Financial literacy: different indicator, different insights?

Despite questions surrounding their reliability, much of the financial literacy literature continues to rely on simple metrics – without checking whether using an alternative might yield different findings. The present paper does just that: we replicate two prior studies and substitute the Standard & Poor’s indicator originally used with the Big Three and OECD/INFE metrics. The results are disconcerting. In the first study, many of the relationships between national culture and country-level financial literacy become weakly significant or disappear altogether. The second study is also of a cross-country nature but analyses the impact of financial literacy on financial inclusion. Here, the magnitude of the coefficient on financial literacy in the OLS regressions decreases substantially once we use alternative metrics. In one case, the coefficient drops by no less than 69 per cent with the OECD/INFE metric, and by between one-third and half with the Big Three. Even more concerning, the Instrumental Variable regressions show no causal relationship anymore. These findings are a strong signal that the literature would benefit from revisiting several key papers, especially those relying on cross-country data.While research employing short financial literacy tests has expanded almost exponentially over recent decades, studies questioning whether these tests are overly simplistic remain relatively scarce. Our work contributes to this emerging literature. Specifically, we replicate two existing studies—one examining the antecedents of financial literacy and another its consequences—and replace the original literacy indicators with metrics that are supposed to be similar. We find that many results either vanish entirely or weaken considerably. We hope this paper inspires further scrutiny of financial literacy tests.