Violence in the Heartland, Part III: City Trends

In earlier posts (here and here), I have explored state-level violence trends since 1960 in the seven midwestern states of Iowa, Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin.  This post focuses on the data from the largest city of each of these states.  Since Chicago does not report its rape numbers in conformity with FBI standards, it is omitted from the analysis.

Here are the city trends since 1985 (reported violent crimes per 100,000 residents):

city data

What stands out most is the very wide, persistent gap between Detroit at the top of the chart and Des Moines at the bottom.  

In 1985, Detroit had 4.02 times as much violent crime per capita as Des Moines.  Twenty-seven years later, in 2012, Detroit had 4.03 times as much violent crime as Des Moines.  Although both cities had some ups and downs in between, the gap remained remarkably consistent.

Most volatile has been Indianapolis, which cut its violent crime rate by more than half in the late 1990s.  However, this decrease was associated with changes in the city’s crime-reporting practices, which may account for much of the apparent drop.  Likewise, a big increase in Milwaukee in the mid-1980s was associated with a change in the reporting of aggravated assaults.  Seeing the apparent impact of changes in reporting practices serves as a reminder that violent-crime statistics are not as objective and reliable as one might hope (a subject I discussed at greater length in my last post).

Also notable is a spike in violent crime in the mid-2000s in all of the cities except Columbus.  Depending on the city, crime started going up between 2003 and 2005 and peaked by 2008.  In most cities, these increases offset most or all of the decreases that had been achieved in the 1990s.  Nationally, there is a well-known story of urban violence spiking in the late 1980s and early 1990s, typically attributed to the fast and chaotic growth of trafficking in crack cocaine, but I’m not sure what was behind this pronounced regional trend of increased violence in the mid-2000s.  Curiously, the spike seems to straddle the economic collapse of 2007, and so does not seem connected in any straightforward way to economic trends.

Or maybe there wasn’t really a spike after all.  Detroit changed its reporting practices in 2005, Indianapolis in 2002, Des Moines in 2005, and Milwaukee in 2005.  Again, inconsistencies in reporting over time require caution in making claims about trends.

It is interesting to compare the city data with the statewide data:

states for city post

This illustrates the convergence phenomenon I described in the first post in this series: the highest-crime states in 1985 (Michigan and Ohio) are down sharply, while the lowest-crime states (Iowa and Wisconsin) are now somewhat higher.  This chart makes the trends in Detroit appear all the more troubling: not only has Detroit failed to make up much ground relative to its regional peers over the past quarter-century, but it is also falling further and further behind the rest of its state.  While Michigan’s rate of violent crime is now thirty-eight percent lower than in 1985, Detroit’s is only eleven percent lower.

To be sure, all of the cities have higher rates of violent crime than their host states.  However, the ratios vary from city to city and over time.  Here are the trends, tracking for each state the ratio of the violent crime rate in the largest city to that of the state as a whole:

city divided by state

Des Moines and Columbus have consistently come the closest to mirroring the overall state rate of violent crime.  Until recently, Minneapolis normally had the biggest disparity, often by a large margin.  Now, however, thanks to the long, sharp drop in violent crime in Michigan outside of Detroit, the Motor City “leads” in this dubious category, with per capita violence almost quintuple that of the state as a whole.

Cross posted at Marquette Law School Faculty Blog.