Last week, New York Gov. Kathy Hochul announced a new policy that made headlines across the country: Following a handful of high-profile violent crimes, she’s sending 750 members of the National Guard and hundreds of state troopers into the city’s subway system.

“No one heading to their job or to visit family or go to a doctor appointment should worry that the person sitting next to them possesses a deadly weapon,” the governor said.

Fair enough. Major crime in the transit system is up 13 percent since the start of the year, according to the police, and post-pandemic, some New Yorkers have felt a fearful and “confused unease about the fact that the trains seemed different.”

But overall the picture is more positive: Crime fell in New York last year; and according to an analysis by the New York Times, the rate of violent crime on the city’s subway system was roughly one per 1 million rides — meaning your chances of being a victim of violent crime on the subway are really low.

There are other reasons a New York Democrat might take an approach to crime that seems aggressively out of proportion: The idea of the city as a crime-ridden hellhole is a perennial of right-wing politics.

Last fall, former President Donald Trump tore into New York’s attorney general for pursuing a tax fraud case against him, claiming falsely that the case was happening “while MURDERS & VIOLENT CRIME HIT UNIMAGINABLE RECORDS!”

In a big city, there’s almost always going to be a recent example critics can point to to say violence is out of control.

So how should we judge how well a city is doing at fighting violent crime?

A novel way to evaluate crime levels

To start trying to unpack that question, let’s look at one major element of violent crime: gun violence.

One way is to judge overall violent crime and per capita crime rates to see how the city is performing compared to past years; that’s what Hochul and the city’s mayor are pointing to when saying their subway deployments are necessary.

But another good way would be to look at how much gun violence there is in a given city compared to how well you’d expect it to do for a city of its size.

And on the gun homicide front, a new study shows, New York City is majorly overperforming. In fact, it’s performing better than any other big city in the country.

That’s one finding of an innovative new study by Rayan Succar and Maurizio Porfiri, the director of the Center for Urban Science and Progress (CUSP) at New York University’s Tandon School of Engineering.

You can read the full study in the journal Nature Cities to learn more about their methodology, but to sum up what makes their research unique: They used urban scaling theory — a form of analysis that has only been around for about 10 years and that has primarily been used to research things like wealth distribution and population growth — and applied it to crime.

They looked at nearly 1,000 US cities, studying a number of different relationships between gun access, crime, and population, and aggregated multiple data sets in the six years leading up to the pandemic. Their modeling allowed them to compare the actual prevalence of gun violence in a given city to how the model predicted a city of its size would behave.

One of their major findings is that gun homicides scale superlinearly to the population in cities — in other words, the bigger the city, the larger the number of gun crimes per capita.

In more rural areas, on the other hand, there are more guns but fewer gun homicides. So while cities have fewer guns per capita, Porfiri says, “they are responsible for more violence than what would happen in a rural area.”

There are some theories as to why this is, and most have to do with increases in social interaction. As Succar puts it, “If you interact with 200 people per day, there’s way more possibility you’ll get shot than if you’re interacting with five people per day.”

Maybe that’s good news for people who believe that they’re safest in a rural area far from big cities with a large stash of firearms (though the research on the risks posed to people who keep firearms in their homes might want to have a word).

But it isn’t the entire story.

New York has fewer gun homicides per capita than expected

The model also predicted that a city of New York’s population size should have way more gun homicides per capita than it does. In fact, of all the big cities they studied, New York had the largest gap between what the model predicted the gun homicide rate per capita would be and what it was.

“New York should be applauded,” Porfiri says. Given how large the city is, it’s outperforming expectations on an important measure of gun violence. (The authors don’t delve into the reasons why, but it’s likely a complex mix of culture, law enforcement, and policy.)

Is this (admittedly nuanced) finding going to convince Republicans who are certain the city is uniquely crime-plagued because of its purportedly soft-on-crime leaders? Probably not.

But the findings matter. Across the country, year after year, cities struggle with crime, and those problems often get spun into political narratives that have little relationship to the facts.

Getting a better understanding of how gun crime concentrates in cities — and distinguishing how a city performs given that reality — is an important development.

This story appeared originally in Today, Explained, Vox’s flagship daily newsletter. Sign up here for future editions.