The fairness/efficiency issue explored through El Farol Bar model

Abstract

The relationship between fairness and efficiency is a central issue for policy makers. To date there is no agreement among economists whether a public policy that pursues fairness entails a loss of efficiency of an economic system, or an oriented towards efficiency policy is able to facilitate the achievement of a higher fairness. In this paper we present the fairness/efficiency issue through the analysis of four agent-based simulations that implement the El Farol Bar Model proposed by Brian Arthur in (1994). As El Farol Bar models are not interested in exploring the equity issue, we modified them by introducing a set of measures of efficiency and fairness. Particularly, we assume that fairness is a short run issue and we develop time-based measurements of fairness. The computational analysis shows that selfish agents are able to achieve an efficient use of available resources, but they are incapable of generating a fair use of them. The analysis shows that random choices of agents generate an apparent fairness, which occurs only in the long run. Thus, a suitable relationship between efficiency and fairness could be reached by a public policy which defines an appropriate pay-off matrix able to drive agents’ choices.