The definition of conflict

Although conflict is a subject of study in all fields of the social sciences, the definition of conflict is not the same across disciplines. Economists, for example, define conflict as a situation in which competitors choose costly inputs in pursuit of private payoffs framed as wins and losses (Garfinkel & Skaperdas Reference Garfinkel and Skaperdas2007; Kimbrough et al. Reference Kimbrough, Laughren and Sheremeta2019). Evolutionary biologists define conflict in terms of Darwinian fitness (Darwin Reference Darwin1859), where at least one of the species negatively affects the others’ fitness (Rusch & Gavrilets Reference Rusch and Gavrilets2019).

De Dreu and Gross (D&G) propose a definition of conflict as a game of attack-and-defense. Indeed, many conflicts have such structure: There is an attacker who wants to revise the status quo and a defender who wants to protect it. Also, such a definition of conflict conceptually fits most disciplines, by pointing out incompatibility of interests by the parties involved in the conflict.

Attack-and-defense games

The main assumption behind the attack-and-defense game of D&G is that conflicting parties have asymmetric objectives. As a result of this asymmetry, the solution to such a game is a mixed-strategy Nash equilibrium.

The continuous version of the attack-and-defense game resembles an all-pay auction (contest) with externalities (Baye et al. Reference Baye, Kovenock and De Vries2012; Chowdhury & Sheremeta Reference Chowdhury and Sheremeta2011). Also, such a game is closely connected to the attack-and-defense games of a weakest-link network of targets (Clark & Konrad Reference Clark and Konrad2007; Kovenock & Roberson Reference Kovenock and Roberson2018; Kovenock et al. Reference Kovenock, Roberson and Sheremeta2019), in which the attacker's objective is to assault at least one target successfully, and the defender's objective is to defend all targets. Finally, the attack-and-defense game is closely related to games of profiling (Holt et al. Reference Holt, Kydd, Razzolini and Sheremeta2016; Kydd Reference Kydd2011), in which attackers choose which demographic “type” to recruit and defenders choose which demographic types to search. All aforementioned conflict games have a structure where there is an attacker and a defender who have asymmetric objectives and whose interests are incompatible. Therefore, these games are directly linked to the attack-and-defense game of D&G.

Behavioral considerations

There are many behavioral considerations that may influence the actual behavior of competitors in the game of attack-and-defense. First, if attackers are inequity averse (Fehr & Schmidt Reference Fehr and Schmidt1999), then conflicts could be less intense than predicted by the standard game theory. However, if attackers are spiteful (which is a more realistic assumption), then conflicts are more likely to escalate (Mago et al. Reference Mago, Samak and Sheremeta2016). Conflicts could also escalate if competitors, in addition to monetary utility, derive a utility from winning itself (Sheremeta Reference Sheremeta2010), or if competitors are regret averse (Filiz-Ozbay & Ozbay Reference Filiz-Ozbay and Ozbay2007).

Other important behavioral considerations include guilt aversion, loss aversion, overconfidence, impulsivity, and various emotional responses. Without proper game-theoretic analysis (Konrad Reference Konrad2009) and experimental testing (Dechenaux et al. Reference Dechenaux, Kovenock and Sheremeta2015), it is not clear how these behavioral factors impact individual behavior of competitors in the game of attack-and-defense.

Intergroup games of attack-and-defense

Many conflicts involve multiple attackers and defenders, resulting in an intergroup conflict game with asymmetric objectives (Chowdhury & Topolyan Reference Chowdhury and Topolyan2016a; Chowdhury et al. Reference Chowdhury, Lee and Sheremeta2013). Scientists from different fields have been fascinated by such games not only because of their prevalence in real life, but also because intergroup conflicts have a number of interesting features with non-trivial trade-offs. For example, attackers may have an incentive to cooperate with each other by expending effort in order to carry out a successful attack; however, because effort is costly, each attacker also has an incentive to abstain from expending any effort and instead free-ride on the efforts of others. Also, because attackers and defenders have asymmetric objectives, the problem of free-riding can be different for the group of attackers and the group of defenders. D&G point out that defender groups share a common fate when they lose, and so their individual interests are more aligned than those of attacker groups. Consequently, this asymmetry results in stronger incentives to free-ride among attackers than among defenders. Another interesting asymmetry pointed out by D&G is that in-group identification could be stronger among defenders than attackers. These are interesting hypotheses that are worth further investigation.

Mechanisms of intragroup cooperation

The asymmetry in free-riding incentives between attackers and defenders creates a greater need for attackers to use various cooperation mechanisms in solving the free-riding problem. Such mechanisms could involve negative reinforcement, such as shaming, sanctioning, punishment, and ostracism, as well as positive reinforcement, such as communication, leadership, and feedback. Although most of these mechanisms have been studied in the context of intergroup conflicts (Sheremeta Reference Sheremeta2018), they have not been studied in the context of intergroup attack-and-defense games. This is an interesting avenue for future research.