Crime rates have fluctuated quite a bit in the last few decades: high and rising through the 1990s; down, somewhat mysteriously, in the early years of the new century; increasing at an alarming rate around 2020; rising disturbingly in many cities in the several years since the pandemic abated, then declining again. But could crime be the main reason so many cities lost so much of their population over the entire period? That might seem a difficult case to make.
Now Henry Canaday, an economist and longtime business journalist, has come up with some compelling data that purports to show that the connection between crime and population loss may be more direct and more powerful than most experts have believed.
In a yet-to-be-published report, Crime & Population Changes in 21 U.S. Cities, 1960-2023, Canaday presents some numbers that seem almost too precise to be true. He divides cities into two categories: those that sustained a rate of less than 2 murders per year per 10,000 population and those that consistently experienced a higher rate. The ones below 2.0 nearly all gained people in the period he examined; those below 1.5 gained substantially. Cities with rates above 2.0 suffered population losses. This holds true, Canaday reports, even when other factors such as climate and economic change are taken into account. “Murder rates around 1 per 10,000, reminiscent of the 1950s, are apparently comfortable for city dwellers,” Canaday writes. “But once that rate gets much above and stays above 2 per 10,000 people start to leave.”
The numbers remain consistent for the last several years as well as for the past half-century, even though the murder rates nationally are considerably lower now than they were in the 1990s. St. Louis had a murder rate of 7.3 per 10,000 residents from 2019 to 2023 and lost 6 percent of its population. New Orleans lost 5 percent of its people amid a murder rate of 6.0. Meanwhile, of large cities that held murder rates below 2.0 in the last four years, more than three-quarters increased or maintained their populations.
I can’t vouch for every single decimal in Canaday’s detailed report of more than 130 pages. But the overall conclusion does carry a ring of plausibility. It rings true not only for the past several years but for the entire period of more than 60 years that he has set out to study.
It would be foolish to attribute Detroit’s long-term population loss solely to its rising crime rate, but the connection is difficult to avoid. The city has lost vast numbers of residents with the shrinking number of jobs in the automobile industry, but it began its greatest decline in the 1960s, just at the time when the murder rate was moving alarmingly upward. Its murder rate per 10,000 residents was well under 1.0 in the 1950s; by 1975 it was up above 5.0. In those years the population was falling by about 15 percent per decade. Only in the past several years has the city begun to see the murder rate go down, but as recently as last year, at a moment when the city was enjoying a downtown renaissance, it was still close to 4.0 per 10,000 Detroiters.
Baltimore has lost nearly 40 percent of its population since 1960; during most of those years it experienced an uninterrupted rise in its murder rate, reaching 6.0 per 10,000 in 2020 before beginning a modest recent improvement.
AS STRIKING AS THE NEGATIVE NUMBERS from these cities is, so is the trajectory of cities that have kept the murder rate under control and managed to gain in population or at least stabilize it. Boston lost about 30 percent of its population between 1950 and 1980 but began growing again in the 1980s, boosted by its emergence as a center of high-tech industry but also by its relative safety. Boston’s murder rate has never risen above 2.0, and it has stood at about half that level for most of the past three decades. Is this a matter of cause and effect? That’s impossible to prove. Canaday believes it may be.
Pittsburgh lost nearly half its population over half a century, but has been able to keep its numbers relatively stable in the past decade. It has consistently kept its murder rate below Canaday’s 2.0 trigger during the entire period. Some of Pittsburgh’s return to stability is undoubtedly a function of its ability to rebuild its economy on the twin pillars of higher education and medical research. But keeping the murder rate down is part of the story.
The clearest successes are in the South and West. Rapidly growing San Antonio has never had a murder rate as high as 2.0 per 10,000 residents. San Diego has held its rate to an enviable 1.0, and its population numbers have soared.
Chicago and Philadelphia come down somewhere in the middle. Philadelphia has lost a quarter of its population since 1960, but it has kept its numbers relatively stable in the past decade as its murder rate has generally stayed around the 2.0 threshold. Chicago has remained below 3.0 in the past 10 years; it has avoided severe population losses during that period.
Then there is New York. It has had the lowest murder rate of any large American city, never even reaching 1.0 after the late 1990s. It has dealt with population losses, but began growing in the 1980s and has seen its numbers drop only in the last three years, a result Canaday attributes to increases in minor crime, since the murder rate has remained extremely low.
THERE’S NO WAY TO DEAL WITH THIS SUBJECT without getting into the delicate topic of race. Canaday doesn’t avoid it, and I won’t. But the numbers don’t support some of the notions we may be holding in our heads. Populations have been declining the most in cities with large Black populations, but this isn’t a simple matter of white residents fleeing in fear of African American violence. Much of the population decline has resulted from middle-class Black residents leaving for suburbs or for other parts of the country. Chicago suffered modest population losses in the last decade mainly because of Black departures. The white numbers actually increased.
And population changes and murder rates don’t necessarily track with numbers of Black city dwellers. Boston and New York have managed to keep violence down with substantial Black populations; Baltimore, Detroit, New Orleans and St. Louis have not. There’s no simple takeaway on this one.
What Canaday does contend is that in the past several years, murder rates have correlated with law enforcement staffing. Baltimore, he reports, has a police force about 8 percent smaller than the one it had in 2019; it also recorded a murder rate, most recently, of 5.6 per 10,000 residents. Buffalo has increased its police staffing by about 3 percent during the same period, while holding its murder rate at between 1.5 and 2.6. A coincidence? Canaday doesn’t think so.
One caution that I would impose on all of these numbers: Population gain often correlates with urban success, but this is another complex relationship. Pittsburgh has managed a significant economic revival during a long period of population decline. No doubt it would have preferred to be growing rather than shrinking, but the shrinkage didn’t turn out to be ruinous. Phoenix has grown enormously in recent decades, but faces a myriad of environmental, climate-related and water-supply problems that it has been unable to solve. All of these are multifaceted issues.
What is impossible to deny is that the United States as a whole is looking at murder rates that are extremely high for a first-world nation. Between 2018 and 2021, U.S. murder rates increased by more than 30 percent. Germany’s rates rose only 4 percent. Murder rates in the United Kingdom went down.
Nothing in Canaday’s report compels any specific solution to urban crime and population flight, except perhaps more assertive policing. But he does lay out the problems very specifically in 132 pages of numbers and correlations. Understanding the situation, even its discouraging elements, is a good way to start.
Related Articles
