Man bites dog. It is one of the oldest cliches in journalism, an acknowledgement of the idea that ordinary events are not newsworthy, whereas oddities, like a puppy-nibbling adult, deserve disproportionate coverage.

The rule is straightforward, but its implications are subtle. If journalists are encouraged to report extreme events, they guide both elite and public attitudes, leading many people, including experts, to feel like extreme events are more common than they actually are. By reporting on only the radically novel, the press can feed a popular illusion that the world is more terrible than it actually is.

Take finance, for example. Professional investors are fretting about the possibility of a massive stock-market crash, on par with 1987’s Black Monday. The statistical odds that such an event will occur within the next six months are about 1-in-60, according to historical data from 1929 to 1988. But when surveys between 1989 and 2015 asked investors to estimate the odds of such a crash in the coming months, the typical response was 1-in-10.

Perhaps professional investors are effectively paid by their clients to be doomsday fanatics. But the more straightforward explanation is that most of them massively overrate the odds of a financial cataclysm—by a factor of six.

A new paper by the economists William N. Goetzmann, Dasol Kim, and Robert J. Schiller blames this pessimism on the minds of news audiences. They point out that the press is more likely to report steep market plunges than moderate gains and losses. Media coverage of negative events—and frequent predictions that they will recur—leads many observers to think that crashes are often right around the corner, when in fact, they are almost always far away. These investors’ fear isn’t rational. It is warped by what researchers call the "availability bias," which gives more weight to information that is top-of-mind, which often begins with cable-news reports that are top-of-the-hour.

The implications of this man-bites-dog bias extend to Americans’ feelings of safety as well. Since September 11, 2001, fewer than 100 people have died in jihadist attacks in the United States, according to the New America Foundation, a think tank. That's about the same number of deaths from motor-vehicle accidents every day. But terrorism feels menacing and personal in a way that even a six-car-pileup does not, and so it receives disproportionate coverage. In December 2015, Americans named terrorism the country's most important problem.

This effect is more complicated when the press begins to report on dramatic events that it previously ignored. For example, media coverage of gun violence has increased in the last few years, particularly when it comes to police shootings of black men. This has created an impression that race relations and gun violence have never been worse. Donald Trump’s speech at the Republican National Convention harped on the escalating violence against police officers and ordinary Americans. The broader truth is that, under Reagan, police fatalities per year were about 60 percent higher than they are today, and homicide rates have been falling for two decades. White Americans are uniquely despondent about race; 32 percent of white Americans think Obama has made race relations worse, while just 5 percent of black Americans think so. One plausible explanation for this gap is that, because of social media and redoubled attention from mainstream news outlets, white Americans are finally seeing the institutional racism that black Americans have known their whole life.

As the futurist Ray Kurzweil put it, it's not that violence is getting so much more common in the U.S., but rather that the information about it is more accurate. It is altogether too easy to conflate primetime attention to violent episodes in America with the overall frequency of American violence.

The media’s emphasis on tragedy and calamity has no obvious solution. After an airplane crash or terrorist attack, should CNN be obligated to scroll, along its chyron, quotidian human activities that are statistically more dangerous than the threat of terrorism or planes? That’s hard to imagine. Should NPR’s Morning Edition begin every hour by reading the names of Americans who died in car accidents the previous day? Probably not.

This week, New York Magazine has a cover story on everything that's wrong with the media, according to members of the media. It's an exhaustive and thoughtful document of insidery self-loathing. But one overlooked detail is that the modern definition of news inherently prejudices journalists such that they focus their limited resources on uncommon stories. There is a strong public interest in reminding readers, viewers, and listeners of the dog-bites-man realities of daily life, as assumed and invisible as air. Sometimes, perhaps, the news ought to feel more like the olds.