Competing studies show radically different numbers than the police study of themselves do. Mayor says the numbers are just right, members of the public have data that shows 65.6 officers could be cut. This argument has been going on since last April or May as you will see from competing memos below. Plus, I had my own additional thoughts yesterday based on serving on the 2003 police staffing committee.
This is the memo from the Mayor, which says, in part: (See full memo for charts and graphsPolice-Related Data to Council — 1-5-17)
Date: January 5, 2017
To: Council President Mike Verveer
Council Pro Tem Marsha Rummel
Council MembersFrom: Mayor Soglin
Subject: Police-related Data
At last week’s Board of Estimates (BOE) meeting, I asked Chief Koval about workload issues, the need for the Midtown Station, and the fact that crime in the city has declined. Chief Koval and Madison Police Department (MPD) command staff generally responded with anecdotes about not returning to the past regarding patrol patterns from a central office, the need to have patrol officers close to their office space in regional districts, and the benefit of community space in a new police station. One effort that’s grounded in data is the MPD staffing study that is currently being finalized by staff. This study will set us on a path to better and continuous improvement in integrating data into decision-making on MPD workload and resource allocation strategies toward achieving meaningful and measurable community-wide outcomes for all Madison residents.
In the interim, I wanted to share some information that I asked be developed which compares Madison with other communities on police officers per 10,000 population, crime rates and median income levels (see graphs below). The overall conclusion – there is no ideal staffing level that is mirrored in every community around the country. Cities much larger than Madison have the same police officer to population ratio as we do. Crime rate is also not a determinant – some with lower crime rates have larger police forces and vice versa. FBI data on officers per 1,000 residents are averages that have fluctuated over time. When you study the data, the conclusion drawn is that we are right about where we should be. We have made, and will continue to make, investments in our public safety services.
Our city is growing about 1% per year. At that rate, it will take us nearly 20 years to reach 300,000 residents. Will we need to add more police staff as we grow? Yes. Do we need to leap from 1.9 to 2.0 officers per thousand residents? No. Are we in a dire situation? Certainly not. Is there an emergency that requires the Midtown Station to be built in 2017 rather than 2018? No. To argue that there is an emergency ignores reason and relies on emotion and fear.
We are a City that has a world class research university – a place with knowledgeable and thoughtful people who understand the importance of reasonable discussion using evidence, data and logic. We have received accolades for being prudent with our resources – finding approaches that balance quality of life with effective and efficient financial decisions. A narrowly focused approach on police staffing and facility needs is inconsistent with what has made our City great. I will continue to emphasize a data-driven culture, continuous quality improvement, a focus on achieving meaningful and measurable outcomes and a thoughtful prioritization of resources consistent with these principles. More police officers alone will not improve educational outcomes, create affordable housing, improve equity in employment or ensure the opportunity for upward mobility for all Madison residents. Must we ensure a safe community? Most certainly…and let’s use data and rational discussion to arrive at reasonable decisions on ensuring all our goals are met.
I appreciate your thoughtful consideration of this information.
Other analysis done shows that we could cut 65.6 officers. See here for a memo (with graphs) from Dr. Gregory Gelembiuk after he attended a Public Safety Review Committee where the police talked about how many officers they needed.
This past Tuesday I attended a Public Safety Review Committee meeting at which Madison Police Department administration advocated for a large increase in staffing, primarily using the claim that the current ratio of officers to residents for Madison is below the national average for comparable cities. Though such an assertion may initially sound plausible, closer examination reveals that it is entirely spurious. Madison, with a police officer to population ratio of 1.87 officers per 1000 residents, already has a substantially higher number of officers than is customary for a city with its population size and crime rate. The Madison Police Department would need to downsize by about 65.6 officers to meet the national average for comparable cities.
I’ll start by pointing out that, among contemporary policing professionals, officer to population ratios are viewed as an inappropriate basis for staffing decision. For example, as noted by the International Association of Chiefs of Police (IACP): “The IACP recognizes that ready-made, universally applicable patrol staffing standards do not exist. As has been stated, the ratio of police officers to population often has been used as a measure of staffing levels. Because of the many variables involved the use of ratios, such as officers-per-thousand in population, has been found to be inappropriate as a basis for staffing decisions and are not included in the IACP methodology.”
I’ll next note that, to make their argument, MPD administration compared the officer to population ratio of Madison to the average for cities with population sizes greater than 250,000 (under the argument that Madison population is increasing to excess of 250,000). However, many U.S. cities with more than 250,000 residents are far larger than Madison (the average U.S. city with >250,000 residents is three times larger than Madison). And most have far higher rates of violent crime (the average U.S. city with >250,000 residents has a violent crime rate that’s twice as high as Madison’s). Of cities with more than 250,000 residents, 69 out of 80 have higher violent crime rates than Madison and only 11 out of 80 have lower violent crime rates than Madison. High violent crime rates necessitate more officers. So a comparison of Madison’s officer to population ratio to the simple mean across all such large cities is not appropriate and highly misleading.
At the PSRC meeting, Alder Cheeks astutely noted that he would like to see staffing ratios for other cities (such as Madison) that are regarded as among the safest in the country. That’s a critical point. Nationally, across cities, the two biggest determinants of the number of officers is the city population size and the city violent crime rate. Here’s a plot showing, for all municipalities with a population size greater than 1000, violent crimes per 1000 residents and police officers per 1000 residents. As you would expect, officer to population ratio correlates with crime rate.
and
Given that city population size and violent crime rate are both generally recognized as the primary drivers of officer to population ratio (as demonstrated in the plots above), I performed multiple regression to account for both these variables. The resulting model: Z = 1.068 + X*3.283*10-7 + Y*0.1194 ,where Z = officers per 1000 residents, X = population size, and Y = violent crimes per 1000 residents. (I’ll note that using log transformed data produced a virtually identical regression outcome).
For a city with Madison’s population size and violent crime rate, the model predicts 1.597 officers per 1000 residents (i.e. this is what would be customary for a city with Madison’s characteristics). Madison actually has 1.87 officers per 1000. I.e. The number of officers employed by MPD is already 17.07% larger (65.6 officers higher) than comparable cities. And looking at data for actual individual cities tends to bear out that Madison falls on the high side (though not outside the normal range).
I’ll also note that this doesn’t even take into account officers with the University of Wisconsin Police Department (62 officers) as well as the Wisconsin State Capitol Police Department, which further increase the effective police officer to population ratio in Madison. Most cities don’t contain such additional independent police departments patrolling residents.
In examining police staffing needs, I think it’s also important to recognize that the U.S. is in the midst of an unprecedented decline in crime levels. That’s underappreciated. Here’s a short video by the Christian Science Monitor, seeking to correct misperceptions about crime and safety, noting that “the U.S. is safer now than at any point in modern history.”
Here’s a graph from the Bureau of Justice Statistics National Crime Victimization Survey, showing the decline in the rate of violent crime. The cause of this dramatic decline is not adequately understood, though there’s considerable evidence that it may reflect the decline in childhood lead exposure (where the cognitive and behavioral consequences of lead exposure predispose to greater impulsivity and risk of involvement in violent crime).
and
Officer to population ratios rose in the 1980’s and 1990’s, largely in response to increases in violent crime during that period (with crime rates peaking nationally in approximately 1993). In 1980 there were 461,810 police officers nationally and by 2008 that number had risen to 765,246. This represents a 23.4% increase in the number of police officers relative to U.S. population size. However, more recently, given the declines in violent crime, officer to population ratios have been dropping. As discussed, for example, in this FiveThirtyEight.com article The Biggest U.S. Police Departments Are Getting Smaller:
“The number of police officers per person is falling nationally, along with reported crime. A report released Thursday by the Bureau of Justice Statistics (BJS) shows there were 15.1 police officers per 10,000 U.S. residents in 2013. That’s down from 15.4 police officers per 10,000 residents in 2007, the last time the BJS surveyed police departments around the country. The country’s biggest police departments are responsible for a big part of the decline, as big-city forces have shrunk even as the populations they serve have grown.”Contrary to what MPD administration is saying, the number of officers Madison has is already unusually high compared to the U.S. average for comparable cities, not too low. MPD wants to push it even higher, to 2.0 per 1000 people. In seeking to further increase officer to population ratios at this time, MPD appears to be out of sync with national trends across cities. Indeed, attention to officer to population ratios actually provides an argument for downsizing (to match norms for comparable cities).
In this letter, I’m not addressing the question of whether Madison should add officers, or downsize, per se. I’m merely pointing out that the primary argument MPD has provided so far, concerning officer to population ratios, is spurious. I’ll also note that police staffing is expensive. How many helping professionals (social workers, mental health workers, etc.) could serve the people of Madison for any given monetary allocation for police? In particular, I think proactive non-law-enforcement-based anti-violence and anti-crime programs with proven efficacy should be explored. For example, the Chicago Becoming a Man program (a cognitive behavior therapy intervention for at-risk youth) and the Richmond CA Office of Neighborhood Safety (a program that provides mentoring and a stipend to proven-risk youth) are two highly cost-effective, evidence-based programs with great outcomes that multiple cities are exploring adopting. Madison would do well to consider such approaches.
Assistant Chief Sue Williams responded to Dr. Gelembiuk and he responded to her if you want to dig in further.
All of this is a long way to show that the police analysis of themselves is questionable at best and their claims that we need 13 – 361 new officers is suspect and in fact, its possible we could cut police officers. If we do that, we could use that funding to fund other services like mental health and social workers – instead of having police perform those functions.
