Do groups fight more? Experimental evidence on conflict initiation

This paper investigates whether distributional conflicts become more likely when groups are involved in the fight. We present results from a laboratory experiment in which two parties can appropriate resources via a contest or, alternatively, take an outside option. Keeping monetary gains expected from fighting constant across all treatments, the experiment compares conflict choices of players in two-against-two, one-against-one, and two-against-one settings. Overall, we find evidence for a higher propensity to opt for conflict when entering the fight in a group than when having to fight as a single player. The effects are strongest in endogenously maintained groups and in the presence of group size advantages (i.e., in two-against-one). The results can be explained by a stronger non-monetary utility from fighting in (endogenous) groups and coincide with a biased perception of the fighting strength in asymmetric conflict.