High School Curriculum and Financial Outcomes: The Impact of Mandated Personal Finance and Mathematics Courses

Three large datasets — the 2000 US Census, the Federal Reserve Bank of New York Consumer Credit Panel/Equifax, and the Survey of Income and Program Participation; state-cohort variation from personal finance mandates enacted 1957-1982 and mathematics graduation requirements enacted 1984-1994

US adults who attended high school under (or just before) a state curriculum mandate; credit outcomes measured in adulthood, well beyond the young-adult window of the mandate literature.

The paper that broke the founding result of the financial-education-mandate literature and replaced it with an uncomfortable substitute. Identification compares individuals in the same state who graduated just before a mandate took effect with those who graduated just after, with state fixed effects — and the state fixed effects are the whole story: Bernheim, Garrett & Maki's (2001) finding that personal-finance mandates raised saving is, in these authors' words, not robust to their simple inclusion. Two limitations are stated by the authors themselves. The personal-finance mandates studied (1957-1982) often did not comprise a course requirement at all but brought in a bank representative for a one-off lecture, so this is a weak treatment; and the maths result is confounded to the extent that maths reforms travelled with other graduation-requirement changes, which the authors control for in aggregate but cannot separate by subject for want of power.

State mandates requiring high-school students to take personal finance courses had NO effect on investment or credit-management outcomes — no effect on the probability of reporting any investment income, the level of investment income, credit score, credit-card delinquency, or the probability of bankruptcy or foreclosure, and no detectable effect on total financial assets or real-estate equity. Bernheim, Garrett & Maki's contrary 2001 result does not survive the inclusion of state fixed effects. Requiring an ADDITIONAL MATHEMATICS course, by contrast, raises home equity by $1,500-$3,000 against a base of $15,500 and cuts the probability of foreclosure by 0.4 percentage points against a base of 9%, and these effects survive controlling flexibly for earned income, educational attainment and occupation — so they are not simply the labour-market return to maths.

Low. Cohort-level exposure to a state curriculum mandate is not chosen by the family, and state fixed effects absorb time-invariant state characteristics. This is precisely the control whose absence, the authors argue, produced the earlier positive result.

This is a failed replication, and an unusually clean one: the same policy variation, a larger dataset, and the addition of state fixed effects turns Bernheim, Garrett & Maki (2001) from positive to null. It in turn stands in tension with Urban et al. (2020) and Brown et al. (2016), who study post-2000 mandates on younger cohorts and find credit-score and delinquency effects. Cole et al. name that tension and offer two reconciliations — the newer mandates are real course requirements while many older ones were a single guest lecture, and their own sample is older (22-28) so effects may dissipate with age. Neither reconciliation has been tested. Any reader who wants to believe the mandate literature has to pick one of them.