Does It Actually Work?

The case against: financial education explains almost nothing

The most influential paper in this literature is a 2014 meta-analysis in Management Science by Daniel Fernandes, John Lynch and Richard Netemeyer, covering 168 papers and 201 prior studies. Its findings are blunt:[1]

• Interventions to improve financial literacy explained about 0.1% of the variance in the financial behaviors studied.
• Effects were weaker still in low-income samples.
• Effects decay. Even large interventions with many hours of instruction had negligible effects on behavior twenty months or more after the intervention.

That last point is the one that should worry anyone proposing to teach mortgages to a fourth grader. The gap between eighth grade and a first credit card is not twenty months; it is closer to a hundred. If knowledge decays on the timescale Fernandes and colleagues measured, front-loading instruction into childhood is exactly the wrong design.

The paper also drew a distinction that is frequently lost in citation: correlational studies measuring financial literacy find much stronger associations with behavior than experimental studies of financial education find. Financially literate people do behave differently. It is far less clear that teaching produces them.

The case for: state mandates changed measurable outcomes

A separate body of work asks a different question. Instead of measuring classroom interventions, it exploits the fact that individual states imposed graduation requirements in particular years, and compares young adults who were in school under a requirement to those in control states who were not. These are quasi-experimental designs measuring the policy, not the lesson.

Credit scores and delinquency

Brown, Collins, Schmeiser and Urban studied Georgia, Idaho and Texas, where mandates took effect in 2007. Young people in school after implementation had higher relative credit scores and lower relative delinquency rates than their counterparts in control states.[2]

The follow-up work by Urban, Schmeiser, Collins and Brown, published in Economics of Education Review, found requirements associated with fewer defaults and higher credit scores — while stressing something advocates usually omit: the aggregate result masks substantial state-level heterogeneity. Some state mandates worked; others did not. The authors identify well-funded teacher preparation as central to which was which.[3]

Payday lending

Melody Harvey, using pooled 2012 and 2015 National Financial Capability Study data, found that state financial education mandates significantly reduced both the likelihood and the frequency of payday borrowing among young adults.[4] Payday loans commonly carry annualized rates around 300%; avoidance is a large, concrete financial gain concentrated among lower-income borrowers — the same group for whom Fernandes and colleagues found the weakest effects.

Student loans

Daniel Mangrum, in the Journal of Financial Economics, found that students subject to high school personal finance mandates showed improved federal student loan repayment. The distributional detail matters: students with higher-income parents responded by borrowing less, reducing median balances by about 7%, while first-generation and low-income borrowers did not significantly change how much they borrowed but were more likely to pay down balances. Mandates that paired career research with personal finance produced better repayment than personal finance alone.[5]

Stoddard and Urban, in the Journal of Money, Credit and Banking, found that requirements increased financial aid applications and acceptance of grants and subsidized federal loans, while decreasing the likelihood of carrying credit card balances — a shift from high-cost to low-cost financing.[6]

Reconciling the two literatures

These findings are not as contradictory as they appear. They differ on three axes.

1. Unit of analysis. Fernandes and colleagues aggregate mostly short-duration interventions, many of them brief, voluntary, or adult-directed. The mandate studies measure a full graduation-required course delivered by trained teachers as part of a state policy. These are different treatments.
2. Outcome measured. The meta-analysis pools a wide range of self-reported behaviors. The mandate studies use administrative credit and loan records — credit scores, delinquency, payday borrowing, repayment. Harder outcomes, less measurement noise.
3. Timing relative to the decision.A high school senior taking a required course is roughly a year from student loans and a credit card. That is close to what Kaiser and Menkhoff call a “teachable moment.”

The finding that matters most here: intensity

Two later meta-analyses by Tim Kaiser and Lukas Menkhoff address the design question directly, and they are the most decision-relevant evidence for North Carolina.

The 2017 analysis in the World Bank Economic Review, covering 126 impact evaluations, found that financial education does significantly affect financial behavior and, more strongly, financial literacy — and that success “depends crucially on increasing education intensity” and on delivery at a teachable moment.[7]

The 2020 analysis in Economics of Education Review restricts attention to school-based experiments — the setting North Carolina is legislating for. Its metaregressions show a beneficial effect of more intensive treatments and of smaller class sizes, with decreasing marginal returns.[8]

Three findings that cut against this site’s own argument

Presenting only supportive evidence would make this page worthless to anyone evaluating the policy seriously. Three qualifications deserve to be stated as plainly as the supportive results.

Mandates may be less effective than voluntary programs

Kaiser and Menkhoff’s 2017 analysis reports that mandatory financial education tentatively appears to be less effective, and that debt-related behaviors are among the hardest to influence.[7] This is an uncomfortable finding for a site arguing for a mandate. It is also, on reflection, an argument about implementation qualityrather than about mandates as such: a requirement delivered by an untrained teacher to a disengaged class is a weak treatment regardless of what compelled it. That reading is consistent with Urban and colleagues finding that teacher preparation separates the mandates that worked from the ones that didn’t.[3] It should still temper any expectation that passing a bill is the same as producing an outcome.

Effects decay, and K-8 is a long way from the decision

The decay finding[1] applies with more force to second graders than to high school seniors. A serious K-8 proposal should not claim that teaching credit scores in middle school will change borrowing at 22. The defensible claim is narrower: early grades build the numeracy, saving habits, and vocabulary that a later, closer-to-decision course can build on — a sequencing argument, not a substitution argument.

Almost none of the causal evidence is about K-8

This is the most important limitation on this page. Brown, Harvey, Mangrum, Stoddard and Urban all study high school graduation requirements. They establish that a dedicated, assessed course close to a financial decision changes measurable behavior. They do not establish that a dedicated requirement for at least one grade before ninth would. The intensity finding[8] comes from school-based studies spanning grade levels, which is the closest available evidence, but it is not direct proof for the elementary grades.

The honest formulation is therefore: the evidence strongly supports dedicated and assessed over embedded and unmeasured, and supports proximity to the financial decision. It does not by itself prove that K-8 is the optimal place to spend new instructional time. What it does establish is that if North Carolina intends to keep its statutory promise of instruction for all students, the current K-8 design is the version of that promise the evidence gives us least reason to expect anything from — and that no one is currently checking whether it happens at all.