The puzzle the authors confront is a familiar one in household finance. Decades of survey evidence show that people with higher measured financial literacy tend to hold more stocks, diversify more widely, and accumulate more wealth. But literacy is not assigned at random. It correlates with education, income, parental background, cognitive ability, and a host of unobserved traits such as patience, confidence, and tolerance for complexity. Any of these factors could simultaneously drive both the acquisition of financial knowledge and the choice to hold risky assets, producing an apparent relationship that reflects underlying differences between people rather than a genuine effect of knowledge on behavior. Standard regression estimates, in other words, may overstate, understate, or even misdirect the true influence of financial literacy on portfolio decisions.

To break this statistical knot, the researchers use parental education as an instrument for financial literacy. The logic rests on intergenerational transmission: children acquire financial knowledge in part through growing up in households where parents attained higher levels of education, an influence that operates largely outside the child’s own control. At the same time, parental education should have no direct effect on an adult child’s current portfolio beyond the channels that run through the child’s own education, income, and financial knowledge, once those factors are held constant in the analysis. This assumption allows the portion of financial literacy that is explained by family background to serve as a kind of natural experiment, revealing how literacy shifts investment behavior when it moves for reasons unrelated to the individual’s current economic circumstances. The approach builds on a long line of research documenting that human capital and financial behaviors pass from parents to children, including work by Chiteji and Stafford on the intergenerational transmission of portfolio choices and by Grohmann and colleagues on the childhood roots of financial literacy.

With the instrument in hand, the authors estimate instrumental variables probit models, a technique suited to binary outcomes such as whether a household owns a particular asset class. The dependent variables span a broad set of financial products: direct holdings of stocks, mutual funds, individual retirement accounts, and simpler instruments such as savings accounts and certificates of deposit. The probit framework models the probability of ownership as a function of financial literacy and an extensive set of controls, while the instrumental variables correction addresses the endogeneity of literacy in the participation equation. Because the Survey of Consumer Finances deliberately oversamples wealthy households through a list sample constructed from administrative tax data, the survey can capture behavior at the upper tail of the wealth distribution that standard household surveys substantially miss, lending the analysis unusual coverage of the American financial landscape.

The headline result is a sharp asymmetry across asset classes. Financial literacy significantly increases the probability that a household participates in complex, growth-oriented investments—individual retirement accounts, mutual funds, and stocks—but shows no significant effect on ownership of simpler financial products such as basic transaction accounts. This pattern is theoretically coherent. Simple products require little knowledge to purchase; nearly anyone can open a savings account. Complex products, by contrast, demand an understanding of risk, return, diversification, fees, and tax treatment. When the costs of comprehension are high, knowledge becomes the binding constraint on participation. The findings align with earlier contributions by van Rooij, Lusardi, and Alessie, who documented a strong relationship between sophisticated financial knowledge and stock market participation, and with work by Calvet, Campbell, and Sodini showing that financially sophisticated households make fewer costly investment mistakes.

Perhaps the most striking finding, however, concerns what the study did not find. The data reveal large demographic differences in both financial literacy and market participation—well-documented gaps by sex, income level, and race. Yet when the authors tested whether the effect of literacy itself differed across these groups, they found no evidence for heterogeneous effects. Once a given level of financial literacy is acquired, its estimated benefit in terms of investment participation appears broadly similar for women and men, for households across the income distribution, and across racial groups. The implication is subtle but consequential: demographic disparities in participation stem primarily from disparities in the acquisition of knowledge and in resources, not from knowledge being intrinsically more useful to some groups than others. A unit of financial understanding, once obtained, translates into roughly comparable behavioral change regardless of who holds it.

This homogeneity result carries significant weight for policy debates. If the returns to financial literacy differed sharply across demographic groups, targeted interventions would be the natural response, with education programs tailored to specific populations. The findings instead point toward broad-based interventions: because the marginal benefit of literacy is similar across groups, expanding financial education universally would be expected to narrow participation gaps wherever those gaps originate in differential knowledge acquisition. The result also reframes the interpretation of documented group differences. Prior research, including Bucher-Koenen and colleagues’ international comparison of gender gaps in financial literacy and Barber and Odean’s classic study of gender and common stock investment, has emphasized differences in knowledge levels and confidence. The new evidence suggests the deeper inequity lies not in how knowledge pays off, but in who has the opportunity to acquire it.

The study enters a literature that has matured considerably since Lusardi and Mitchell’s foundational work established financial literacy as a distinct field of economic inquiry. Their subsequent theoretical contributions, including the analysis of optimal financial knowledge and wealth inequality with Michaud, modeled literacy as a form of human capital that households choose to acquire, with compounding consequences for lifetime wealth. Empirical work by Jappelli and Padula linked investment in literacy to portfolio choice, while Behrman, Mitchell, Soo, and Bravo used Chilean data and an instrumental variables approach similar in spirit to document effects on wealth accumulation. The new American evidence complements this body of work by providing causal estimates for a comprehensive set of asset categories in a single, richly measured dataset, and by explicitly probing whether the causal effect varies across demographic lines—a question earlier studies largely left open.

The timing of the analysis is notable. The 2022 Survey of Consumer Finances captures households after a period of extraordinary market turbulence, rapid inflation, and the broad democratization of trading through commission-free brokerage apps. Against this backdrop, the persistence of a strong literacy gradient in participation underscores how structural the knowledge barrier remains, even as the transaction costs of investing have fallen toward zero. The paper’s notes on measurement also illuminate the care taken in operationalizing key concepts: risk tolerance in the survey is scored on a scale from zero to ten, with respondents classified as risk averse, risk neutral, or risk loving depending on where they fall, and self-declared financial knowledge is banded into low, medium, and high categories using parallel thresholds. These classifications allow the authors to distinguish genuine knowledge effects from mere differences in risk appetite, a confound that has complicated earlier analyses.

For a field increasingly concerned with consumer financial regulation, the findings speak directly to questions raised by Campbell and others about how policy should treat households that struggle with complex financial environments. If literacy causally unlocks participation in retirement accounts and equity markets—the very instruments through which long-run wealth is built in most advanced economies—then failures of financial education translate directly into foregone wealth accumulation and deepened inequality. The authors’ evidence that the benefits of knowledge are distributed evenly once acquired offers a measure of optimism: the barrier to broader and more equitable market participation appears to be an educational one, and educational barriers, unlike many other determinants of financial fortune, can be lowered by deliberate policy. The study’s results will likely inform ongoing efforts in the United States and elsewhere to mandate financial education in schools and to design default architecture in retirement systems, adding rigorous causal evidence to a debate that has long relied on correlational findings.