Financial literacy and resilience when survey respondents prefer guessing to admitting ignorance

The respondent’s propensity to guess randomly even though it is possible to answer "I do not know" is a plausible source of measurement error in assessments of financial literacy. Using data from two consecutive waves of a survey on financial literacy and resilience of Italian households during the Covid-19 pandemic we model and estimate the respondents’ propensity to guess rather than admit ignorance and find that it implies sizable probabilities of misclassification for standard financial literacy indicators. Classifying as financially literate only respondents who answer correctly in both waves of the survey yields more significant plausible coefficients of financial literacy as an explanatory variable for financial resilience: truly literate individuals have a lower probability of having difficulty to make ends meet at the end of the month and of not being able to face a mid-size emergency expense; individuals who guessed and were lucky enough to appear literate show lower financial resilience.