A0390
Title: A two-part income-based model for estimating underreported income in the presence of consumption measurement error
Authors: Maria Felice Arezzo - Sapienza University of Rome (Italy) [presenting]
Giuseppina Guagnano - Sapienza University of Rome (Italy)
Domenico Vitale - Sapienza University of Rome (Italy)
Roberta Di Stefano - Sapienza University of Rome (Italy)
Alberto Arcagni - Sapienza University of Roma (Italy)
Abstract: A novel two-part income-based model is introduced to estimate underreported income, addressing limitations of expenditure-based approaches in datasets with consumption measurement error, such as the Bank of Italy's SHIW. Traditional methods often require controlled data or assumptions about evading units, hindering broader applicability. While prior works use expenditure frameworks, they critically assume error-free consumption data, a condition often violated. The model circumvents this by directly incorporating measurement error in reported income. Part one establishes a preliminary classification of statistical units into potential "evader" and "non-evader" categories based on the income-consumption gap, explicitly acknowledging misclassification risk. Part two focuses on the provisional evader group, statistically estimating their expected true income to quantify underreporting, using advanced econometrics to mitigate misclassification and measurement error bias. A significant methodological advancement is offered for estimating income underreporting when comprehensive, error-free consumption data is unavailable, contributing to more robust empirical tools for economic measurement and informing fiscal policy.
