Conspiracy against the public - An experiment on collusion

Åshild A Johnsen,Ola Kvaløy

doi:10.1016/j.socec.2021.101742

Åshild A Johnsen, Ola Kvaløy

Open Access

https://doi.org/10.1016/j.socec.2021.101742

Copy DOI

Abstract

We study to what extent collusive behavior is affected by the awareness of negative externalities. Theories of outcome-based social preferences suggest that negative externalities make collusion harder to sustain than predicted by standard economic theory, while sociological theories of social ties and intergroup comparisons suggest that bilateral cooperation can be strengthened if there exist outsiders that gain from cooperative break down. We investigate this in a laboratory experiment. Subjects play the infinitely repeated prisoner's dilemma with and without a negative externality. The externality is implemented by letting subjects make a positive contribution to a public good if they choose to deviate from cooperation between the two, i.e. cooperation is collusive since the gains are at the expense of the public. We find that this negative externality tends to increase collusive behavior. Initially, the level of cooperation is lower, but as subjects gain experience and observe that their partners choose to cooperate despite the negative externality, they cooperate as least as much as in the baseline treatment.

Highlights

There are two main motives for cooperative behavior in repeated game prisoner’s dilemma
There are numerous equilibria in super games like this, and it is straightforward to construct equilibrium strategies where the players after n periods cooperate more in the public good game. This is our behavioral prediction: We will initially see less, but more cooperation when cooperation is at the expense of the public good
This paper studies to what extent collusive behavior is affected by the awareness of negative externalities

Summary

Introduction

There are two main motives for cooperative behavior in repeated game prisoner’s dilemma. Our paper relates most closely to Malmendier and Schmidt (2017), who investigate the negative externalities of gift giving, e.g. business gifts, where the giver hopes to get favorable treatment from the receiver They find that the gift triggers an obligation to repay the gift, even if the gift is given with the intention to affect the decision of the recipient at the expense of a third party. Consistent with Bland and Nikiforakis, we find that subjects put positive weight on the welfare of third parties in the first round of the repeated game Another strand of literature that (implicitly) studies the effect of negative externalities on prosocial behavior is the experimental papers on bribery games, which are similar to repeated gift exchange games.

Experimental design and procedure

Behavioral Predictions

Experimental results

The individual’s collusion rate is calculated in two steps

Conclusion