by Giovanna Paladino; CeRP WP 209/22
Abstract. This paper takes its cue from the relevance of the framing effect related to behavioural biases associated with economic decision-making. Most attempts to measure financial literacy rely on surveys that include standardized questions about the knowledge of three or four fundamental concepts. A survey conducted in October 2021 that involved 2500 individuals representative of the Italian population made it possible to evaluate whether questions with different wording created higher respondent engagement, determined other answers and improved performance in terms of financial literacy. The descriptive and regression analysis showed that the wording mattered in three out of four questions. More engaging wording mitigated the gender effect by reducing the probability of women choosing the ‘I do not know’ option. However, while there was evidence of an increase in the percentage of correct answers in single questions, the overall level of financial literacy showed no signs of improvement. The regression analysis found that the likelihood of being financially literate, independently of the type of question, depends on sociodemographic variables (gender, age, geographical area and level of education) and on self-evaluation of digital and economic skills. In addition, knowledge of basic maths plays a key role. Whoever knows how to compute a percentage correctly has a notably higher probability of being financially literate. This evidence has clear policy indications.
Published: March 2022