We thank Nada Gligorov for her careful reading of our paper. Although she raises some interesting questions that merit further investigation, several points of disagreement between us remain.
Our paper, “Affective Forecasting and Its Implications for Medical Ethics,” delivers a cautionary message. It calls for awareness of our inclinations to exaggerate the predicted seriousness and duration of future psychological states. In that light we explain how all of the actors involved in medical decisions are susceptible to affective forecasting distortions and warn of the danger that such biased thinking introduces into deliberations. We go on to call for critical evaluation of the grounds for our fears and to suggest that we pause to assess evidence and determine whether or not the fears that are moving us are, in fact, well grounded.
Gligorov is correct to note that the literature on affective forecasting describes “a first-person phenomenon.” Our paper discusses the first-person phenomenon of how patients’ biased judgment infects the decisions that they make for themselves. Gligorov rightly notes, however, that we go beyond the evidence when we ascribe the phenomenon to family members, clinicians, and policymakers who are making decisions based on how they expect others to feel. Again, the decisionmaking patterns that we identify seem to support our presumption that the phenomenon is more general. Yet, without evidence that judgments about how others will feel in the future is biased in the same way as judgments about how we ourselves will feel, this remains an untested hypothesis. If nothing more, our discussion could be taken as an argument that this might be a fruitful direction for further psychological investigation.
Tied to her point about affective forecasting research's focus on the first person is her claim about the phenomenon being “incorrigible.” By this, Gligorov means that regardless of the disconfirmation that we would note if we examined how well we recover from emotionally traumatic experiences, we fail to notice our bias and continue to make the same sort of distorted judgments about our future emotional states. Gligorov expresses hope that third persons (e.g., clinicians, family members, and policymakers) will be able to overcome the incorrigibility of the bias when making decisions about how to treat others, even though people cannot do it for themselves.
We are more hopeful about the prospects for everyone. In a recent New York Times interview, Daniel T. Gilbert, a prominent researcher on affective forecasting, offered that he now uses the findings of his research to guide his choices. For example, Gilbert explained that he now chooses to spend his money on experiences with others that are more likely to provide enduring pleasure rather than on purchasing things because they are less likely to provide enduring pleasure. If Gilbert can overcome the natural first-person bias, it may be less incorrigible than Gligorov supposes. And, as we suggest, armed with insight or aided by cognitive behavioral therapy, we all may be able to avoid making choices that we would not have made but for falling victim to systematic affective forecasting distortions.
Gligorov goes on to criticize us for not appreciating that the point of affective forecasting is “explanatory, and not predictive.” The theoretical point is well taken, and we do appreciate it. Nevertheless, when we note that affective forecasting distortions in medical decisions often amount to a systematic exaggeration of the affect of bad news, it is not a far stretch to appreciate that judgments informed by such bias incline people in a particular direction, in this case, against giving bad news. In our paper we offer several examples of what we take to be evidence of this phenomenon in decisions relating to medical communication and policy. Unless Gligorov can offer reasons to reject our interpretation or an alternative account to explain these patterns, affective forecasting bias seems to be a useful interpretation.
In the same line, she then criticizes us for failing to provide a general conception of happiness to guide health policy. Offering such direction would go well beyond the limited scope of our article, so we make no claims at all about what is needed for happiness. It is enough for us to agree with policymakers that avoiding likely, serious, and enduring unhappiness is a legitimate policy goal. We also inherently endorse broadly accepted views that happiness is in the eyes of the beholder and that most people (at least most U.S. citizens) treasure their liberty to pursue happiness as they see it. In that light, we caution policymakers that they must take care in their assessment that serious, likely, and enduring unhappiness will actually occur without their imposition of liberty-restricting policy. Such judgments should reflect hard evidence rather than rely on distorted presumptions. Our take-home message is that policymakers are irresponsible when they allow biased thinking to guide their imposition of liberty-limiting rules and guidelines and that by allowing distortions to infect their judgment they may actually do more harm than good.
Turning to her final objection, Gligorov argues that we are mistaken in concluding that affective forecasting biases impair decisional capacity and, therefore, justifies some measure of paternalistic correction (e.g., unrequested cognitive behavioral therapy) when there is good reason to conclude that the distortion is at work. Most authors who discuss decisional capacity list the abilities to “understand and appreciate” the relevant facts among its defining component abilities. As we see it, those in the grips of affective forecasting distortions are significantly impaired in their ability to “appreciate” the seriousness, likelihood, and duration of their painful emotional reaction to the anticipated event. When people grossly exaggerate some facts and dramatically discount other relevant factors that would be given much more significant weight but for the unrecognized psychological distortion, they fail to appreciate their situation. That means they lack decisional capacity. Gligorov interprets the “appreciate” requirement to simply mean “understands medical information about treatment alternatives.” Her assumption seems arbitrary and misses the point.
Consider an illustrative case. A man with a large, leaking aortic aneurism refused surgery and refused to provide reasons for his decision. He said that he wanted to live, and he understood that he could die if the aneurism should rupture. Pointing out the window of his hospital room he declared that a bomb could go off in Central Park, too. This man did appear to understand his predicament, that he could die without surgery. It is, however, hard to imagine that he appreciated the situation precisely because he seemed to think that the likelihood of dying from a bomb going off in Central Park was equivalent to his dying from the imminent rupture of his aorta. In sum, we see him as lacking decisional capacity because he fails to appreciate his circumstances, and we, therefore, conclude that paternalistically imposing the surgery is well justified.
Consider another case that appears to involve an affective forecasting bias. A woman was diagnosed with a malignant tumor in her jaw. Chemotherapy was likely to cure the cancer, but it was also likely to cause hair loss. She refused the treatment, claiming that she cared too much about her appearance. The patient could not envision that her loss of hair was transitory. As we see it, affective forecasting distortions magnified the imagined emotional pain related to hair loss and that singular focus prevented her from recognizing and appreciating how her appearance would be affected by the growing tumor and cost her life. The treating physician felt noble because he had respected his patient's choice. We see the physician's decision as a negligent failure to identify the patient's lack of sufficient capacity to make the decision at hand because of a gross and fatal misconception. Instead of seeing the patient's choice as autonomous and leaving her alone, we see her refusal of treatment as a conclusion infected by bias, something for which she could not be held responsible. Because of the seriousness of her decision and the likelihood of her judgment being the effect of distortion rather than enduring values, we would conclude that “soft paternalism,” in the form of cognitive behavioral therapy to correct the distortion, would be in order.
In sum, we remain convinced by the powerful conclusions of the studies on affective forecasting. We see the broad relevance of the findings, particularly in medicine. We, therefore, caution health professionals to be aware of the influence that affective forecasting can have on judgment and to use that awareness to protect their judgment and their patient's judgment from its distorting influence.
