A few months ago I did a quick review of the literature on financial literacy training to find out whether there was any indication that those programs have been successful.  I was struck by the lack of evidence.  Out of all the studies I could find only a couple stood out as being remotely credible.  So I took notice when one of those studies, the best of the lot, was put under new scrutiny by Shawn Cole in a presentation last month at a World Bank conference on financial literacy.

The original study, conducted in 2001 by Douglas Bernheim, Daniel Garrett, and Dean Maki, measured the impact of a mandatory, state-sponsored, high school financial course on subsequent household savings levels. The study concluded that individuals in states where financial education is mandatory have higher savings rates than residents in states with no mandatory financial education requirement.

But establishing causality is often a tricky matter.  A 2007 paper by Shawn Cole and Gauri Kartini Shastry reopened the question that Bernheim et al. seemed to have answered. Cole and Shastry use a larger pool of data, allowing for more finely tuned analysis, to answer questions about the role of education, financial literacy and cognitive ability on financial market participation. Their study shows that the amount of general education received corresponds positively with an individual's investment income, but financial literacy training does not seem to contribute to that effect. Cole and Shastry were able to attribute the increased savings seen in Bernheim et al's study to strong GDP growth in participating states prior to the curriculum implementation. In short, it was economic growth, not financial literacy, that was responsible for the increase in savings that Bernheim et al found.

Cole and Shastry are not ready to entirely dismiss the importance of financial education, though. They suggest that the length of time spent in financial courses may have a positive impact on financial participation, as may the math skill level of program participants. Financial literacy likely also helps people make better decisions with the money they do have, whether those decisions involve formal financial channels or not.

Cole's presentation reminded me it can be dangerous to put too much faith in a study just because it's the best we've got.  The Bernheim et al. study was perfectly well executed, but this is exactly why randomized trials are so important - to eliminate the unknowns.  Of course, no one study is conclusive, but if financial literacy doesn't necessarily lead to financial participation, what other options for driving increased use of formal financial services are there? One possibility is simple incentives -- the equivalent of the classic, "Open a bank account and get a free toaster!" promotion. Cole, in another study conducted with the World Bank's Bilal Zia, this time in Indonesia, tested the impact of both financial literacy training and cash incentives on participants' propensity to open bank accounts. He found that financial literacy training has no impact but that offering financial incentives does. Perhaps marrying financial literacy training, to promote better decision making, with financial incentives, to promote use of services, will result in more of the world's poor building assets through savings and integrating into formal financial networks.