Safety in Numbers?

Pittsburgh is a safe city. Pittsburgh is a safe region. This has long been the case, and the latest data on crime indicate that the shoe still fits.

These four statements need to be tempered, however. This is because with public safety what matters is not regional or municipal crime rates as much as whatis going on in your neighborhood, at the house around the corner, on the street in the next block or over at the high school. Unfortunately, the Uniform Crime Reports of the Federal Bureau of Investigation do not provide that data. They track only crime in municipalities. They also aggregate municipal data to provide additional crime reports for Metropolitan Statistical Areas and the nation as a whole.

This means that the citizen does not get the crime data he most needs, but only the crime data the federal government is able to provide. Average rates for a particular crime per 100,000 population over the past 12 months is the basic building block of all federal crime reports.

The alternative is the daily TV and newspaper story, which by its nature is anecdotal and episodic. A furthercomplication: Because local police departments are the sources that the FBI uses to assemble its reports, datafromoutside major cities are uneven. Many communities, particularly in the Pittsburgh region, have no local police department or an under-funded police department (see Figure 1). This affects the number of reported crimes from year to year. It also suggests that regional comparisons, while valid, should be qualified under particular circumstances.

The Pittsburgh Regional Indicators project has attempted to minimize these pitfalls by concentrating on three major crimes: murder, robbery and burglary (see Figures 2, 3, and 4). As Carnegie Mellon University’s Alfred Blumstein points out in his background paper on www.pittsburghtoday.org, “we chose these [crimes] because they are seen as serious and are well-defined offenses (in contrast, for example, to aggravated assault which requires subjective judgment to distinguish ‘aggravated’ from ‘simple’) and they are reasonably well reported (in contrast for example, to forcible rape, which is poorly reported and whose reporting rate could fluctuate from year to year or from jurisdiction to jurisdiction).”

It is notable that only two regions, Minneapolis and Boston, have a lower murder rate than Pittsburgh; that only one region, Denver, has a lower robbery rate and one region, Boston, has a lower burglary rate. To repeat, those are not one-year rates but average rates for the years 1999 to 2006.

The numbers for the city of Pittsburgh are not as low as the regional rates, but they are well below the benchmark city medians for all three crimes: Denver, Boston, Charlotte, Minneapolis and Indianapolis in that order have lower murder rates than Pittsburgh. Denver, Indianapolis, Boston and Charlotte in that order have a lower robbery rate than Pittsburgh. Philadelphia and Boston in that order have lowerburglary rates than Pittsburgh. All other benchmark cities have higher rates for the three crimes.

Figure 5, assembled by Blumstein’s assistant Hee Jun, is particularly helpful in providing perspective. It presents the rates for each of the three crimes for both regions and cities. It also contains a “ratio” column that reports how much higher the city rate for each crime is than the regional rate. This ratio provides insight about the core cities in benchmark regions, because, as Blumstein points out, “one important feature of the geographic distribution of crime is that most crime occurs in‘hot spots,’ typically neighborhoods of significant social disadvantage, including single-parent households and those with low income, high unemployment, high mobility of residents and high housing vacancy rates.”

Consider two cities at the extreme. In St. Louis, the urban/non-urban disparities for all three crimes are higher than all other benchmark cities, while in Charlotte, they are the lowest. The housing segregation index on the regional indicator site provides one possible explanation for this disparity. St. Louis is the most segregated benchmark city with the exception of Milwaukee. On the other hand, Milwaukee’s urban/non-urban ratios, which are low for all three crimes, suggest that it is dangerous to generalize too much on the basis of measures like the segregation index. The fact that poverty rates in the Pittsburgh region are the highest of all benchmark regions also suggests that crime rates cannot be explained by financial deprivation alone.

And this is where we came in. Pittsburgh is a safe city. Pittsburghis a safe region. But because what matters most when it comes to crime are events very close to home, do not jump to any conclusions. With crime, everything is relative.