Friday's tragic mass shooting at Sandy Hook Elementary is a harrowing reminder of the polar worlds of violent crime in this country. At the very same time that the murder rates in our cities are declining, mass killings, which tend to occur in the suburbs, are happening more often.
In an enlightening piece here on Cities on Monday, John Roman, an urban crime expert at the Urban Institute, took a detailed look at the declining homicide rates in America's largest 24 urban centers. After noting the caveats about lack of comparability of crime data between cities, he focused on data for "the one type of crime that is really hard to fake: homicide." Following a strategy originally introduced by The University of Chicago economist and Freakonomics co-author Steven Levitt, Roman arrayed these cities by the year their homicide rank peaked (about the early 1990s, when homicide rates peaked across the board in large cities), and compared this to their homicide rates in 2001 and 2010. In sync with previous research, he noted the overall decline in the homicide rate across all of these cities, and summarized Levitt's now-famous maxim: "the homicide decline was so consistent across cities that there was no need to think about local explanations."
But according to Roman, the nature of the change actually begs for a geographic explanation. On the one hand, a bunch of cities saw dramatic improvements in their homicide rates — New York, Washington, D.C., Los Angeles, Dallas, and Houston among them. On the other hand, there were cities that "are going in the wrong direction." While Las Vegas was the only city where the homicide rate actually increased, several cities moved up in the rankings — Columbus, Indianapolis, Milwaukee, Philadelphia, Baltimore, and Detroit.
After citing a variety of data he advanced the simple notion that for cities the "easy story" might well be: "Economic growth = crime reduction." But he adds that the story may not be as simple as it at first sounds, writing that we might be looking at "two crime declines."
I liked the piece so much that I immediately emailed Roman and we engaged in a productive dialogue. I agree that the declining homicide rate begs for more "local explanations" based on the experience of different cities and kinds of cities.
From where I sit, the real story is not the short-term rate of growth, but the type of growth and the factors that underpin it.
Houston, Dallas, and San Antonio have benefited from relatively stable housing markets, and in Houston's case, the energy boom. D.C. and New York have been at the leading edge of the urban shift in knowledge, profession, and creative work. Austin, San Jose, and San Francisco are among the nation's leaders in the share of college grads and of knowledge work. These knowledge-based, high-tech economies are underpinned by clustering and concentration of the sort that has fueled greater density and infill development. Many of these places — New York, L.A., San Francisco, and San Jose — have seen massive immigration as well, which has helped to stabilize and bring back disadvantaged neighborhoods and damp down violent crime.
The cities which have crept up the homicide rankings — Detroit, Baltimore, Philadelphia, and Milwaukee, as well Indianapolis and Columbus which have stronger economies — have seen far less re-urbanization, have had far lower rates of immigration, and in many cases continue to suffer from the classic "hole in the donut" syndrome. (It's worth noting that metros like Detroit and Milwaukee do very well on the Brookings MetroMonitor mainly because they have bounced back from the serious economic trauma brought on by the Great Recession, even though they lag behind metros like New York, D.C., or San Francisco in terms of income or wage levels).
My own recent look at the geographic variation in gun death and gun-related murders across cities and metros suggests a substantial part of the issue ties to the economic and geographic structure of urban economies. Although the data are spotty and while correlation does not imply causation, my analysis found gun-related homicides to be lower in metros with higher levels of human capital, more knowledge-based economies, and greater concentrations of high-tech industry. On the flip side, I found gun-related murders to be higher in metros with higher poverty levels, higher levels of inequality, more blue-collar working class economies, and higher shares of commuters who drive to work alone (a proxy for sprawl, among other factors).
Simply put, the ongoing urban transformation of the past two decades has had, and is having, a positive effect on murder rates. While it is not as simple as murder rates being a function of local economic growth, forces acting on economic structure and geography of cities — the clustering of the knowledge and creative economy, the ongoing shift toward urban tech, and continued immigration — have helped to reduce homicide rates. Part of the explanation may involve increased policing and security. And certainly there remain the thorny issues of housing affordability and displacement, not to mention growing geographic inequality and polarization. In our extended email exchange Roman pointed out that:
The other issue to note is that the forces at work here, including gentrification, have not had the unintended consequence of just pushing crime into close-in suburbs that were safe. I think the Washington, D.C. experience is pretty typical where gentrification pushed poor people into Prince George's County where crime has gone up some, but nowhere near as much as crime went down in DC. So, overall, it seems to be a net positive. I'm not clear on the causal mechanism(s), but there does seem to be something of a virtuous cycle at work. But there is a lot that still needs to be understood.
This kind of thinking is in fact in line with a wide body of urban research which suggests the growing divergence or bifurcation of cities and metros in terms of their human capital levels and overall economic performance. An article on Grist on Monday noted that:
Safer, healthier cities draw and keep new residents away from the unsustainable suburbs and exurbs. But while the numbers point to positive trends on the whole, they also reveal our sacrifice zones: Cities that have not been revitalized in this recent wave, where we have allowed poverty and violence to concentrate, out of sight and mind — cities that go unmentioned in the wake of mass murders like the one in Newtown, though they are actually our mass murder capitals.
I reached out to the distinguished criminologist, Alfred Blumstein of Carnegie Mellon University, who I learned a great deal from during my time there (he was the dean that hired me) for his perspective on the geography of gun-related homicides versus mass shootings. "The social [and] geographical distinction is certainly interesting," he writes to me in an email. "Much homicide is black-on-black, much inner-city, much gang-on-gang, unlicensed illegal guns, whereas the mass shootings have all been individual, white-on-white, mostly non-urban, often licensed guns (although often licensed to others as in [Sandy Hook])."
Roman, in his emails to me, writes that it's time to essentially flip the construct of concentration and clustering on its head. Instead of focusing just on "concentrated disadvantage" to describe crime, we should also focus on "concentrated prosperity." He adds that the main problem in making the connection between the concentration of urban prosperity and disadvantage is that our measures are far too broad. To do so would require more micro-level data at the neighborhood or Census tract level, something like a "micro place-based Gini coefficient."
I could not agree more. The concentrated nature of economic advantage is a central feature, if not the central feature, of the rapidly evolving urban knowledge economy. We are seeing the rise of an increasingly spiky world where prosperity and economic advantage are becoming increasingly uneven and concentrated at every single geographic scale — between as well as within cities and metros, globally and national, and across neighborhoods and communities.
We have tons of research that compares metros. What we need is better data and more research which looks inside our cities — at their increasing economic and geographic divides, tracing differences in income and wage levels, class and industry composition, density and other factors at a much more fine-grained scale.
These are the most pressing issues that face our increasingly urban world and ones that I will be focusing on a great deal more here on Cities in the coming year.
Top image: A girl places flowers at a memorial at a sign for Sandy Hook Elementary School in Sandy Hook, Connecticut on Dec. 15. (Eric Thayer/Reuters)