Games of corruption: how to suppress illegal logging.

Corruption is one of the most serious obstacles for ecosystem management and biodiversity conservation. In particular, more than half of the loss of forested area in many tropical countries is due to illegal logging, with corruption implicated in a lack of enforcement. Here we study an evolutionary game model to analyze the illegal harvesting of forest trees, coupled with the corruption of rule enforcers. We consider several types of harvesters, who may or may not be committed towards supporting an enforcer service, and who may cooperate (log legally) or defect (log illegally). We also consider two types of rule enforcers, honest and corrupt: while honest enforcers fulfill their function, corrupt enforcers accept bribes from defecting harvesters and refrain from fining them. We report three key findings. First, in the absence of strategy exploration, the harvester-enforcer dynamics are bistable: one continuum of equilibria consists of defecting harvesters and a low fraction of honest enforcers, while another consists of cooperating harvesters and a high fraction of honest enforcers. Both continua attract nearby strategy mixtures. Second, even a small rate of strategy exploration removes this bistability, rendering one of the outcomes globally stable. It is the relative rate of exploration among enforcers that then determines whether most harvesters cooperate or defect and most enforcers are honest or corrupt, respectively. This suggests that the education of enforcers, causing their more frequent trialing of honest conduct, can be a potent means of curbing corruption. Third, if information on corrupt enforcers is available, and players react opportunistically to it, the domain of attraction of cooperative outcomes widens considerably. We conclude by discussing policy implications of our results.