Política fiscal e crime ambiental no Brasil: o impacto das renúncias fiscais na dinâmica dos ilícitos ambientais

Abstract

Tax incentives are widely recognized as public policy instruments to promote desirable business practices, such as sustainability. However, coherence in managing these benefits presents a significant challenge. Therefore, this study aimed to investigate the relationship between tax incentives granted to Brazilian companies and the occurrence of environmental crimes, analyzing this phenomenon through the lens of Rational Choice Theory, taking the number of infractions and the value of infraction notices as negative proxies for socio-environmental performance. For this purpose, a quantitative approach was used, employing regression analysis on a sample of 61 Brazilian companies registered in the Transparency Portal and the IBAMA platform, with data from 2024. The statistical analysis utilized linear regression models, specifically Ordinary Least Squares (OLS) and, subsequently, the Generalized Least Squares (GLS) model, to estimate and ensure the efficiency of the results. The estimations evidenced a positive and statistically significant correlation between the total value of tax waivers and the occurrence of environmental crimes. This finding suggests a fragility in the implementation of public policies through tax waivers, where the State may be indirectly financing the persistence of illegal activities. Therefore, for public management, this research suggests considering the proposal of a Bill that enhances the combat against environmental impunity. It is recommended that such a legal structure establish the automated integration of data between oversight bodies, such as IBAMA, and platforms for transparency and benefit concession. This would allow for the establishment of a mechanism for the automatic suspension or exclusion of infringing companies from accessing tax incentives, thereby ensuring that public resources do not subsidize, even indirectly, the perpetuation of environmental crimes.