From the Chart Room: Seattle’s Intended Police Defunding — Expected Side Effects

More rapes, robberies, and murders; less inequality from traffic stops


roposals to reform American policing are not new. Nor are critiques of the violence American police employ. Ubiquitous calls to “defund the police,” however, distinguish the movement catalyzed by the killing of George Floyd from previous surges of concern over  these issues. But what does “defund the police” entail as a policy proposal? Among advocates and activists, interpretations vary.

City governments, however, have now started to flesh out concrete policies in response to the demands of this new movement. Earlier this month, the city council of Seattle, Wash., voted to enact what one New York Times reporter described as one of the “most ambitious” overhauls of policing in America. The “budget amendments passed by the council,” a Seattle Times staff reporter noted, “are intended to shrink the force by up layoffs and attrition this year.” A reduction of 100 would represent a decrease of 7 percent in sworn law-enforcement officers employed at the Seattle Police Department.

Ever the dismal science, economics offers insights into what lies ahead for Seattle’s residents as the city moves down the path its council has chosen. However dismal arbitrary cutoffs may be to experience, they can create opportunities in an economist’s dataset to treat observations close to the cutoff threshold as if they randomly passed, or didn’t. This “quasi-randomness” can then permit economists to distinguish true underlying effects from misleading correlations. In President Obama’s 2009 fiscal-stimulus legislation, a scoring system with a cutoff rule determined how much federal funding a city received for its police force, generating this type of quasi-randomness. And from that, it was possible to derive a set of estimates about what the effects of increasing the size of a city police force on crime tend to be.

Today, that analysis offers a looking glass through which to view some aspects of Seattle’s future if its city council fully achieves its intended objective of shrinking its police force by up to 100 officers. The chart above shows the grim picture that comes into focus with respect to violent crime. The study indicates that each additional officer prevents 0.107 murders, 0.532 rapes, and 1.98 robberies in a given year; the estimate for the annual effect of a reduction of 100 officers is the multiple of these values by 100, and that’s the quantity shown in the chart. To estimate the effects of a police reduction of Seattle’s intended size for longer timeframes, multiply the shifts in annual crime rates displayed in the chart by as many years as in your time horizon. To derive an estimate for the cumulative effect on violent crime over five years, for instance, multiply by five.

These estimates certainly cast a shadow over Seattle’s initiative. But they do not necessarily eclipse the entirety of the policy per se. Evaluating a policy as a whole involves weighing its intended as well as its unintended consequences. And these increases in violent crime would, of course, be unintended.

For instance, new research from Steven Mello (the same author of this study on police size and crime) suggests that less policing is likely to reduce the extent to which police interactions perpetuate racial inequalities. According to that research, police tend to cut breaks on speeding tickets to white more often than to minority drivers, who then face higher potential fines for the same underlying traffic offense. Fewer police and therefore fewer traffic stops, then, would reduce the extent to which policing exacerbates inequalities. In the case of Seattle, such an effect would represent an intended benefit of the city council’s recent policy.

The safest conclusion from the existing evidence, then, is simply that criminal-justice policy in America entails difficult tradeoffs. And increases in violent crime, at least in principle, may represent costs that come with benefits sufficient to outweigh them. But even the soundest statistical analysis of societal costs and benefits cannot evade the gravity of what’s at stake. Regardless of whatever other benefits accompany a policy that increases murder, for instance, a murder victim is losing his only life.

We don’t yet know how these effects will play out in the case of Seattle. Only time will tell. But all indications are that any increases in crime may wind up, from the perspective of the policy’s architects, among the more predictable of its consequences. Already, within weeks of the city council’s vote, the constituents of Councilmember Lisa Herbold, who heads Seattle’s Public Safety Committee, have launched a campaign to recall her.

From the Chart Room: Seattle’s Intended Police Defunding — Expected Side Effects

Joseph W. Sullivan served from 2017 until 2019 as special adviser to the chairman and staff economist at the White House Council of Economic Advisers.