Modeling Financial Literacy Using Multilevel Item Response Theory and the COVID-19 Pandemic

In this paper, we analyze the financial literacy of account holders of a large Brazilian financial institution. Using a sample of 2,234 respondents, we estimated the latent characteristic of financial literacy using a Bayesian multilevel item response model. Our results suggest that financial literacy is intrinsically related to individual differences in Brazil, particularly regarding gender, political views, schooling, the type of organization the individual works in, and the region they live in. Also, people with low financial literacy were the ones who found it more challenging to manage their financial resources during the COVID-19 pandemic. Our results suggest that public policies to increase financial literacy should also include basic economic and political knowledge