Misclassification of the dependent variable in a debt-repayment behavior context

This paper estimates a model of the household debt-repayment decision that accounts for the possibility of misclassification of self-reported debt-repayment status. It likewise estimates the extent of misclassification in a sample of data from different European countries. The evidence suggests that many households that are in arrears do not report this condition, so that the true level of arrears is, on average, 24% higher than that observed in our data. Furthermore, the effects on the incidence of arrears of adverse income and expense shocks are substantially greater than those predicted by estimators that ignore the possibility of misclassification.