Note: The following article is reprinted by permission of POLITICO LLC, and originally appeared on June 2, 2020.

___________________________________________________________________________

I Was the Mayor of Minneapolis and I Know Our Cops Have a Problem

Racism permeated the culture of the department. But there are ways to change that culture that other cities can copy.

By R.T. Rybak

The searing images from the past several nights of anger and violence in dozens of cities across the country have shocked and horrified the nation. But there is one image that we need to keep fixed in our minds, the one that started it all:

A human being, staring calmly off into the middle distance, while his knee slowly suffocates another human being.

Our repulsion should boil over as we realize that the white police officer, who took an oath to protect and serve that person on the ground, who is black, would not have acted so brutally if the man he was restraining were white. Until every one of us can see that image for what it is—an example of a two-tiered justice system that treats black and white people differently—we cannot move another inch forward. We need to acknowledge that on some level, every one of us had a role in keeping this inequity in place.

I’ll go first, because after living in Minneapolis all my life, covering the Minneapolis Police Department as a Minneapolis Star Tribune reporter and, more directly, serving 12 years as mayor of this city, I should.

My own efforts to change a police department and its culture failed badly. That starts with appointing three different police chiefs who all made change but not enough. It includes attempts to diversify the force, to change practices in mental health and numerous efforts to work with individual officers on softening their approach so they could empathize more deeply with community. These failures will haunt me for the rest of my life, and it should. As each of us sees and acknowledges our own part it can be paralyzing. It was for me.

But I was heartened by something a colleague at the Minneapolis Foundation said to me the other day. Chanda Smith Baker grew up and raised a family as an African American in north Minneapolis, and for years has lead the Pillsbury United Communities. She has seen so many more of the consequences of our deep, endemic racism than I ever will. But as we surveyed the damage and pain in our community she said simply: “We have no choice but to act.”

So we are acting. Our foundation, which has been centered on racial equity for decades, is granting $1 million in the next few weeks to community-based solutions that strive for justice and healing in the aftermath of George Floyd’s death in police custody. Knowing we need to have tough conversations about race and culture, we launched our “Conversations with Chanda” podcast that will give our community, which has avoided those tough conversations for too long, the space to “go there.”

Like everyone in this city, we know that is still not enough. A very well-intentioned friend asked me what one thing he could do to make this situation better. I had to say, “There’s no one thing.” You can’t fully stop racism in policing without understanding the racism in the laws we ask our police to enforce, the racism in a criminal justice system that over-incarcerates black men, the racism in how we white Americans perceive a threat when we see someone who is black. An unjust economic system matters, and so does the issue where I focus most these days: the intolerable racial inequities in education. So does the classism that allows so many of us with privilege to have someone else’s child put on a police uniform and walk into tough situations so we can safely, mindlessly go about our lives.

But, right now, nothing matters more in Minneapolis than reforming the city’s police. An obvious first step would be to demilitarize the department. As a mayor who took office right after 9/11, I quickly saw that the community-based preventive programs like Bill Clinton’s “cops on the streets” initiative lost funding while we seemingly had a blank check for equipment and weapon systems that too often have the officers we want to “protect and serve” separated from their communities by shields and armored turtle suits.

Fortunately, we don’t need to invent a solution from scratch. We already have the Obama administration’s “21st Century Policing Plan,” which lays out in detail how our country’s police departments can be rebuilt around six pillars: building trust and legitimacy, policy and oversight, technology and social media, community policing and crime reduction, training and education and officer safety and wellness.

One of the most important values I took from that plan is something I learned on a deeply personal level as a mayor: Police officers are human beings. We then train them, put them with others we have trained into cultures that develop around the job and expect them to perform in the most high-stress situations imaginable.

We also know a lot about what makes that human being performing as a police officer thrive in the job or become a headline from a searing incident we could have prevented. The Center for Data Science and Public Policy at the University of Chicago has studied officer conduct over time in major departments and analyzed what actions signal when behavior starts to go off track. This helps us act more quickly when we need to intervene so that officers can be retrained or treated, and get back on track.

When I first saw this research, I realized that if, as mayor, the police chief and I, and the department’s supervisors, had known early when our officers needed our help and attention, we could have saved tens of millions in settlements costs and scores of lives. The problem was we never had the technology or tools to connect in real time what was happening with each officer and we didn’t have access to what we now know about how to step in.

That’s why I joined the founding board of Benchmark Analytics, which is now using that work in 60 cities and the state of New Jersey to connect department internal personnel systems to that deep research so mayors and chiefs can do what I never could to prevent the next tragic incident.

There are many more specific actions that can be taken but above all we need to address police culture. I have never been a police officer, so my experience is limited to what I have seen as a reporter and mayor. But I have come to know so many officers and continue to struggle with how I can know so many truly committed people whose collective actions I don’t recognize. In my city, at least, we have a majority of officers who let a minority of officers create an us-vs.-them culture that over time dehumanizes the people and neighborhoods the officers are supposed to protect and serve. Throw race into this toxic mix and you end up with behavior that often has to be named for what it is: racism. It plays itself out when a knee stays on the neck of a human being treated like he’s not human.

Much has been written by people who know more than I about police culture, but I do know it can be reformed only from within. That means the majority of officers need to rise up and take control of their culture. To the many good officers I know exist, I say this: I know the consequences of being shunned by your co-workers, but I also know you know in your heart that George Floyd should not be dead. Your silence is deafening and this city, and this country, cannot move forward until we hear your voices.

There is good news. We have stood at this place before, in Minneapolis and across the country. Yes, this might seem like the beginning of a familiar and dispiriting cycle: a terrible incident, a few days of promises and then, as the attention fades, so does the hope of change. But I also know that this is not a predestined conclusion. Change is possible. I know because I have seen it before in this very city.

Forty-one years ago, I was a young crime reporter. Night after night, I covered a police department that had deep issues of trust with two communities: residents who were black, and residents who were gay.

All these years later, one of those groups has seen enormous change. The Minneapolis police, which back then routinely beat and humiliated gay residents, is now one of the most gay-friendly departments in the country with openly gay officers serving in every part of the force, including at one point, the role of chief. There was no one action that made that possible, instead, in thousands of interactions, that wall creating an us vs. them turned into a we because each group recognized we are human beings on the other side.

The fact that we have seen so much progress with gay residents and almost none with black residents says a lot about the perniciousness of racism. We need to own that. But it does also say that change is possible, and now we have to prove that is true.

Copyright 2016 POLITICO LLC.

What could you do, as a leader in law enforcement, with a truly holistic and preventative early intervention system? The proposed benefits are well-documented:

  • More insight into negative and positive behavior trends
  • Provide officers an opportunity to self-correct ahead of behavior necessitating disciplinary action
  • Increased police legitimacy and through that improved relationships with the community your agency serves.

Implementing such an EI system sounds like the obvious choice. Unfortunately, most EI systems don’t actually provide this level of insight because they rely on threshold-based systems and simple triggers.

This type of system doesn’t enable supervisors to be proactive when addressing off-track behavior. A research-driven EI system has arrived to meet this need. Knowing a solution exists, what else might prevent your agency from successfully implementing a 21st century early intervention program?

Challenges to Successful EI System Implementation

Despite being around for 30 years (at least in concept), EI systems still face some resistance within agencies. Little of it has to do with the outcome of the system as much as misunderstandings around how the system works and why it exists. The outcome is clear: prevent the need for discipline by addressing off-track behavior before it leads to adverse incidents.  So why do some agencies still struggle to fully realize the potential of EI systems?

Officer Perception

“It is imperative to communicate that the system does not focus on disciplining employees but on assisting supervisors and leadership in preventing disciplinary issues from arising and to improve the overall performance of all agency members.” 1

Because existing EI systems use inaccurate methods for evaluating risk, many officers associate early intervention with being flagged for doing their job. The EI system is not a mechanism for doling out disciplinary measures to your officers. Rather, it’s meant to help supervisors avoid situations where formal discipline is necessary. You need to help your agency understand EI systems from this perspective, highlighting this type of technology is intended to offer additional support to officers, ensuring they are equipped with the training and feedback to do their best possible job.

Uncertainty about Data

“It is also critical for employees to understand how the system works and the factors it considers.” 2

Law enforcement produces a significant amount of data just in the course of doing everyday police work. Sophisticated EI systems use this data to determine how likely an officer is to engage in off-track behavior. Some officers might perceive this as additional tracking or oversight, or worry that personal data is being used to evaluate their ability to do the job. However, this is far from the reality. For example, consider how data and analytics factors into professional sports. Athletes want their data to be captured and analyzed so they can improve their game to the nth degree. EI systems offer a sliver of that some opportunity for excellence. Also, having a clear policy in place delineating between what data is captured, what data is exempt, and how it’s used to power the system will also help alleviate uncertainty.

Inaccurate Data

“Another challenge arises when those responsible for entering data into an EI system are not consistent and diligent in doing so when only some supervisors enter data, the database does not become the robust system needed for an effective EI system.” 3

We’ve all heard the idiom, “garbage in, garbage out.” Because your EI system runs on data, its quality determines the actionability of the insights generated by the software’s analytics. An EIS typically draws on existing data sources like CAD and RMS platforms along with its own data. Ensure your agency is operating in such a way that accurate data becomes the standard operating procedure.

Supervisors Fail to Adopt the System

“Similarly, if supervisors and command personnel do not monitor the EI system, they miss opportunities for proactive intervention with at-risk employees.” 4

The guiding principle of an EI system is its proactivity. You want to provide supervisors with data points they need to intervene ahead of an officer having an adverse incident with a civilian. In fact, your officers should advocate for having supervisors monitor the system – in effect, they are using analytics to support their officers. Subtle changes in behavior can often be imperceptible to the person experiencing them. And supervisors can’t be with every officer all the time, close enough to observe changes in behavior that might indicate a problem. Having an EI system in place that supervisors readily adopt scales the ability of your supervisors to anticipate the needs of your officers. This will lead to better rapport between ranks, and ensure personnel aren’t left without the appropriate support.

For more information about the challenges of implementing EI systems, read the full COPS report, Law Enforcement Best Practices: Lessons Learned from the Field, here.

 

1,2,3,4
U.S. Department of Justice. 2019. Law Enforcement Best Practices: Lessons Learned from the Field. Washington, DC: Office of Community Oriented Policing Services

Even though Early Intervention Systems have existed in concept and practice since 1989, most agencies haven’t been able to implement a truly proactive instance of one. Historically, EI systems have been developed using thresholds and simple triggers to identify personnel exhibiting off-track behavior. What results is a necessarily retroactive method of supporting your officers, as the problematic behavior must occur before a supervisor receives an alert to address it. This simple system will always fail to realize the full potential of the Early Intervention programs.

In its latest report, Law Enforcement Best Practices: Lessons Learned from the Field, the Office of Community Oriented Policing Services (COPS) identified Early Intervention Systems as an important issue for law enforcement executives and personnel. Why now? Because computing power has finally caught up to what was required by the original vision for effective EI systems. And because more than ever we need to be providing officers with the right type of research-backed support at the right time.

What Type of Agency Uses an Early Intervention System?

According to the COPS report, “65% of police departments with over 250 officers had [an EI system] in place.” Because simple triggers were the norm until recently, we know these systems are less effective than they could be.

According to research conducted by our partners at the University of Chicago, EI systems using triggers produce 89% false negatives (i.e., officers who are likely to exhibit off-track behavior are not flagged by the EI system) and 71% false positives (i.e., the officers flagged by your system aren’t actually at-risk of off-track behavior).

Put simply, if you’re relying on triggers to flag at-risk officers, it’s unlikely you’ll ever succeed in identifying the right personnel at the right time. Trigger systems produce these results because they treat every officer – regardless of when or where they’re patrolling – in the same way. This doesn’t reflect the real world of policing.

Early Intervention as a Management Process

The COPS report defines an Early Intervention system as “a management process used in law enforcement agencies to monitor employee performance or behavior via administrative data.” The key word here is data. As mentioned, simple triggers won’t provide the insight you really need to have a meaningful impact in supporting your personnel. Without data, you are just taking a shot in the dark. New EI systems, like Benchmark’s First Sign®, take a data-driven approach. Meaning the system is designed to reflect the specifics of your agency’s data, allowing you and your supervisors to draw informed conclusions based on patterns of behavior and risk factors unavailable to those using simple triggers.

The report also emphasizes an EI System is “not designed to be punitive, but rather a proactive tool.” Again, this only becomes possible when you start with the data.

Anatomy of a Successful Early Intervention Program

If you’re starting to evaluate EI systems for your agency – or re-evaluating an existing one – you’ll find the COPS report valuable. In it, they lay out what makes for a successful EI program.

– It identifies personnel in need of support while also identifying the right support, intervention or training to put the officer back on track.
– It helps an agency better audit the types of training and support it currently has in place, surfacing data that enables supervisors to develop better interventions and provide tailored support.
– It necessitates policies and procedures to ensure personnel are trained to use it, thus ensuring a broad awareness of both the benefits of an EI system as well as the types of data factored into its analysis.
– It’s communicated effectively to the rank-and-file. This offsets any fear or apprehension about a system that could easily be perceived as disciplinary, though it’s critical to emphasize the proactive and non-disciplinary nature.
– It’s easily leveraged by First-Line Supervisors (another focus of this report), ensuring adoption and data fidelity to allow the system to provide valuable insights to stave off at-risk behavior in personnel.

The report is full of invaluable perspective for police executives. We’ll continue to share what we think are the most salient points for agencies looking to implement truly proactive, data-driven EI systems. In the meantime, you can read the full report here.

Most chiefs already know why a law enforcement agency would implement an Early Intervention System (EIS). Police early intervention systems are designed to help agency leadership identify officers who need additional support. All in an effort to prevent those individuals from having an adverse incident with a citizen or other personnel.

But what conditions result in an alert, and its accuracy in helping agency leadership effectively intervene, might not be as widely understood.

Traditionally, these systems use triggers tuned to flag an officer whose activity exceeds agency-decided thresholds. Critically, this type of system enables police executives to remain aware of potential personnel issues. However, that awareness comes with a caveat:

Researchers found that trigger-based EIS typically flag the wrong officers:

Visual representation of trigger-based EIS flagging the wrong officers.

 

In fact, studies found trigger-based solutions produce false negatives 89% of the time, and false positives 71% of the time. If, as studies have found, 17% of your officers will have an adverse incident within a year, can you afford to make decisions using such an inaccurate view of your agency?

How do agencies improve at identifying these officers and successfully intervening?

When we first start working with a partner agency, we help them identify where they’re at on the path to digital transformation.

Modern organizations evolve through several stages of technology adoption and integration

 

As you can see, there are five stages, beginning with Undefined and ending with Predictive. Traditional EIS can get you as far as the Analytic stage, but the nature of the software means you can’t effectively, and proactively, support your officers using triggers and thresholds.

Watch Nick Montgomery, Chief Research Officer at Benchmark Analytics, share the studies behind Benchmark’s research-based Early Intervention System. You’ll learn how to identify your agency’s current transformational stage and what you gain by evolving to a predictive approach.

You’re no doubt familiar with Early Intervention Systems (EIS) used by law enforcement agencies to identify off-track behavior in officers. However, in recent years, agencies have sought a more proactive and preventative solution that can identify officers before they’re involved in a career-damaging adverse incident.

They’ve found that little has changed in traditional EIS platforms since the initial days of trigger-based systems. How they look has barely kept up with other technology, and the way they work hasn’t been updated since the 70’s.

The emergence of modern problems leads many agencies to reevaluate their current EIS and consider investing in a modern one.

Old technology and new technology

Non-Disciplinary by Design

Far from being a means to discipline, EIS platforms were intended to be non-disciplinary by design. However, using triggers to identify officers in need of additional support has since been found to create situations where EIS platforms are used as hindsight-driven, punitive tools, if they’re used at all.

Unlike classic EIS platforms, a modern EIS is configured to capture and analyze indicators beyond the simple mechanism of triggers and thresholds. This enables them to provide insight that can be used to address an officer before they have an incident that could require disciplinary action. It also provides a greater depth of insight into an agency’s overall health, including officers deserving of recognition, and individuals who have been exposed to a critical number of stressful events.

An EIS is intended to be an objective, smart, and automated mechanism that supports the health, safety and efficacy of your officers. Still, the idea of trusting an algorithm to interpret officer data can be off-putting to those who prefer a human touch. However, a truly modern EIS blends seamlessly with your department, informing and amplifying the impact of your frontline supervisors rather than replacing them.

Similar to the communication and tactical improvement brought about by two-way radios, an EIS extends your supervising staff’s ability to respond to officers in need of support. Furthermore, in recent years agencies have been using EIS to identify officers worthy of recognition for exceptional service.

How Do Early Intervention Systems Work?

Initially, EIS platforms were called “Early Warning” systems. While “Intervention” and “Warning” are sometimes used interchangeably, the latter has gradually fallen out of favor. Consider this –a warning will tell supervisors something is wrong, but it won’t tell supervisors how to help the officer who triggered it.

According to the National Police Foundation, using intervention terminology “emphasizes the role of the agency in providing officers with support and resources to address problems at their earliest stage.” Most police executives would agree helping officers is preferable to punishing them.

The important part, regardless of what term you use, is that you are aware an officer is exhibiting behavior pre-determined to be atypical, enabling your supervisory team to take preventative actions that “promote officer safety, health and wellness, and success.”

old typewriter new computer

The first generation of EIS platforms relied heavily on the mechanics of triggers and thresholds. Agencies would select a set of indicators and then define thresholds for each one. For example, if an officer had three use-of-force instances in a single month, that would trigger an alert for a supervisor.

Though innovative at the time, recent advancements in technology allow for the development of sophisticated systems that refine triggers to allow for truly preventative action. A peer-reviewed study from our partners at the University of Chicago identified that most trigger-based systems result in a 78% false positive and 90% false negative rate.

These trigger-based systems can often orient the attention of supervisors in the wrong direction, wasting time and introducing the risk that officers in need of support will slip through the cracks. Beyond that, they often aren’t configured to provide an early enough warning.

Regardless, an EIS is an essential part of any modern agency’s personnel management toolkit for myriad reasons.

For more information about alternatives to a trigger-based approach, such as using analytics to convert data into insights, check out First Sign®.

Why would your agency want an EIS?

In a report on EIS best practices, the National Police Foundation defined one as “a personnel management tool designed to identify potential individual or group concerns at the earliest possible stage so that intervention and support can be offered in an effort to re-direct performance and behaviors toward organizational goals.”

In other words, an EIS can help you know where to focus your management efforts. Most companies and organizations have some type of support system or tool in place to position employees to succeed. Usually this falls under the purview of human resources. Agencies can apply some of these existing principles to how they think about supporting officers. EIS platforms are particularly well-suited to support agencies and their officers in the following ways.

Protect Your Officers (and Their Careers)

Law enforcement officers regularly must adapt to high-stress, complex (often unfamiliar) situations. They’re also expected to thoroughly document these events, including everything from when use of force is required to vehicle pursuits. This makes their profession unusual in that a lot of data is produced but not much is done with it to help personnel.

Early EIS platforms worked off indicators chosen through intuition. While this gets agencies part of the way towards a system that can help prevent officers from drifting into adverse behavior, these indicators alone simply aren’t enough to make sense of all the data generated by today’s LEOs.

A modern EIS allows supervisors and police executives to take truly preventative action, without getting bogged down in false positives and false negatives.

Accreditation Compliance

CALEA offers multiple tiers, but table stakes for accreditation include guidance on EIS usage. According to CALEA standard 35.1.9, agencies with an EIS must also have a “written directive” to provide structure around definitions of behavioral indicators, reviews of identified employees, remedial action, and so on. In CALEA’s own words, “the failure of an agency to develop a comprehensive system can lead to the erosion of public confidence in the agency’s ability to investigate itself, while putting the public and agency employees in greater risk of danger.”

Relationship with Community

A survey of nearly 2,000 residents of U.S. metro areas, “found that 75% of white respondents and 80% of Black and Hispanic respondents favored the use of early warning systems as an accountability mechanism within police agencies.” As communities continue to ask for more transparency and insight into how police agencies identify and address officers whose behavior does not align with their expectations, modern EIS platforms offer a solution that signals an agency is investing in not only the well-being of their officers, but the well-being of the community.

start here

6 Baseline Functions to Look for in an EIS

As you prepare to either reevaluate your existing early intervention system or consider purchasing one for your agency, here are some baseline functions you should seek.

Trigger or Threshold Mechanisms

Though research indicates these mechanisms are no longer enough (and can often mislead supervisors), they still provide some insight into the frequency of certain events. Though you might find yourself

A Research Base

While it’s great to gain insight into your officers, there’s only so much you can do comparing your agency against itself. Implementing an EIS built on top of a longitudinal research base helps you understand your officers in the context of policing across a variety of departments.

21st Century Analytics

Data without analytics isn’t very useful. If your EIS doesn’t come with advanced analytics, your team will be left to crunch the numbers on their own. And at the rate that law enforcement produces data, it’s nearly impossible for an individual to derive any meaningful insight without the help of modern technology.

Situational Evaluation

Your EIS should be able to interpret information based on situational data. This is critical to avoiding false positive and false negatives. If your EIS can’t process data in the context of a situation, the results won’t differentiate between a justifiably active officer and an off-track one.

Temporal Evaluation

When something occurred should also factor into the information you receive from your EIS. Whether a series of events happened during third watch, on the weekend, in the morning, or during a large event should factor into which officers are flagged, if any.

Command Channel Review Support

While all agencies have some review process, our experience has revealed that each one is unique. That’s why it’s key for your EIS to be flexible enough to align with your command channel review.

Choosing an EIS is an important decision for an agency. Consider seeking out a partner who understands the complexities of policing who can also leverage the power of advanced analytics.