A number of American colleges and universities have decided to bring students back to campus this fall, believing they can diminish the risk of coronavirus transmission if everyone wears masks, uses hand sanitizer and social distances. Some schools also plan to reconfigure dorms to create family-sized clusters of uninfected students, who could socialize in relative safety, if only with their suite mates.

These plans are so unrealistically optimistic that they border on delusional and could lead to outbreaks of Covid-19 among students, faculty and staff.

My skepticism about the strategies under consideration is not based on videos of college students frolicking on Florida’s beaches when they were explicitly told to avoid large gatherings. Rather, it comes from more than 40 years teaching and researching young people.

Most types of risky behavior — reckless driving, criminal activity, fighting, unsafe sex and binge drinking, to name just a few — peak during the late teens and early 20s. Moreover, interventions designed to diminish risk-taking in this age group, such as attempts to squelch binge drinking on campus, have an underwhelming track record. There is little reason to think that the approaches proposed to mitigate transmission of the coronavirus among college students will fare any better. A series of studies that compare the ways in which young people and adults think and make decisions about risk-taking confirms this.

The late-adolescent peak in risky behavior has been found pretty much around the world. Although risky behavior is more common in some countries than others, the heightened risk-taking characteristic of adolescents, relative to adults, is more or less universal. My colleagues and I recently completed a study of more than 5,000 people between the ages of 10 and 30 from 11 different countries (including both Western and non-Western ones). Respondents answered a series of questions about the extent to which they had engaged in various types of risk-taking. Consistent with large-scale epidemiological studies, we found a peak in risk-taking somewhere between age 20 and 24 in virtually every country.

Our team has also conducted experiments in which we test participants on various risk-taking tasks under controlled conditions, which allows us to rule out any age differences in real-world risk-taking that might be caused by environmental factors, such as opportunity or cultural norms. As in our survey studies, risk-taking peaked during adolescence. Other studies, using different samples, have reached similar conclusions.

We’ve also conducted a series of experiments designed to identify just what it is about college-age individuals that accounts for their relatively greater propensity to take risks. Three factors appear to be most important.

First, this is the age at which we are most sensitive and responsive to the potential rewards of a risky choice, relative to the potential costs. College-age people are just as good as their elders at perceiving these benefits and dangers, but compared with older people, those who are college-aged give more weight to the potential gains. They are especially drawn to short-term rewards.

Second, college-aged people have more trouble exercising self-control than do those in their late 20s and beyond, an age difference that is amplified when people are emotionally aroused. Under calm conditions, college-age individuals can control their impulses as well as their elders, but when they are emotionally aroused, they evince the poor self-control of teenagers.

Author Bio: Dr. Steinberg is a professor of psychology at Temple University and the author of “Age of Opportunity: Lessons From the New Science of Adolescence.”