Curriculum·G701 How to Teach This Material·about 33 min
What teaching this material has to overcome
By the end of this lesson you can
- →State what the best available evidence says financial education changes, and by how much, with the figure and its source
- →Distinguish knowledge that can be recited from behavior that changed, and name the measurement that tells them apart
- →Compute what a 0.1 percent share of variance means as a correlation, and what a 20-month decay means for a course schedule
- →Identify the four features of this Academy's design that exist because of this evidence, and say which one each answers
Graduate · enrolled learners
This lesson opens with Financial Literacy, Financial Education, and Downstream Financial Behaviors, 2014.
- What happened
- Daniel Fernandes, John Lynch and Richard Netemeyer published a meta-analysis in Management Science in 2014 covering 168 papers and 201 prior studies of financial literacy and financial education. Interventions to improve financial literacy explained 0.1 percent of the variance in the financial behaviors those studies measured, with weaker effects in low-income samples. The effect of financial education decayed with time: even large interventions with many hours of instruction had negligible effects on behavior 20 months or more after delivery. Measured financial literacy correlated with good outcomes, but in three surveys of their own, when the authors controlled for traits omitted from prior research (propensity to plan, confidence in financial information search, willingness to take financial risks and numeracy), the partial effect of literacy diminished dramatically and for most behaviors was no longer significant. The authors say their cross-sectional data cannot settle which way causation runs; one reading is that literacy was a marker of traits that also produced the behavior.
- The decision point
- The field had read the correlational studies, in which people who knew more behaved better, as evidence that teaching people more would make them behave better. When the studies that actually delivered an education and then measured behavior were pooled, the effect was close to zero. Nothing about the finding says people cannot learn; it says that a course which ends at the quiz has not been shown to change a decision made twenty months later. Every design choice in this Academy that looks unusual, from the autopsy at the top of every course to the hand-computed lab at the end of it, is an answer to this paper.
What you will be able to answer
- →What did the 2014 financial education meta-analysis find?
- →Why did literacy correlate with outcomes but not cause them?
- →What is the difference between the recitation number and the behavior number?
- →What does 0.1 percent of variance mean as a correlation?
Orientation and Year One are open: anyone can read them without an account. From Year Two onward the lessons are for enrolled learners, because progress through the later years only means anything if it is tracked against a record.
It is free. We do not sell the list and there is nothing to buy at the end of it.
Sources and review
- https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2333898
- https://ideas.repec.org/a/inm/ormnsc/v60y2014i8p1861-1883.html
- https://www.nber.org/system/files/working_papers/w27057/w27057.pdf
- https://www.nefe.org/_images/research/Effect-of-Financial-Literacy-on-Financial-Behaviors/Effect-of-Financial-Education-on-Financial-Behaviors-Final-Report.pdf
Confidence high·Volatility low·Reviewed 2026-09-14·Owner unassigned
Contested
A later meta-analysis, Kaiser, Lusardi, Menkhoff and Urban (Journal of Financial Economics, 2022; NBER working paper 2020), pooled 76 randomized experiments with over 160,000 people and found positive effects on knowledge and behavior at least three times the 2014 average. It found no difference for low-income samples, found some decay after six months but no further significant decline, and attributes the 2014 decay result to a meta-regression on only 29 observations, while saying it cannot rule out near-zero effects at long horizons either. The two papers disagree, and the gap is largely a difference of sample (76 randomized trials against 13 in 2014) and statistical model. Reading the 2014 paper as a finding about the field as practiced, and the later one as what better-run programs can do, is this lesson's reading, not either paper's. Neither supports a course that ends at the quiz.
The traits that took the weight off literacy come from the authors' own three cross-sectional surveys, not from the meta-analysis, and the authors state those data are equally consistent with several causal orderings.
This lesson does not claim that people cannot learn to handle money or keys. It claims that the usual way of teaching has not been shown to change decisions made later, and that the burden is on the teacher to measure the decision rather than the recitation.