KSHB recently reported that Kansas City homicides are down 22% compared with the five-year average, nearly two years after the launch of a coalition of city agencies and non-profits called SAVE KC. That is good news. It is not proof that SAVE KC caused the decline.
One of the most common mistakes in public policy is assuming that because one event follows another, the first event must have caused the second. Crime declines after a new program is launched, so the program gets credit. Crime rises after a policy change, so the policy gets blamed. Often, the evidence for either conclusion amounts to little more than timing.
The KSHB story quotes Jackson County Prosecutor Melesa Johnson, who said she believes SAVE KC is playing “a real role in the success that we are seeing.” Perhaps it is. The problem is that belief is not evidence.
Violence rises and falls for many reasons: gang conflicts, police deployment, prosecution decisions, demographics, economic conditions, and the churn of individual offenders. A before-and-after comparison cannot isolate any one cause.
That is why researchers do not determine whether a program works by simply comparing crime rates before and after implementation. They look for evidence that the intervention itself produced measurable changes that would not otherwise have occurred. Jackson County’s COMBAT program has long suffered from this same problem: public claims of success without rigorous evaluation.
To its credit, SAVE KC has been careful to acknowledge on its website that multiple factors influence violence trends; it does not claim sole responsibility for recent declines. That’s a welcome departure from what we’ve seen before. But public officials are already drawing connections between the program and declining violence. That may ultimately prove justified. But Kansas City has heard similar claims before.
The Kansas City No Violence Alliance (KC NoVA) offers a warning. KC NoVA was once praised as an innovative violent-crime strategy. But a U.S. Department of Justice review found no statistically significant effect on homicides, group-member homicides, or aggravated assaults after two years.
In 2014, city leaders were celebrating the lowest number of homicides since 1972. Public officials were quick to claim credit. “We’re making progress,” proclaimed then-Mayor Sly James, citing targeted police work, community engagement, and anti-crime initiatives for the decline. But after homicides continued to rise in subsequent years, Mayor James’s confidence disappeared.
The lesson is not that violence-reduction initiatives never work. The lesson is that confidence should follow evidence, not precede it.
Rather than asking whether a new program coincides with lower crime, reporters should ask what evidence exists that the program caused the decline. Has an independent evaluation been conducted? Are outcomes being measured against comparable groups? What metrics are being tracked? How will success be defined? What would constitute failure?
Lower homicide numbers are worth celebrating. But celebration is not evaluation. Before officials claim victory, and before reporters repeat the claim, Kansas City deserves evidence that the program worked.
