With the rise of the #MeToo movement, many are left wondering why so few victims of sexual harassment spoke up before. A psychological disconnect called “affective forecasting error” helps to explain this failure. Affective forecasting error is the disparity between how one believes they would react in a stressful situation compared to their actual response.

What does affective forecasting error have to do with harassment, especially sexual harassment? As it turns out, a lot. A recent article by Roseanna Sommers in Behavioral Scientist notes, “Our affective forecasting errors lead us to blame victims for failing to exit harassing situations, because we incorrectly believe that if it happened to us, we would have marched straight to HR.”

There’s good news, however. Developing an understanding about this fundamental psychological phenomenon, and a sensitivity to it, will help you better respond to harassment complaints and avoid liability.

Reporting Delays Could Provide Defenses

By now, all employers are—or should be—aware of the need to promptly investigate and address complaints of harassment. In cases of harassment by a supervisor that did not result in a tangible job detriment (such as a termination or demotion), the Supreme Court has held that an employer can avoid liability under certain circumstances. By showing that the employer exercised reasonable care to prevent harassment and promptly corrected any harassing behavior, and the complaining employee failed to take advantage of those preventative or corrective opportunities or otherwise avoid harm, employers can use the Faragher v. Boca Raton affirmative defense. Likewise, in most jurisdictions, an employer can avoid liability for co-worker harassment if it takes prompt, effective remedial action after it becomes aware of harassment.

How, then, should you handle complaints that are made weeks, months, or even years after the alleged harassment was said to have occurred? Some employers conclude (mistakenly in many instances) that the delay in reporting relieves them of any obligation to investigate, while many decide, consciously or not, that the delay undermines the credibility of the complainant. Consequently, they do not perform a thorough investigation.

Fortunately, not every failure to investigate will result in liability, as the law provides support for such conclusions in certain circumstances. The applicable statutes of limitations may bar claims based on conduct that took place many years prior. In one such case, Milligan- Grimstad v. Morgan Stanley, the plaintiff claimed that she had been harassed throughout her 11-year employment with the company. The 7th Circuit Court of Appeals found that, because the plaintiff had not shown a course of continuing harassment, her claim was untimely to the extent it was based only on the older allegations.

An employee’s delay in complaining may provide the employer with other defenses as well. For example, in Conatzer v. Medical Professional Building Services Corp., the 10th Circuit Court of Appeals held that an unexplained three-week delay in making a complaint “constitutes an unreasonable failure ‘to take advantage of any preventative or corrective opportunities provided by the employee or to avoid harm otherwise.’”

While these defenses could be useful, when considering the “reasonable” response, you should be aware that there may be hidden psychological forces at work, and those forces may cause an error in judgment that could lead to liability. This is where affective forecasting error comes into play.

People Might Not React The Way You—Or They—Would Expect

In the 1990s, psychologists Julie Woodzicka and Marianne LaFrance asked approximately 200 women how they would respond to certain discriminatory and sexually harassing questions in a job interview. The purposefully offensive questions included “Do you have a boyfriend?” “Do people find you desirable?” and “Do you think it’s important for women to wear bras to work?” The study, published in the Journal of Social Issues in 2001, showed that most of the women would be angry and claimed they would confront the interviewer.

In the next phase of the study, Woodzicka and LaFrance ran job advertisements and invited a number of women to participate in what they thought were real job interviews. The male interviewer asked the same discriminatory questions that were posed as hypotheticals to the women in the first phase. Surprisingly, few of the women appeared angry or confronted the interviewer. Even more surprising: none of them even reported him after the fact.

This is a cogent example of affective forecasting error. Stated simply, people are generally poor prognosticators when it comes to their own responses to future events. Moreover, they tend to be overconfident in their erroneous predictions—that is, they imagine they will be more assertive and confrontational than they typically are.

What Should You Do?

Why does it matter? Employers who ignore or discredit a harassment complaint based on the belief that the complainant should have acted differently may reach the wrong result for the wrong reason. Imagine that you ignored or discredited a substantial harassment complaint simply based on the belief that a “reasonable” employee would not have waited weeks, months, or even years to complain. If you end up in litigation, your defense likely will succeed or fail on that single issue. Almost invariably, you would be better off had you investigated and addressed the complaint as best you could under the circumstances. Moreover, it is simply good employee relations practice to take all complaints seriously.

This is not to suggest that every complaint about conduct occurring in the distant past should be treated the same as a complaint about recent conduct. Actions that allegedly occurred years ago, or even in recent months, may be difficult to investigate. Employees come and go, documents get lost or destroyed, and memories fade. And as noted above, the applicable statutes of limitations might preclude a viable legal claim if an employee complains about conduct that allegedly occurred years ago. There may be other considerations not covered here that could also impact your assessment of an older claim.

However, you should be mindful of judging the employee’s actual behavior based on a prediction of how you believe you would have behaved in similar circumstances. Stated simply, you should remove the following question and response from your corporate lexicon: “If it was so bad, why did she wait so long to report it? If I had been in her shoes, I wouldn’t have waited. I would have told the harasser to stop and then marched straight to HR.”