We are Not Alone: The Impact of Externalities on Public Good Provision

Abstract

Abstract:Public good provision is often local and also affects bystanders. Is provision harder if contributions harm bystanders, and is provision easier if outsiders gain a windfall profit? In an experiment we observe that both positive and negative externalities reduce provision levels whenever actors risk falling back behind bystanders. The mere presence of unaffected bystanders already dampens contributions. This behavior seems to result from the interplay of two motives: the desire to realize opportunities for joint gains, and concerns for comparative performance. Individual payoff comparisons to the other actors as well as to individual bystanders drive contributions down.

Harm on an Innocent Outsider as a Lubricant of Cooperation – An Experiment

“…It turned out that contribution decisions were not driven by the direction of the externality, but by the comparison with bystander payoffs (Engel and Rockenbach 2011). In the present experiment, by the design of the Externality treatment, the outside player is always worse off.…”

Tacit Coordination in Games with Third-Party Externalities

“…The evidence suggests that participants care less about the welfare of inactive third parties than the welfare of other decision makers (e.g., Engel and Zhurakhovska, 2012; Güth and van Damme, 1998; Kagel and Wolfe, 2001). Nevertheless, the existence of third parties has been shown to have a signi…cant impact on behavior (e.g., Ellman and Pezanis-Christou, 201), especially when decision makers are in a worse monetary position than the third parties (Engel and Rockenbach, 2011;McDonald et al, 2013). 5 Furthermore, even if agents care less about third parties, the uncertainty about the social preferences of other decision makers in our experiment could serve as a multiplier increasing the number of agents willing to act in the interest of the third party.…”

Competition as a Socially Desirable Dilemma - Theory vs. Experimental Evidence

“…The experimental evidence used to back the behavioural claims is taken from three different papers (Engel and Normann 2008;Engel, Irlenbusch et al 2009;Engel and Rockenbach 2009) and a meta-study (Engel 2007a) of my own 1 .…”

“…While these are interesting hints, neither of these findings can serve as proof. This motivated an experiment by Bettina Rockenbach and myself (Engel and Rockenbach 2009). We had groups of 7 interacting over 10 announced rounds, who were randomly split into 4 active and 3 passive players.…”