Cuadernos de Administración
versão impressa ISSN 0120-3592
MENDEZ SAYAGO, Jhon Alexander. Adapting genetic algorithms to simulate the strategic behavior of contaminating agent companies regarding retribution tax charges. Cuad. Adm. [online]. 2008, vol.21, n.35, pp. 161-187. ISSN 0120-3592.
olombian environmental legislation has resorted to an economic instrument called retribution tax -based on the mechanism proposed by Baumol and Oates in their 1973 paper-, to regulate the contamination of bodies of water. This article employs standard genetic algorithms to simulate the behavior of a set of artificial companies that spill contaminants into bodies of water, when they are faced with a uniform tax charge with the same characteristics as the retribution tax. The simulations are aimed at reproducing company behavior in distinct tax application scenarios, for the purpose of evaluating their efficiency and proposing measures aimed at improvement. Such assessment is needed because the contaminating companies have adopted a strategic behavior that Baumol and Oates did not contemplate. It is due to the companies limited rationality and to the environmental authoritys but partial compliance with its invoicing and monitoring activities. The article concludes that, faced with such strategic behavior by the contaminating agent companies, a non-declared higher contamination tax would increase the effectiveness of the control instrument.
Palavras-chave : Retribution tax; genetic algorithms; limited rationality; social learning.