All you need to know on the Hindsight Bias

A few months ago, I published a #30in30 regarding the hindsight bias. As I promised, during this essay, I will give you more information on the impacts on the legal profession as well as on debiasing this classic behavioral bias. This article will be a thorough introduction. Hence, there is no need to read the original #30in30 post.

As you might remember from my earlier posting: There is no rational reason to be annoyed by people claiming “had you asked me, I could have told you that would happen”. Our brains cannot help but think this way: Tricked by the hindsight bias, people tend to rate the likelihood of events higher when already knowing the actual outcome.

What has the potential to annoy us in our everyday lives can turn hazardous when giving legal advice to your clients. Moreover, both judges and juries are likely to be misguided into applying a much higher standard of proof – especially in cases involving unresolved questions of negligence.

Ex Post ≠ Ex Ante

In 1975, B. Fischhoff conducted seminal research that is serving up to today as one of the most vivid examples of what was formerly known as the “Knew-it-all-Along Effect”. Five groups of participants learned about a random historical event of minor importance. It was unknown enough that none of the participants had ever heard of it. One group was asked to evaluate the likelihood of four different outcomes. All other groups were told that one of these outcomes actually happened. Each of the four groups was significantly rating the outcome of their respective scenarios higher than the control group did at the beginning.

Twenty years later, Kim A. Kamin and Jeffrey J. Rachlinski researched the implications on the legal profession by asking three groups of participants whether they considered the precautions of a municipality against flood damage to be negligent. The first group was given exactly the information the municipality had when making the decision of whether they should increase their precautions against floodwaters. A blunt 75 % of participants of this group concluded that the costs were too disproportionate given the estimated likelihood of a flood. The second group however was given the facts plus information that the flood damaged a riparian property. It will not surprise you that this group gave higher weight to the unfortunate outcome and thus found the evaluation of the municipality with a majority of 57 % to be negligent. It is quite interesting that the third group was concluding the city to be legally negligent at 56 %, although they were expressly told about the risks of hindsight distortion.

In another study, Judges had been confronted with trial court judgements and different information regarding the outcome of an appeal in front of the relevant appellate court. The participants were asked to give their opinion on the outcome of the appellate court. Those participants that were told that the appeal court sustained the trial court’s judgement found it with a majority of 81.5 percent reasonable. On the other hand, those participants that were told the appellate court revoked the trial court’s decision found it in 27.8 more reasonable to go with the trial court’s view. (Guthrie/Rachlinski/Wistrich (1997), Cornell Law Rev. 86, 2000-2001, 777)

German Laws prone to the Hindsight Bias?

Judges in Germany are exposed to the hindsight bias in many different situations. Section 276 Para. 2 German Civil Code for example states, that an obligor is behaving negligent, if danger for a legally protected right was foreseeable. Relevant is in this case the foreseeability for an objective member of the obligor’s business branch. Hence, a judge needs to rule on what an average objective person would have been able to foresee. In tort law however, the German Federal Supreme Court rules continually that the behavior of an understanding and faithful objective person shall be the standard to determine a breach of the duty of care. (Steinbeck/Lachenmaier in: NJW 2014, 2086)

These examples show quite clearly that the hindsight bias has a major impact on our judgement. Even if we know that there is a significant risk of misjudgement based on knowledge of an outcome, we cannot change our analyses without an efficient debiasing technique.

The Legal Profession reacting to the Hindsight Bias

When facing the hindsight bias, there are basically two different kinds of methods to combat it in general: Firstly, methods that are set in written law. Secondly, methods that try to change the thinking of those who apply the law.

A good example for the first category is the business judgement rule in Sec. 93 Para. 1 Sent. 2 of the German Stock Corporation Act (Aktiengesetz). It states: “A breach of duty is not given in cases the member of the board thought with due reasons that his actions would turn out beneficial for the company.” This is what some scholars call “debiasing through procedural rules”. The Stock Corporation Act itself states rules that take counter measures against the hindsight bias. (Redeke in: ZIP 2011, 59)

The audit industry faces the hindsight bias in the annual audit of large companies’ balance sheets. This is a perfect example of both, an industry developing standards to protect itself against the hindsight bias as well as for a useful benchmark for deciding on the legal reasons of ones behavior. The Institute of Public Auditors in Germany publicizes standards for prudent auditing. These standards are written to ensure a rightful audit process. However, they are trying to protect the auditors from their hindsight bias by explicit instructions, such as: “By comparing the rough estimate to the true data the major assessment in former report periods should not be questioned by using other assessment strategies than possible at the time of the former report.” (IDW PS 314 n.F. S. 14)

Steinbeck/Lachenmaier discuss the positive effects of the decision-maker searching for arguments for the opposite view on their decision. Further it can be helpful to think at any time about rules that were in place when the incident happened. This way, the decision maker can change from a subjective ex post view to a rather objective ex ante view. In many cases the relevant industry developed rules and standards to minimize hazard – these could turn out to be a useful benchmark.

Another possible way to confront the hindsight bias is to rely on legal case groups and previous decisions. This proved effective in the following example: In a study, the judges on the one hand showed hindsight bias when being asked to evaluate the probability that evidence will be found in a search of a car (after being told, that at the end they found the evidence in an illegitimate search). However, on the other hand, the judges did not show any bias regarding the lawfulness of the illegitimate search. (Steinbeck/Lachenmaier in: NJW 2014, 2086) And while this shows that there is still a long way to go there is as well hope for a more rational future in the law. Join us, join the Rational Think Tank in our quest to foster equity and objective judgement in front of our law.

PHILIP HATTEMER