How to Reduce Self-Serving Bias

Self-serving bias can be a powerful roadblock to flexible and adaptive thinking. It usually leads us to selectively process information and focus on favorable aspects of a problem, and as a result, we usually overestimate our strengths and underestimate our weaknesses.

Fortunately, it is possible to reduce self-serving bias and think more effectively overall. To achieve that, we first of all have to become aware of this bias. However, awareness of self-serving bias by itself is not enough. Some popular debiasing techniques such as arguing our opponent’s case are not only ineffective but probably will backfire. A more effective technique is to list weaknesses in our own position.

Research on Self-Serving Bias

Self-serving bias was perfectly demonstrated in a series of studies by George Lowenstein, Linda Babcock, Samuel Issacharoff, and Colin Camerer, published in 1990s.[1] Their participants had to negotiate an auto-injury settlement on behalf of either a plaintiff or a defendant. They read a tort case, based on a real Texas case, in which a motorcyclist was injured by an automobile driver and now the motorcyclist claimed $100,000 in damages. All the subjects read the same 27 pages of the original Texas case, including police reports, witness testimony, maps, and testimony of the parties.

Before negotiating the settlement, the subjects had to guess what a Texas judge would award if they failed to settle the case. (It was $30,560). The subjects would also receive a bonus if they guessed within $5,000 of the award. So at this stage, all subjects read the same materials and they didn’t have to engage in advocacy, only to guess what the neutral party would award. The plaintiffs’ predictions, however, were on average almost $15,000 higher than those of the defendants.

Yet, when the experimenters didn’t tell which side they would represent, the subjects had no problem guessing the correct award. This happened because when they didn’t know which side they would represent, the subjects were not sure which arguments would be favorable and this prevented them from selectively processing the information they read.

Debiasing Techniques That Don’t Work Against Self-Serving Bias

The researchers also tried various methods to cure this bias, but most of them failed:

  • Financial Penalties. The subjects who failed to negotiate the settlement, not only wouldn’t get the bonus, but they would also suffer penalties because of the legal costs. Such financial incentives failed to change anything.
  • Bias Awareness. The researchers also tried telling the subjects about the self-serving bias and that they might fall for it. It didn’t help. The subjects thought that the other person might fall for it, but that it didn’t apply to them.
  • Arguing the Opponent’s Case. The researchers also asked subjects to write an essay arguing their opponent’s case. This method not only failed to reduce the bias, but it backfired because now they knew their opponent’s weaknesses better and so they thought they deserved even more.

Acknowledging Weaknesses

The only effective and practical debiasing technique was to tell subjects about the self-serving bias and also ask them to list weaknesses in their own case. Being aware of their own weaknesses, the subjects now thought more objectively about the judge’s estimates.

This technique effectively reduced the bias. In the control group, the difference between the parties’ estimates was $21,783 and 35% failed to settle. Among the subjects who thought about their weaknesses, the difference was only $4,674 and only 4% failed to settle.


[1] George Loewenstein, Samuel Issacharoff, Colin Camerer, and Linda Babcock, Self-Serving Assessments of Fairness and Pretrial Bargaining, Journal of Legal Studies, 100, 73-126 (1993); Linda Babcock, George Loewenstein, Samuel Issacharoff, and Colin Camerer, Biased Judgments of Fairness in Bargaining, American Economic Review, 85, 1337-1343 (1995); Linda Babcock and George Loewenstein, Explaining Bargaining Impasse: The Role of Self-serving Biases, Journal of Economic Perspectives, 11, 109-126 (1997).

%d bloggers like this: