Blog

Don’t Hate the Cities, Hate the Rankings

By Louis Bach, Communications

City crime rankings imply distributions of crime like illustration C, with uniform crime rates within a city that change suddenly at a city border. Crime occurs more frequently in patterns like illustration D, with crime focused around a hot spot and diminishing with distance. Source: NIJ

In October, D3 responded to Forbes’ annual rankings of “America’s Most Dangerous Cities,” criticizing their methodology and reporting. Last week, The Atlantic Cities published its own response to city crime rankings. Theirs dealt not with Forbes, but with the annual report published by CQ Press:

What exactly is wrong with City Crime Rankings? “Where does one begin?” asks criminologist Richard Rosenfeld, a longtime critic of the publication. “The fundamental problem with the rankings is that they are supposed to inform individuals about their chances of becoming victims of crime, yet they tell you almost nothing about a person’s risk for crime. A person’s age, gender, their activities—all those help tell you a person’s risk for crime.” …

The rankings also fail to account for the different configurations of American cities, what criminologists call the denominator problem. For instance, St. Louis, which topped the list in 2010, has much of its crime concentrated in inner-city neighborhoods, as do cities like Indianapolis, Louisville, and Memphis. But these cities have either annexed or merged with many surrounding, safer communities and are physically several times larger than the city of St. Louis; as a result they have significantly diluted crime rates within their city borders. Camden, New Jersey, which took second place last year as well, fell to the 225th most dangerous in 2010 when metro areas were considered.

This piece adds important points to the dialogue surrounding crime rankings, particularly that comparing crime of relatively aggregated geographies, like cities, loses the level of detail that makes crime analysis useful. Crime tends to cluster around hot spots, not spread evenly across cities. This is as true in Detroit as it is anywhere else; certain neighborhoods in Detroit are necessarily safer than others. You wouldn’t know it, though, from the rankings that give readers the impression that Detroit is uniformly unsafe (and therefore implicitly suggest that they avoid the city altogether.)

That distinction is, as The Atlantic Cities notes, lost on CQ Press, which stated in a press release:

The FBI, the Conference of Mayors, American Society of Criminology and various police chiefs — notably of cities that received “dangerous” rankings — say such rankings are “simplistic.” They argue that crime levels are affected by many different factors, including population density, economic conditions and education and drop-out levels.

We agree, of course, that crime-ranking information contains many variables and that all must be considered carefully. But as journalists, we take very seriously our responsibility to keep Americans informed — even if the news is not good. So we publish such data, even if it causes cities and officials to feel aggrieved.

While crime has many causes, its results are simple to understand. Reporting on cities’ crime rates, and presenting the results in understandable formats, is an elementary journalistic exercise.

The incidence of crime — is it growing, or dropping? — is of real concern to millions of Americans, and the rankings help them better understand what is happening in their communities.

CQ Press implies that the controversy surrounding crime ranking is the result of city officials who would prefer that Americans remain in the dark about crime rates; they claim that crime is “simple to understand” and “elementary” to report; and they claim that these rankings help to inform decision-making by citizens.

None of those claims are true. The controversy about crime rankings stems from the fact that comparing crime rates between cities obfuscates, rather than illuminates, the complex reality of crime. Crime, like so many other social ills, is complicated not just in its causes but also in its effects, regardless of CQ Press’ belief that crime is “simple.” As Rosenfeld notes for The Atlantic Cities, informing people about crime in a useful way involves expressing not just the aggregated presence of crime but also the risk factors for types of crime, something that rankings like those published by CQ Press entirely fail to do.

Rather than “help[ing Americans] better understand what is happening in their communities,” the rankings published by Forbes, CQ Press, and others hinder Americans from being truly informed about the state of crime in their neighborhoods by lumping safe and unsafe neighborhoods together. CQ Press is correct that “the incidence of crime… is a real concern to millions of Americans,” but they are cynically exploiting that concern by simplistically spreading fear, uncertainty, and doubt about major American cities, regardless of whether any given area is genuinely safe.