Estimation and Simulation of RURO Labor Supply Models with Administrative Data: Re-assessing the Evidence from Belgium

This paper estimates a Random Utility Random Opportunity model of labor supply using linked Belgian administrative data . The framework allows individuals to choose among stochastic wage and hours offers, capturing both participation decisions and hours adjustments within a unified structure. By combining tax records, social security data, and demographic registers, we construct precise measures of earnings, hours, and household characteristics.

A detailed micro simulator tailored to the Belgian tax system maps gross income into disposable income, ensuring an exact representation of institutional rules and nonlinear budget constraints. Compared to survey based data, administrative records substantially reduce measurement error and improve the credibility of simulated behavioral responses.

The model closely replicates observed distributions of hours, wages, and income across gender and household types. We then revisit an in-work benefit reform previously analyzed using survey data, highlighting how data precision and institutional detail affect predicted labor supply responses and budgetary implications.