Elsevier

Economics Letters

Volume 188, March 2020, 108941
Economics Letters

Conditional cooperation: Type stability across games

https://doi.org/10.1016/j.econlet.2020.108941Get rights and content

Highlights

  • We identify behavioral types by a prisoner’s dilemma and a public goods game (PGG).

  • Conditional cooperation is a stable pattern across games.

  • Selfish types in the prisoner’s dilemma, however, are more cooperative in the PGG.

Abstract

To classify cooperation types, a sequential prisoner’s dilemma and a one-shot public goods game are convenient experimental setups. We explore the within subject stability of cooperation preferences in these two games. We find that the prisoner’s dilemma performs well in identifying conditional cooperators while it is only an imperfect tool for identifying selfish types in the public goods game.

Introduction

An important contribution of behavioral economics is to establish the relevance of an additional behavioral type beyond the purely payoff-maximizing “homo oeconomicus”, namely “homo reciprocans”, who has a preference for reciprocal social behavior and represents a large fraction of the population.1 Consequently, type dependency regarding the intrinsic motivation to cooperate is a key question in the literature.2 Two methods have been frequently applied when researchers need to determine the discrete behavioral type of subjects. On the one hand, one can use the method of Fischbacher et al. (2001) which relies on a conditional contribution vector elicited by the strategy method in a one-shot public goods game (FGF hereafter).3 This method is typically based on a set of 22 questions. On the other hand, a simple sequential prisoner’s dilemma (SPD hereafter), for which only three questions are sufficient, can be used for type classification as well (Miettinen et al., 2017, Kosfeld, 2019, Eichenseer and Moser, 2019).

The discrete behavioral type of individuals towards cooperation can be used as an explanatory variable for various economic research questions. The FGF method has, for example, been used by Rustagi et al. (2010), to investigate whether the share of conditional cooperators in a group of Ethiopian forest workers has an impact on the quality of joint forest management, thereby connecting the lab to the field.4 Since FGF is still the standard tool for such applications, the question arises whether using the simpler method is sufficient for type classification as it may save time and reduce cognitive load for the participants. To the best of our knowledge, there exists no systematic comparison of classification congruence between these two procedures.

Consequently, the aim of this paper is to assess the stability of classifications across games thereby contributing to the literature on the within subject stability of cooperation preferences (Blanco et al., 2011, Volk et al., 2012). To this end, we compare the types assigned by SPD to those assigned by FGF in its latest refinements (Fallucchi et al., 2018, Thöni and Volk, 2018).

Section snippets

Protocol

The experiment was conducted on Amazon Mechanical Turk (MTurk henceforth) in December 2018 using a sample of MTurk experienced US residents. In total, 232 participants took part in the experiment earning $2.85 on average with an average completion time of approximately 13 min. About half of the subjects (120) played SPD first, while the other half (112) was doing the FGF task first. Subsequently, the participants completed a short questionnaire on age, gender, and education. Instructions for

Results

In our data analysis, we investigate the relationship between the discrete behavioral types classified by SPD and FGF in the refinements of Thöni and Volk (2018) and Fallucchi et al. (2018). The refinement of Thöni and Volk (2018) of FGF (FGF-T hereafter) resembles a theory-driven approach and is based on the Pearson correlation coefficient. It distinguishes the five behavioral types depicted in Table 2.

In our sample, we can categorize 184 out of 232 subjects (79.3%) as conditional cooperators

Summary and conclusion

With regard to the consistency of discrete behavioral types, our results indicate that SPD performs well in identifying subjects with a stable pattern of conditional cooperation. Given that a subject is of CC type in SPD, the probability is 93.3% to be classified as CC as well according to FGF-T (refinement of Thöni and Volk, 2018) and 88.8% according to FGF-F (refinement of Fallucchi et al., 2018), respectively. We further observe that the distinction between WCC and SCC is helpful for

Acknowledgments

Michael Eichenseer and Johannes Moser acknowledge funding by the International Doctoral Program “Evidence-Based Economics” of the Elite Network of Bavaria. We would like to thank Wolfgang Buchholz, Francesco Fallucchi, Moritz Janas, Martin Kocher, Michael Kosfeld, Andreas Roider, and Christian Thöni for helpful comments.

References (20)

There are more references available in the full text version of this article.
View full text