An Experimental Study On Individual’s Corrupt Behaviours

Corruption is considered as one of the most important factors which impede the economic and social development, an issue of long term and continuous research concerns to economists and other social scientists. Much effort has been devoted to explore the causes and the determinants of corruption in order to identify the most efficient anti-corruption mechanisms. Despite the fact that some developments of theoretical and empirical research on corruption have been made over the past 20 years and institutional arrangement for anti-corruption was put in place based on economic, social and cultural variables, many obvious deficiencies still exist in the above-mentioned methods, which may result in inefficiency in our effort to fight corruption.Compared to the traditional studies, more articles address the issue of corruption through experimental methods on well-known international journals, showing that the internal validity and external validity of experimentation have gained increasing recognition.. Being the first thesis that tries to investigate corruption by experimentation systematically, this article focuses on the bribery and embezzlement that involves relatively small amount of money, identifies key influencing factors on taking bribes on the basis of three groups of experiments, and conducts a comparative study of several anti-corruption mechanisms. By so doing, this thesis can address the gap of experimental research of corruption in China and contribute due theoretical and empirical innovation on the one hand, while provide food for thought for the institutional design and building of anti-corruption mechanism amidst new situations facing China.Why does corruption still linger, and even prevail in some areas under ever increasing pressure of anti-corruption efforts? This thesis introduced in both injunctive social norms and descriptive social norms in a modified bribery game experiment to examine the effects of the two social norms on corruption systemically for the first time in the laboratory. The results showed that social norms exerted the framing effect and the herding effect on corruption. On the one hand, reducing explicit use of the word "corruption" could prohibit the framing effect from curbing corruption. On the other, when corruption became a norm, passive herding effect would strengthen the tendency of individual towards corruption and when corruption was not yet a social norm, it would only strengthen the framing effect. These behavioral patterns could be well predicted by people’ sexpectations. This thesis holds the view that corruption as a social norm is the key factor that contributes to the persistence of corruption, whereby we should take into account the social norm factor for corruption when designing anti-corruption institutions and policies.Secondly, this thesis introduced in two independent officials and one player with the identity of intermediary in an improved bribery strategy game experiment to study the "four eyes" principle and effect of intermediary on the principle for the first time. The experimental results indicated that the "four eye" principle could significantly reduce the completion of corrupt exchanges, but it had no effect on the official’s own behavior It could neither reduce the frequency of officials taking bribes by the influence of external incentives, nor could it increase officials’ psychological costs in taking bribes by internal incentives; and the real intermediary could not only increase the frequency of corrupt transactions by eliminating uncertainty, but also could reduce the psychological cost of both corruption participants, thereby further increasing the frequency of corrupt transactions. These results showed that although the "four eyes" principle is respected by many international organizations, its effectiveness should be further examined; and the role of intermediary organizations in the corrupt transactions should be given enough attention, and requires strengthened supervision and needs to be properly regulated. Finally, this thesis throws some light on the Property Declaration System, which has been regarded as an efficient mechanism to crack down upon corruption. By an incomplete information dynamic game model, this thesis examined the feasibility of the Property Declaration System from a purely theoretical perspective for the first time. The equilibrium result showed that there was no separate equilibrium between the corrupt officials and honest officials in declaring their real income. The Property Declaration System under current institutional arrangement proved no effect on unveiling corruption. Furthermore, the evidences from the laboratory confirmed this theoretical result. In fact, there is either no empirical evidence that proves a corrupt official will confess his corruption by the system. By contrast, a property declaration mechanism sampling at a set ratio showed more significant effect on prevention of corruption.