…or “Why can’t we all just get along?”
In a world of economically rational agents, there would be no strikes, civil legal disputes wouldn’t come to court, marital break-up would not be so acrimonious and there probably wouldn’t be any wars. One strand of bias research deals with why there is so much conflict in human life.
A labour dispute, for example, could be resolved on paper by rational agents. The union would threaten to strike, the employer threaten to dock pay, and as each party spells out their position, they would negotiate an equilibrium. They could then sign a contract giving each other some concessions but committing not to take any of the threatened action. There’s a worked numerical example in Dixit and Nalebuff’s Thinking Strategically. Strikes cost the company and the strikers a great deal in lost production and lost pay, so it’s in their rational interest to have the strike “on paper” like this. (Perhaps it would be rational for unions to strike once in a while, just so that their threats are credible).
When there is a civil legal case, there are usually many pre-trial hearings, during which the two sides set out all the arguments they will use, and look for precedents in similar cases. Each side becomes increasingly informed about the strength of the other’s position, and about how this case is likely to turn out. Eventually both sides are working from the same information about the case, So there’s a scientific question of why such cases ever come to trial. Given that actually having a trial is very costly in legal fees, psychological stress, time spent and so on, it’s in the collective rational interest of the parties to agree a settlement.
War is enormously costly, obviously in terms of dead and maimed soldiers and civilians but in a more narrowly economic sense. As well as the costs of transporting and supporting people and equipment, there is lost productivity in that men and women in their prime are taken out of the economy to go and fight. War may be very profitable for the arms industry, but costly to the economy as a whole, because we would be richer if we could spend the resources on other things.
So, if military solutions are so costly, why are they so often taken?
A number of the biases discovered by psychologists bear on why conflicts of any sort are so acrimonious; in particular why agents can have totally opposed perspectives on an issue when, in a strictly rational sense, they have the same information.
The fundamental attribution error is an effect where a person’s reaction to circumstances is misread as part of their personality. Imagine that you are frustrated by someone’s apparent stand-offishness and become angry while talking to them. If they conclude you are an angry person by nature, and fail to see the role of their own attitude, they are making an attribution error. Conversely, if you fail to see that their attitude might be a result of the way you treat them, you’d be making the same error.
An already tense disagreement can become more aggressive because, with the fundamental attribution error, each side becomes convinced that the other is unreasonable and belligerent. Chapter 6 of Elliot and Aronson’s Mistakes Were Made (but not by me) has more about how attribution errors contribute to conflict in marriage.
Memory biases, especially self-serving biases, can distort so that each side can see their own behaviour as more reasonable than the other. See Chapter 3 of Elliot and Aronson.
There is a demon effect where someone with a salient negative quality in judged to be negative in other respects (the opposite process is called the halo effect). This predicts that if you aren’t familiar with someone but know something about them that you find objectionable, you might well be prejudiced to think them unintelligent, selfish or unreasonable. So when, say, capitalists and communists meet across a negotiating table, they might have pessimistic expectations of what could be achieved by negotiation.
The hostile media effect might make both sides feel that neutral parties are treating them unfairly.
Another process that may worsen conflict is polarisation of attitude due to social pressure. Say that Alice and Bob differ on some political issue that they both care about. They might go to different meetings, listen to different pundits, and generally identify with different groups. Many social effects might effect how their attitudes evolve (e.g. conformity), how they see the other group (e.g. stereotyping) and how they decide to handle the disagreement (e.g. risky shift). These processes could escalate an initially small disagreement.
Another bias that worsens conflict is what I would call the egocentric assessment of pain i.e. people considering their own pain and suffering to be more significant than other people’s. Elliot and Aronson report on an experiment by Shergill et al. (2003) where subjects were connected in tit-for-tat situation where they could each apply pressure to the other’s index finger. Although both participants were told to apply the same amount of pressure that they had just received, the pressure escalated rapidly. It seems that, to each subject, a great deal of pressure on the other person’s finger was judged “the same as” a relatively small pressure on their own.
An area where there has been specific research on the role of biases in conflict is a series of papers by Linda Babcock, George Loewenstein and colleagues on self-serving biases. This theory says that, where there is uncertainty in the evidence, each party will interpret that uncertainty in their own favour, and form an optimistic picture of what they can get out of the dispute. There are known optimism biases and superiority biases affecting a lot of human judgement. So it’s not surprising that one or both sides will think they can do better than the negotiated settlement, and reject it.
Babcock and Loewenstein in particular looked at pay negotiations for public school teachers in one state of the USA. In order to decide on a pay settlement, the union and employer (the school board) would each look for “comparable” districts to decide what amount is fair. Assuming that unions will be biased towards high pay settlements and employers towards low settlements, negotiations should be more likely to break down when there a greater range of potential comparison districts. This is exactly what the researchers found.
In a laboratory context, Babcock and Loewenstein used a role-play in which subjects took the role of plaintiff or defendant in a civil case. Their task was to negotiate a settlement, with a penalty if they failed to agree. The background information was taken from a real case, although they were not told the size of the award. This experiment clearly demonstrated the self-serving bias, in that plantiffs thought a much higher award would be paid out (and was fair) than did defendants. The greater the discrepancy between the two parties, the more likely the negotiation would fail.
A conventional economic explanation of failed negotiation would involve each side having different information, but here the two parties had the same information (i.e. the case study) but different randomly-assigned roles.
References
Linda Babcock, George Loewenstein, Samuel Issacharoff and Colin Camerer (1995) “Biased Judgments of Fairness in Bargaining” The American Economic Review, Vol. 85, No. 5., pp. 1337-1343.
Linda Babcock and George Loewenstein (1997) “Explaining Bargaining Impasse: The Role of Self-Serving Biases” The Journal of Economic Perspectives, Vol. 11, No. 1. (Winter, 1997), pp. 109-126.
Sukhwinder S. Shergill, Paul M. Bays, Chris D. Frith, Danel M. Wolpert (2003) “Two Eyes for an Eye: the Neuroscience of force escalation” Science Vol. 301. no. 5630, p. 187 DOI: 10.1126/science.1085327
Thanks to Paul Latreille at Swansea University for directing me to the Babcock et al. research.
Bias and Belief 