B'ham leaders dispute "10th Most Dangerous" ranking

Birmingham Mayor Bernard Kincaid described the survey as "unchecked propaganda."

Kincaid said figures from the Birmingham Police Department show the city had an across-the-board reduction in crime, including homicides, based on 2004 statistics. ...

Barry Copeland, executive vice president of the Birmingham Regional Chamber of Commerce...said he, too, is suspicious of the survey's methodology, which includes giving equal weight to the six basic crime categories reported: homicide, rape, robbery, aggravated assault, burglary and motor-vehicle theft.

While it's an overstatement to call this survey "unchecked propaganda," it is important to interpret the results properly.

First of all, the methodology used was objective and straightforward:

First, 2004 city and metro area crime rates per 100,000 population (the most recent comparable final numbers available, released by the FBI in October 2005) for six basic crime categories — murder, rape, robbery, aggravated assault, burglary and motor vehicle theft — were plugged into a formula that measured how a particular city or metro area compared to the national average for a given crime category.

The outcome of this equation was then multiplied by a weight assigned to each of the six crime categories. Each of the six crimes was given equal weight. By weighting each crime equally, cities are compared based purely on their crime rates and how they stack up to the national average for a particular crime category. These weighted numbers then were added together for a city or metro area’s final score.

Finally, these scores were ranked from lowest to highest to determine which cities and metropolitan areas were safest and most dangerous.

Mayor Kincaid's contention that Birmingham's crime rate went down in 2004 certainly indicates progress, but it is irrelevant to what was actually measured; i.e., how each city's crime rate compares to the national average.

Nationwide, crime has declined substantially over the past decade or so. In Baltimore, for example, it is down 40% since 1999. Nonetheless, that city is still ranked as the 6th most dangerous in the country. Crime is down in Camden, New Jersey, as well, but it still earned the #1 ranking.

All of that illustrates the big problem with rankings like this. By only showing relative statistics, they obscure the underlying data, which may be more meaningful. In this case, we know that crime rates have improved in Birmingham and other cities on the "most dangerous" list in recent years, but that fact doesn't show up explicitly in the rankings. For example, an individual city may move up or down in rank from one year to the next, even if its own crime rate remains unchanged. Likewise, every city on the list could have better crime statistics than it had the previous year, while their placement in the overall rankings remains the same. In the end, somebody has to be "best" and someone else has to be "worst." Whether you're looking at college football polls or comparisons of poverty rates, rankings may be useful and interesting, but they never paint the complete picture. That seems pretty obvious, but it's a point that's often lost among the media and politicians.