Lies, damned lies, and statistics: Politico's argument for reducing police spending is ... something else

Do we spend too much on policing? Probably true, although not for the reasons Politico offers in this narrow look at data around crime rates and policing. This attempts to give an answer to the question of “defunding the police,” a narrative which the media has forcefully morphed into cutting police spending. The only context provided for this argument is homicide rates, which assumes that homicide is the only crime for which policing is needed — already one bias in an approach that contains several.

The crux of this argument is that homicide rates have gone down while police spending has gone up over the last 25 years or so. Could that be a causal relationship? Perhaps, Politico allows, but it’s an uneven distribution if so:

A POLITICO analysis of city finance data shows police departments have grown dramatically in size since the mid-1990s, when federal funding first put thousands of new cops on the streets. During the same period, the national homicide rate dropped — a sign to many policymakers that spending on law and order works.

But over time, police spending has outpaced the corresponding decreases in crime. Since 1995, the year after the landmark national crime bill went into effect, the U.S. homicide rate has dropped by a third. Police spending per capita, however, has increased by 46 percent nationally — and in some cities, by far more.

Studies have shown that an increase in sworn police officers reduces instances of crime. However, increases in other factors — such as social welfare, access to health care, employment and other social services — have also been shown to decrease crime rates. It’s unclear the extent to which increases in police spending are responsible for falling rates of violent crime.

“The perception was that the police have a direct relationship with crime, so the more police … the lower the rate of crime, we thought. But that has not been the case for some time,” said Dr. Howard Henderson, founding director of the Texas Southern University Center for Justice Research. “There are other factors that are at play that affect that relationship beyond simply just the police’s presence.”

In the first place, why are we only discussing homicide rates? Police enforce a panoply of laws besides homicides — sexual assaults, property crimes, traffic laws, domestic violence incidents, battery, public drunkenness, and so on. In fact, one context that never does get mentioned is just how many more laws have been passed in that time requiring more enforcement by police. Between federal funding mandates and grassroots grievance campaigns, police likely have gotten a bigger job in this period of time, which would explain the increase in resources.

A better metric would be total arrests, but even that might be misleading. Those didn’t really start declining significantly until a decade or so ago, and the reporting on arrests may not be entirely reliable. Arrests still only cover a portion of what police do; how many calls do they get year-on-year? How many traffic stops do they conduct for things like, say, texting while driving — which didn’t exist in 1990?

Politico also leaves out broader context of increased demand for more governance overall. The Urban Institute analyzed police spending not “per capita,” as Politico did, but as the portion of overall state and local spending over the last 40 years. Police funding increased in almost exact proportion as overall government spending, they determined, even when adjusting for inflation:

From 1977 to 2017, state and local government spending on police increased from $42 billion to $115 billion (in 2017 inflation-adjusted dollars). However, as a percentage of direct general expenditures, police spending has remained consistently at just under 4 percent for the past 40 years.

That 273% increase raises a question about the impact of demands for bigger government, of course. It also raises a question about the efficiency of that spending. How many police officers do we get for the expanded investment? For instance, it costs the city of Minneapolis over $180,000 per police employee (not just sworn officers), which means they can afford fewer officers on the street. The Politico analysis assumes that more spending means more policing, when it might just mean more expensive policing.

In other words, we might well be spending too much on policing, but not because we have too much policing. The opposite may well be true: we may not be able to implement the successful model of community-based policing because we have too few officers, thanks to legacy contracts with police (and other public employee unions) that make it much more costly than 25 or 40 years ago. And who negotiated those contracts at the city and county level? Almost exclusively, it was the party of Bigger Government.

Let’s put it another way: Does anyone really feel that we have gotten 273% improvement from municipal services in state and local government in any sense over the last 40 years? If the answer to that is no — and it should be — then why limit this discussion to policing? And for that matter, why don’t voters in these jurisdictions hold the progressives accountable that ran all of these municipalities, including police departments, during that 40-year era?

Those are the questions we should be asking. It’s hardly a surprise that the media keeps actively avoiding them by torturously parsed statistical analysis like this.