to experience fraud. Among those who experience fraud or a scam, individuals who believe they are “just getting by” also experience more substantial financial exposure or loss. The same patterns emerge for individuals who “are concerned the money [they] have or save won’t last” and who feel that they “will never have the things [they] want.” Moreover, those who believe their money will not last or that they will never have the things they want are more likely to report an incident. In other words, the data indicate that lower levels of financial well-being are tied to a greater risk of fraud and loss, but also a higher propensity to report an incident.
Overall, analysis of the MEMS data supports a receptivity mechanism: financially stressed individuals may be more susceptible to fraud and scams promising improvement in their financial situation. This is consistent with the idea that psychological factors can significantly amplify vulnerability.[11]
Conclusion
Analysis of the MEMS suggests that consumers’ demographic and personality characteristics are important determinants of their experiences with financial fraud and scams. Individuals with lower or more variable income are more likely to experience fraud, as are those with less patience, those who have gambled in the previous year, or those who are less satisfied with their financial situation.
Especially salient in the research is the analysis of consumer traits that are not readily observed but are nonetheless crucial for assessing why and what types of consumers may be vulnerable to fraud and scams. Analysis of these traits suggests a clear and intuitive mechanism: financially stressed individuals are more susceptible to fraud and scams promising improvement in their financial situation.
The findings also suggest systematic differences in consumers’ propensity to report their experiences, often based on unobserved traits, which can be relevant for policymakers who rely on these reports to learn about how fraud and scams affect different parts of the U.S. population.
[1] This research, titled Frauds and Scams: A Post-Pandemic Consumer Perspective,can be found here: Frauds and Scams: A Post-Pandemic Consumer Perspective by Daniel Grodzicki :: SSRN.
[2] For a discussion of income volatility and financial fragility, see, for example: https://cdi.mecon.gob.ar/bases/doc/oecd/sem/125.pdf.
[3] The MEMS is a nationally representative survey of adults with a credit record. It was administered in the first quarter of 2025.
[4] A fraud or scam is determined to have led to significant exposure if monetary loss prior to recoveries exceeds $100 and some portion of the amount exposed is not recovered.
[5] Each likelihood is estimated relative to a base group, which is reported in the figure. For example, the likelihood of experiencing fraud by income is estimated relative to a base group comprised of consumers earning less than $100k per year.
[6] Confidence intervals are based on robust standard errors. See research paper for additional information.
[7] Specifically, survey respondents are asked the following question: Which best describes your household’s income from month to month: (1) income is about the same each month, (2) income varies somewhat from month to month, (3) income varies a lot from month to month. Income variability in this analysis takes on a value of one if respondents said their income varies somewhat or a lot. It takes on a value of zero otherwise.
[8] This is based on the method of Coller and Williams (1999). The paper can be found here: https://www.cambridge.org/core/journals/experimental-economics/article/abs/eliciting-individual-discount-rates/F7D1E72F9BB2E17A9033CE2011EF41D5/. In the MEMS, respondents are asked a series of three questions: (1) prefer $1,000 in one month or $1,050 in six months, (2) prefer $1,000 in one month or $1,100 in six months, (3) prefer $1,000 in one month or $1,150 in six months. Consistent preferences imply a nesting of these questions not imposed in the questionnaire. For example, one cannot prefer $1,050 in six months to $1,000 in one month and $1,000 in one month to $1,150 in six months (i.e., be willing to wait five months for an additional $50 but not for an additional $150). About 97 percent of responses are consistent. For the analysis plotted in Figure 2, mid-patience takes on a value of one if the respondent prefers $1,100 in six months to $1,000 in one month but not $1,050 in six months to $1,000 in one month, and zero otherwise. Low patience takes on a value of one if the respondent prefers $1,150 or more in six months to $1,000 in one month, but not $1,100 in six months to $1,000 in one month, and zero otherwise.
[9] For more details regarding the construction and interpretation of this scale, see https://www.cambridge.org/core/journals/journal-of-financial-literacy-and-wellbeing/article/consumer-financial-wellbeing-does-scale-choice-alter-the-measure/19B02EF4573599FDE02794F3327B66B3.
[10] These questions are framed as: “How well does this statement describe you or your situation?” Respondents answer on the following Likert scale: (1) not at all, (2) very little, (3) somewhat, (4) very well, (5) completely. All these questions are asked together and early in the survey, prior to those on fraud.
[11] A more comprehensive discussion of this mechanism can be found here: https://www.tandfonline.com/doi/abs/10.1080/07317115.2012.749323.
