By Dr. Ken Broda Bahm:
As of press time for this post, 173,094 Americans have died from the novel coronavirus. If the United States had followed the example of other developed countries by taking earlier and more decisive actions against the virus, our country could have prevented approximately 70 percent of these deaths. That, at least, is the conclusion of an analysis conducted by Isaac and James Sebenius of Harvard University. It is also a causal statement that attributes responsibility: If American leaders had acted differently, many of these lives would have been saved, our reopening would be safer, and our economy would be closer to normal. Logically, this is an expression called a “counterfactual.” If past actions had been different, then different results would have been obtained. Hypothetical reasoning like this can feel speculative, but it is a mainstay of legal judgement: If the driver had been more careful, the accident would have been avoided; If the details had not been hidden, the party would not have signed the contract.
Research has shown that people will generate their own counterfactuals as a way of testing claims, and the law even enshrines this in the “but for” test (as in, “But for the doctor’s negligence, the injury would not have occurred“). An attorney or an expert witness can frame their arguments as counterfactuals to help jurors with this process. There are also strategic choices to be made, since varying one actor or event, and not another, can make the varied action appear more salient. So constructing counterfactuals is a way of directing conclusions of causal responsibility. A recent study appearing in the journal Applied Cognitive Psychology (Catellani et al., 2020) confirms this for expert witnesses: When they use counterfactuals, they are able to enhance the perception of causation and responsibility for the factor being varied. They concluded, “expert witnesses’ counterfactuals are a powerful communicative tool to influence the judicial reasoning of both lay jurors and expert judges.” In this post, I’ll take a look at the study and its implications for expert testimony.
The Study:
The study says to experts, if you use counterfactuals, then you’ll do better. The authors, psychology researchers in Italy, tested the theory using mock jurors in the first study and actual judges in the second. Both sets of participants reviewed a medical malpractice scenario that included elements of patient choice, physician choice, and external circumstances. Different versions of an expert report included counterfactuals focused on the plaintiff (If the patient had strictly followed the doctor’s orders...), on the defendant (If the doctor had considered the symptoms more carefully…), or on external circumstances (If the mountain town had a hospital…).
What studies found is that an expert’s use of a counterfactual focused on one party increased the causal focus and responsibility placed on that party. “Our findings showed for the first time that the focus of counterfactuals included in an expert witness’s report consistently influences causal and responsibility attributions regarding a judicial case, and that this is true for both lay jurors and expert judges.”
Prior research has shown that other factors matter as well — i.e., when the injury is more severe or more foreseeable, we are more likely to focus on responsibility. The researchers found, however, that an expert’s use of counterfactuals can moderate these other factors, instead focusing responsibility on the factor being varied.
The Advice:
Use Counterfactuals
Invite your fact-finders to think about what would have happened if the facts had been different. Granted, that can feel speculative, but there are likely to be facts that support the counterfactual as a reasonable inference. Experts will likely reach their conclusions analytically, and not through a simple thought experiment. But in communicating those conclusions, the counterfactual form provides a way for fact-finders to visualize and understand those conclusions.
Be Conscious of What You’re Varying
There is never just one counterfactual, and when there are rival counterfactuals (e.g. either the doctor or the patient behaving differently) then you need to account for rival theories. For example, a plaintiff’s expert might argue that if the doctor had taken greater care with the diagnosis, then that would have avoided harm whether the patient followed instructions or not, and that, combined with the doctor’s professional obligations, makes the doctor’s counterfactual the relevant one. Strategically, experts want to construct counterfactuals that highlight the analyses that are already done, and support those conclusions.
And Remember that Credibility Still Matters
Of course, counterfactuals have to be rational and supported by the evidence as well. There are a number of factors that could have been different, and not all of them are equally plausible. For example, the counterfactual that involves the patient following instructions is going to be more salient and more available than the counterfactual of the small mountain town gaining a first-rate hospital. Generally speaking, counterfactuals that involve less overall change to the world are going to be seen as more credible counterfactuals. The bottom line is that the counterfactuals, and the rest of the testimony, needs to be grounded in reality to be credible. Based on past research, the authors note, “the influence of counterfactual communication can diminish or even disappear when the counterfactual source is perceived as unreliable.” Counterfactuals provide a good communication tool, but the expert on the whole needs to come across as credible.
______
______
Other Posts on Counterfactuals:
- Make Events Meaningful by Inviting Counterfactuals
- Ask If Your Jurors’ Causal Thinking Is Based on Facts or Possibilities
- Address the Central Unknown
______
Catellani, P., Bertolotti, M., Vagni, M., & Pajardi, D. (2020), How Expert Witnesses’ Counterfactuals Influence Causal and Responsibility Attributions of Mock Jurors and Expert Judges. Applied Cognitive Psychology. DOI: 10:1002/acp3720