Fuzzy Facts Can Make Crime Rankings Suspect

by Carl Bialik

Wall Street Journal
January 28, 2012

Homicide might be the most concrete of crimes. In most cases, if there was a homicide, there's a body. That makes it the preferred benchmark for experts delving into the fuzzy world of crime statistics to compare cities or track trends.

But, as two recent rankings demonstrate, even homicide figures are subject to interpretation and uncertainty.

This month, a Mexican advocacy group released a ranking of the 50 cities around the world with the highest homicide rates. Forty of the cities on the list were in Latin America, but some of the most violent hot spots in Africa and Middle East were left off for lack of data. And critics say many of the underlying numbers originate with local police, which have varying levels of commitment to accurate reporting.

The underlying numbers are more solid, criminologists say, for the news that homicide has fallen out of the 15 leading causes of death in the U.S., according to a federal report, also out this month. But they question how much this slight shift in the rankings reflects a drop in society's level of violence, as opposed to demographic trends. Also, the homicide counts are preliminary and might be revised slightly for several reasons.

If interpreting homicide rates is this tricky, what does that say about overall crime rates, which experts say are more vulnerable to underreporting and subjectivity? Those issues are part of the reason the Federal Bureau of Investigation warns efforts to use its U.S. crime statistics to rank cities or states.

More

See also