Tax bunching at the kink in the presence of low capacity of enforcement : evidence from Uruguay

Bérgolo, Marcelo - Burdín, Gabriel - De Rosa, Mauricio - Giaccobasso, Matías - Leites, Martín

Resumen:

A first-order policy issue in low and middle income countries is how to design optimal tax systems in order to improve the state’s potential of supporting economic development. Although information regarding behavioral responses to taxation is a key input for tax design, the evidence in developing contexts is still scarce. In this paper we contribute to fill this gap by exploring in detail how individual taxpayers respond to personal income taxation in Uruguay. To do this, we rely on rich administrative tax records covering the universe of Uruguayan taxpayers and implement a bunching design. First, we find a moderate implied elasticity of taxable income (0.16) in the first kink point of the tax schedule. Second, we investigate the mechanisms driving these responses extensively. We find that the observed responses are a combination of both gross labor income and deductions responses. In particular, we document a more intensive use of personal deductions for taxpayers close to the kink point, and suggestive evidence of evasion responses through unilateral and employer-employee collusion labor income misreporting. Our results suggest that policy efforts should be directed at broadening the tax base and improving the enforcement capacities of tax authorities rather than eroding tax progressivity.


Una cuestión de política de primer orden en países de ingresos bajos y medios es cómo diseñar sistemas impositivos óptimos para mejorar la capacidad del estado de apoyar el desarrollo económico. Aunque la información acerca de qué tanto y cómo los individuos responden a los impuestos es importante para el diseño de política, la evidencia para países en desarrollo es escasa. En este documento, contribuimos a llenar esta brecha al explorar cómo los contribuyentes individuales responden a los impuestos personales a las rentas del trabajo en Uruguay (IRPF). Para hacer esto, nuestro diseño de investigación utiliza registros administrativos de impuestos que cubren el universo de los contribuyentes de Uruguay, y se basa en un enfoque de bunching. Primero, encontramos una elasticidad del ingreso reportado moderada (0.16) en el primer kink del esquema de IRPF (i.e., mínimo no imponible). En segundo lugar, investigamos en detalles los mecanismos que explican estas respuestas. Encontramos que las respuestas observadas son una combinación de ajustes, tanto en el ingreso laboral bruto como en las deducciones. En particular, documentamos un uso más intensivo de las deducciones personales para los contribuyentes cercanos al kink, y evidencia sugerente de respuestas de evasión a través de la sub-declaración de ingresos laborales, tanto de manera unilateral como a partir de mecanismos de colusión entre empleadores y empleados. Nuestros resultados sugieren que las políticas deben dirigirse a ampliar la base impositiva y mejorar las capacidades de enforcement de las autoridades fiscales, en lugar de erosionar la progresividad del esquema de IRPF.


Detalles Bibliográficos
2019
Personal income taxation
Tax bunching
Elasticity of labor income
Deductions behavior
Misreporting
Developing economies
Brunching
Impuestos a la renta personal
Elasticidad del ingreso reportado
Deducciones
Sub declaración de ingresos
INGRESOS
DEDUCCIONES FISCALES
PAISES EN DESARROLLO
Inglés
Universidad de la República
COLIBRI
http://hdl.handle.net/20.500.12008/20441
Acceso abierto
Resumen:
Sumario:A first-order policy issue in low and middle income countries is how to design optimal tax systems in order to improve the state’s potential of supporting economic development. Although information regarding behavioral responses to taxation is a key input for tax design, the evidence in developing contexts is still scarce. In this paper we contribute to fill this gap by exploring in detail how individual taxpayers respond to personal income taxation in Uruguay. To do this, we rely on rich administrative tax records covering the universe of Uruguayan taxpayers and implement a bunching design. First, we find a moderate implied elasticity of taxable income (0.16) in the first kink point of the tax schedule. Second, we investigate the mechanisms driving these responses extensively. We find that the observed responses are a combination of both gross labor income and deductions responses. In particular, we document a more intensive use of personal deductions for taxpayers close to the kink point, and suggestive evidence of evasion responses through unilateral and employer-employee collusion labor income misreporting. Our results suggest that policy efforts should be directed at broadening the tax base and improving the enforcement capacities of tax authorities rather than eroding tax progressivity.