As Gov. Martin O'Malley and Baltimore Mayor Stephanie Rawlings-Blake have battled in recent weeks over crime-fighting strategies, both have come up with diagrams to illustrate their positions.
O'Malley makes his argument with dueling trend lines that show crime declines leveled off as arrests dropped since his time at City Hall. Rawlings-Blake submits that arrests alone aren't the key to stopping violence, citing a chart that shows shootings and homicides at lower levels now than during O'Malley's tenure.
But experts say neither politician has presented a compelling visual argument, adding that crime also follows national trends and the causes of declines are far from clear.
O'Malley plotted his two trends on vastly different scales in a way that some say could distort the relationship between enforcement and violence. And criminal justice observers say Rawlings-Blake chose metrics that don't tell the whole story.
Fewer than half as many people are arrested in Baltimore now than under O'Malley, who advocated a zero-tolerance style of policing.
His display plots the number of arrests and incidents of violent crime from 2000 to 2013. It shows annual arrests rising to a peak of about 110,000 in 2003, then dropping.
The arrest rate continued to decline under the mayoral administrations of Sheila Dixon and Rawlings-Blake, with about 50,000 arrests made annually now. Meanwhile, violent crimes have declined in almost every year, though the rate has slowed. The governor has argued that the lower arrests are hurting the city's ability to deter crime.
Criminologists who reviewed O'Malley's chart agreed that it was hard to draw scientific conclusions from it because the reasons for crime's rising and falling are complex and result from many factors.
"I see crime dropping here year after year," said Richard Rosenfeld, a criminology professor at the University of Missouri in St. Louis and former president of the American Society of Criminology. "Case not made."
Rawlings-Blake argues that a zero-tolerance approach alienates community members who would otherwise be willing to help police identify violent offenders. The number of arrests that aren't pursued by prosecutors has also dropped significantly during Rawlings-Blake's time as mayor.
Her staff released its own chart plotting homicides and nonfatal shootings during the same 14-year period. That chart shows homicides and shootings rising and falling during O'Malley's tenure, with homicides actually higher when he left office in 2006 than when took the job six years earlier.
But looking at just homicides and shootings, as Rawlings-Blake did, does not give a full picture of a city's violence, said Bill Bales, a Florida State University criminology professor. The violent-crime rate is a "more accurate representation of what's occurring in the communities," he said.
"Violent crime" is defined as homicides, rapes, robberies and aggravated assaults.
Bales added that O'Malley's method of examining incidents of violent crime, rather than looking at the violent-crime rate, which is adjusted for population size, is also inadvisable.
Baltimore's population declined by thousands of people in that period before increasing slightly last year, which may have played a role in the incidents of violent crime falling, he said.
He noted that crime has generally declined across the country in the past decade, and criminologists cannot settle on why.
"There's a myriad of possible reasons why crime at this stage in our history is declining in general," he said. O'Malley's chart is "overly simplistic," he said.
High levels of arrests can play a role in reducing violent crime, but not always, said Rosenfeld, who added that other strategies, like a heavy police presence in crime "hot spots," are more effective.
O'Malley's use of crime statistics has drawn fire from critics in the past. The governor has said violent crime dropped by 40 percent in Baltimore during his time as mayor, which would be the second-largest drop of violent crime for large American cities.
Rivals said O'Malley manipulated the statistics by ordering an audit of crime figures that revised those numbers upward for 1999, and then did not do a similar audit in his later years at City Hall.
The governor noted on Twitter last week that The Baltimore Sun's editorial board did not run his chart with an opinion piece O'Malley wrote to defend his position. In a blog post, editorial page editor Andrew A. Green wrote that the chart was inconclusive and might prove confusing to readers.
"For one thing," Green wrote, "The scale used for violent crime is not the same as the scale for arrests, but the way the two lines are overlaid leads to the incorrect impression that they cross at various points in the last 14 years."
The scale for arrests runs from 40,000 to 120,000 on the chart, while the violent crime scale covers 6,000 to 18,000.