Financial Literacy and the Big Three

Financial Literacy and the Big Three

Financial literacy is the empirical proposition that households’ competence with basic financial concepts is far below the level the rational household-finance benchmark assumes. The standard measure is the Big Three survey instrument introduced by Lusardi and Mitchell ($2006$), three multiple-choice questions covering compound interest, real interest rates, and risk diversification.

The Big Three.

1. Compound interest. “Suppose you had \100$inasavingsaccountandtheinterestratewas$2%$peryear.After$5$years,howmuchdoyouthinkyouwouldhaveintheaccountifyouleftthemoneytogrow?"—Morethan$$102$/Exactly$$102$/Lessthan$$102$ / Don’t know.
2. Real interest rates. “Imagine that the interest rate on your savings account was $1\%$ per year and inflation was $2\%$ per year. After $1$ year, how much would you be able to buy with the money in this account?” — More than today / Exactly the same / Less than today / Don’t know.
3. Risk diversification. “Buying a single company stock usually provides a safer return than a stock mutual fund.” — True / False / Don’t know.

Headline result. Only $30\%$ of U.S. respondents answer all three correctly. The conditional rates are individually high (compound-interest correct for $\sim 65\%$, inflation correct for $\sim 64\%$, diversification correct for $\sim 52\%$), but the joint rate is only about $30\%$. Literacy correlates strongly with retirement-saving participation, portfolio diversification, and avoidance of high-fee products.

Cross-sectional pattern. Literacy is lower among women, younger respondents, lower-income respondents, and those without college education. The pattern has been replicated in over $20$ countries (van Rooij-Lusardi-Alessie $2011$ for the Netherlands; Bucher-Koenen-Lusardi $2011$ for Germany; Klapper-Lusardi-van Oudheusden $2015$ globally). The gender gap in particular is large and stable across countries.

Policy challenge. Direct financial education has produced disappointing field results. Skimmyhorn ($2012$) studies the U.S. Army’s mandatory financial-literacy training for new recruits and finds little long-run effect on financial outcomes. Cole and Shastry ($2012$) study U.S. state-level high-school personal-finance mandates and find little effect on adult financial behaviors. The empirical pattern motivates the shift toward nudges as a policy response — leaving literacy as it is and instead redesigning the choice environment.

Intuition

The rational household-finance benchmark — Merton’s life-cycle model, the standard portfolio-choice problem — assumes the household solves a complex multi-period optimization problem with stochastic returns and uncertain longevity. The Lusardi-Mitchell Big Three survey is, in comparison, almost trivial: each question can be answered with elementary arithmetic and a basic intuition for risk. The fact that fewer than a third of respondents answer all three correctly is a stark indictment of the assumption that households operate at the level of the rational benchmark.

Each of the three questions targets a foundational concept. Compound interest is the building block of saving; without it, the household cannot evaluate a retirement projection. Real interest rates are the building block of valuation under inflation; without them, the household cannot tell whether a “high yield” savings account is actually preserving purchasing power. Diversification is the building block of portfolio choice; without it, the household holds undiversified positions in employer stock or single-company concentrations. The questions are not chosen as exotic challenges; they are chosen because each answers a question the household must know to make any reasonable financial decision.

The disappointing results from financial-education interventions reflect the difficulty of moving these foundational concepts from “Don’t know” to “Correct” in adult populations through structured curricula. Even where curricula succeed in raising survey-test scores, the gains often fail to translate into changed financial behavior. The interpretation in the prescriptive-finance literature is that the binding constraint may not be the absence of information but the cost of attention, the immediate temptation of consumption, and the cognitive burden of optimization — features that more education does not directly address.

Properties

(Joint Correct Rate Is $30\%$) Most U.S. adults cannot answer all three Big Three questions.

Lusardi and Mitchell ($2006$) administer the Big Three to a representative U.S. sample. The joint correct rate — answering all three correctly — is $30\%$. The marginal rates are higher: compound interest at $\sim 65\%$, inflation at $\sim 64\%$, diversification at $\sim 52\%$. The gap between the marginal and joint rates means that those who get any one question right are not strongly more likely to get the others right; literacy is unevenly distributed across the three concepts.

(Cross-Sectional Gradient) Literacy is lower among women, young respondents, and those with less education.

The gender gap in particular is large and stable: women in the U.S. sample are roughly $10$ percentage points less likely to answer all three correctly than men, and the gap appears in nearly every country tested. Klapper-Lusardi-van Oudheusden ($2015$) survey $140$+ countries with the same instrument and confirm the gender gap on every continent. Possible explanations include differences in financial-decision-making roles within households, differences in confidence and willingness to attempt the question, and differences in financial-product exposure. The gap survives controls for income, education, and age.

(Predicts Financial Outcomes) Higher literacy correlates with better retirement saving, more diversification, lower fees.

van Rooij-Lusardi-Alessie ($2011$) find that high-literacy Dutch households are more likely to participate in the stock market, to hold diversified portfolios, and to plan for retirement. The associations are economically substantial — a one-standard-deviation increase in literacy is associated with a $20$-percentage-point increase in stock-market participation — and survive controls for income, wealth, and education. Similar correlations have been documented in U.S., German, Italian, and Japanese samples.

(Skimmyhorn 2012) Military financial education has small long-run effects.

The U.S. Army began administering mandatory $8$-hour financial-literacy training to all new enlisted soldiers in $2007$. Skimmyhorn ($2012$) compares pre- and post-mandate cohorts using personnel-finance administrative data. He finds short-run effects on retirement-plan participation but small or zero effects on durable measures (savings balances, debt management, credit utilization) several years later. The estimated improvements are roughly an order of magnitude smaller than the effects from auto-enrollment defaults.

(Cole-Shastry 2012) High-school personal-finance mandates have negligible effects.

A number of U.S. states required personal-finance instruction in high school in the $1980$s and $1990$s. Cole and Shastry ($2012$) use cross-state and cross-cohort variation to measure the long-run effect on adult financial behaviors (savings, asset accumulation, credit). They find essentially zero effect — the coefficients are statistically indistinguishable from zero and economically negligible. The result mirrors Skimmyhorn’s military finding and casts doubt on direct education as a policy lever.

(Why Education Fails) The binding constraint may be attention and self-control, not information.

The standard interpretation of these null results in the prescriptive-finance literature is that the obstacle to good financial decisions is not the absence of information per se but the cognitive cost of attention, the temptation to consume now rather than save, and the computational difficulty of multi-period optimization (see hyperbolic discounting and bounded rationality). A literate but inattentive or present-biased household will still under-save. This shifts the policy lever from teaching the chooser toward redesigning the choice environment.

Examples

The compound-interest question

“Suppose you had \100$inasavingsaccountandtheinterestratewas$2%$peryear.After$5$ years, how much do you think you would have if you left the money to grow?”

The correct answer is “More than \102$"—specifically,$$100 \times 1.02^5 \approx $110.41$.Thewrong-answerpatternisinformative:respondentswhoanswer"Exactly$$102$"arecomputingsimpleinterest($$100 + 5 \times $2 = $110$wouldbethecorrectsimple-interestanswer,soevensimpleinterestisbeingmis-appliedas$$2$peryear\ast foroneyearonly\ast ).Respondentswhoanswer"Lessthan$$102$"areconfusedaboutthedirection.About$35%$ of the U.S. sample fall into one of the wrong-answer categories.

Why diversification confuses respondents

The diversification question states “Buying a single company stock usually provides a safer return than a stock mutual fund.” The correct answer is false — a single stock has both market risk and idiosyncratic risk; a fund has only market risk. But the question’s phrasing exploits a common mistake: respondents who interpret “safer” as “more familiar” or “more controllable” select true. The framing reveals a deeper confusion between familiarity and risk — a single well-known stock feels safer than a basket of unfamiliar names — that drives the home-bias and own-employer-stock anomalies in 8.10.2. Under-Diversification.

An $8$-hour course versus a one-time default change

Skimmyhorn ($2012$) shows that an $8$-hour mandatory financial-literacy course in the U.S. Army raises retirement-plan participation by about $5$ percentage points relative to baseline. Madrian-Shea ($2001$) show that switching a single firm’s $401(k)$ from opt-in to auto-enrollment raises participation by approximately $40$ percentage points. The two interventions are not directly comparable — different populations, different time horizons — but the orders-of-magnitude difference in effect size is striking. The implication that the prescriptive-finance literature draws is that changing the chooser (literacy training) is harder than changing the choice environment (defaults), and policy should focus on the latter.

See Also

• 8.12. Prescriptive Behavioral Finance — parent hub.
• 8.12.2. Financial Advice and Audit Studies — sibling: why market provision of advice does not solve the literacy gap.
• 8.12.3. Nudges and Default Options — sibling: the policy lever that has worked where literacy training has not.
• 8.10. Behavioral Household Finance — household-finance puzzles motivating the prescriptive agenda.
• 8.10.1. Stock Market Non-Participation — the empirical puzzle most directly tied to literacy.
• 8.10.2. Under-Diversification — anomaly driven in part by the diversification question’s underlying confusion.