Police on Camera

Surveillance, Privacy, and Accountability

Bryce Clayton Newell

• Publisher: Routledge
• Year: 2020
• Print ISBN:
  9781138342439
• eBook ISBN:
  9780429800962
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

Police body-worn cameras (BWCs) are at the cutting edge of policing. They have sparked important conversations about the proper role and extent of police in society and about balancing security, oversight, accountability, privacy, and surveillance in our modern world. Police on Camera address the conceptual and empirical evidence surrounding the use of BWCs by police officers in societies around the globe, offering a variety of differing opinions from experts in the field.

The book provides the reader with conceptual and empirical analyses of the role and impact of police body-worn cameras in society. These analyses are complimented by invited commentaries designed to open up dialogue and generate debate on these important social issues. The book offers informed, critical commentary to the ongoing debates about the implications that BWCs have for society in various parts of the world, with special attention to issues of police accountability and discretion, privacy, and surveillance.

This book is designed to be accessible to a broad audience, and is targeted at scholars and students of surveillance, law and policy, and the police, as well as policymakers and others interested in how surveillance technologies are impacting our modern world and criminal justice institutions.

Section 1

Setting the stage: Theory and practice

Chapter 1

Taking off the blinders

A general framework to understand how bodycams work

Sander Flight

Just five years ago, when reporters would ask me: “Do body cameras work?” I would answer with an unequivocal “Yes”: the evidence was clear and seemed indisputable. Now, after five years of conducting evaluations of bodycams for the police, local law enforcement, and public transport in The Netherlands, and after studying over 200 books, articles, meta-evaluations, and reports on bodycams from all over the world, including chapters contained in this book, I’m not so certain. Even within the same jurisdiction, following the same departmental guidelines, and using the same technical equipment, bodycams can produce strikingly differing outcomes. So, my answer now is: “It depends,” which always prompts the next question: “On what?” In this contribution I attempt to answer that question.

I present a general framework that describes all the factors that influence the way in which body cameras “work.” The framework consists of three groups of variables that need to be considered in order to better understand how, where, when, and why bodycams have an impact and why they sometimes don’t. My aim is to support practitioners within law enforcement agencies who want to make optimal use of the latest academic insights, but find they are struggling to make sense of the sometimes contradictory findings from different studies.

The importance of context

The starting point

In the year 2015 we could all still dream about this new tool that improved policing and made everybody happy: politicians, press, public, privacy advocates—even the police were starting to embrace the technology. We could all be sure that bodycams “worked” because we had all seen the evidence from Rialto where bodycams significantly reduced the number of citizens’ complaints against the police. Because the findings were based on the “gold standard” of a randomized controlled trial in an experimental setting (Ariel, Farrar, and Sutherland 2015) there was no reason to doubt the cameras could not work their magic elsewhere. Two years later, everyone was forced to wake up from that dream when Forbes published an article entitled: “Why Don’t Police Body Cameras Work Like We Expected?” (Durkheimer 2017). Based on “the largest study to date,” an impact assessment of bodycams in Washington D.C., the conclusion was that we had to abandon the generally accepted wisdom on body camera effectiveness: body cameras did not reduce citizen complaints. Experts suggested that Washington was “an outlier” because even before the body cameras were introduced, this department went through a decade of federal oversight with effective reforms, making large reductions in complaints unlikely. We learned an important lesson: your starting point matters.

There is no holy grail

The starting point is not the only intervening variable that can change the outcomes of bodycams—there are many more. The first step towards a better understanding of bodycams is to accept the fact that this little piece of technology is actually a very complicated tool. We will never find one “crucial” variable that will be able to explain all variation in success or failure of bodycams. Accountability and discretion have been promising candidates for some time to play that central role, but even they could not have predicted that bodycams work in Rialto, but not in Washington D.C. The more we learn about bodycams, the more of these crucial and intervening variables are discovered. That is not surprising because bodycams may look like a simple tool from the outside, but they are supposed to influence the complex interactions between law enforcement officers and the public. Any effort to truly understand how bodycams impact on these interactions will have to combine insights from different academic disciplines and make use of a variety of quantitative and qualitative research methods. To paraphrase Jane Jacobs (1961), the science of bodycams has not yet broken with the specious comfort of oversimplifications and has not yet embarked upon the adventure of probing the real world.

Widen the lens

The vast majority of information on bodycams comes from the United States. An international comparative perspective, such as the one adopted in the succeeding chapters with contributions from Canada, Australia, and Uruguay, is the exception to the rule. The most recent overview of all available high-quality bodycam evaluations included 70 studies, of which 58 were from the United States (Lum et al. 2019).¹ This focus on just one country has consequences for the applicability of the findings in the rest of the world: in several important ways the United States is incomparable to other countries. One important difference is that in the studies that are done in the United States, complaints against the police and use-of-force by the police are often the most important, if not the only, outcome measures studied. This makes sense in a country where many police officers were forced to wear bodycams by court order, consent decree, or political pressure.² In many if not all other countries, the main reason to use bodycams is almost the opposite: to prevent violence against the police.³ This means that a sizeable chunk of what we know about bodycams is not very relevant in most countries outside of the United States, even if the studies are executed according to the highest possible academic standards. The political, social, and policing contexts influence the goals people hope to achieve with bodycams; they impact on how the bodycams “work” and they determine how we measure their effects. Even within the United States it is clear that bodycams can produce a wide range of outcomes. In the overview of 70 studies that was mentioned above, the findings were categorized into six groups of outcome measures: officer behaviors (32), officer attitudes (32), citizen behaviors (16), citizen/community attitudes (16), investigations (7), and organizational impacts (8).⁴ But outcomes are not the only way in which bodycam projects differ from each other: the picture is much more complicated than that.

The importance of mechanisms

Understanding what happens

As early as 2015, some experts were asking tough questions about bodycams. In his assessment of the five “state-of-the-art” evaluations of bodycams that were available at the time, Michael “mister bodycam” White posed an intriguing question. He pointed to the fact that a reduction in the number of complaints against the police, like the one that was found in Rialto, may seem like a positive outcome, but is not necessarily a good thing. If bodycams make police officers do a better job and act more professionally, this could lead to a reduction in the number of complaints and that’s good. If the bodycams prevent citizens from filing frivolous or vexatious complaints, the number of complaints decreases and that is good, too. But what if the bodycams make police officers avoid risky situations because they are afraid that the bodycam footage might reveal something that could harm their career? That would reduce the number of police–public contacts and thus the number of complaints against the police and that would be bad. In order to find out which of these three explanations is the right one, we will need to open up the “black box” of the experimental design. A randomized controlled trial (RCT) is the best research design if we want to go from correlation to causation. But experiments are unable to explain what caused the effect. We learned another important lesson: we need to understand what caused a certain outcome. Mechanisms matter.

Officers’ behavior matters

Based on experiments with bodycams in nearly a dozen police departments in different jurisdictions, the main conclusion was that “the use of BWC’s in police operations dramatically reduces the incidence of complaints lodged against police officers” (Ariel et al. 2016). This amounted to “a profound sea change in modern policing” or even “a turning point in policing.” The second conclusion was a little more complicated: bodycams produced positive, negative, or no results at all depending on three variables. The first “quintessential component” was the verbal warning given by officers: bodycams do not “work” if police officers do not give a verbal warning as soon as possible when engaging with members of the public. The second variable was the activation of the bodycam: “activation is paramount to the success of BWCs.” Different levels of adherence by officers to the protocol prescribing activation of the bodycam were able to explain all heterogeneity between sites: “the more officers can opt-out from mandatory activation procedures (and without consequences for deactivations), the less we should expect the BWCs to effect policing” (Ariel et al. 2016). The importance of compliance with department policy was confirmed in Phoenix. There, the number of complaints was reduced too, but “BWC activation was relatively limited. … If the officers would have complied with department policy there would have been an estimated 96% reduction in the number of complaints” (Hedberg, Katz, and Choate 2017). Finally, a third “most important” variable was introduced: the use of the recordings by law enforcement: “We need to pay attention to how cameras are used, bearing in mind that ‘used’ denotes not only the actual activation and recording of evidence but also, most importantly, whether and how the footage is then used by the law enforcement establishment” (Ariel et al. 2018). The lesson to take away from this is that the guidelines and the compliance to those guidelines by officers determines to a large extent whether bodycams will produce the intended results or make things worse. In short: policies matter.

Technology matters

Recently, scholarly interest for the design of the bodycams themselves has grown (Taylor and Lee, Chapter 5; Suss et al. 2018; Flight 2018). We know that giving a verbal warning can make a big difference: if people are not aware of the bodycam, they will not adjust their behavior. But that logically means that the design of the bodycams is important too. Anyone who has ever been on a ride-along with police officers trying out different types of bodycams will agree that different designs have an enormous impact on the visibility of the bodycam. Some bodycams are all but invisible because they are a black box on a black uniform. On the other end of the spectrum, there are highly visible bodycams with flashing lights or even outward facing screens showing what is recorded. In some areas of Germany, police officers wear large signs on the front and the back of their uniform informing passers-by of the presence of body cameras (see Flight 2018 for examples). A highly visible bodycam has more preventative capabilities than an invisible one. But a very noticeable design can also back-fire: the outward facing screens on the bodycam of German officers were so attractive to suspects that they wanted to take a look up close, even when the officer told them to keep their distance (Kersting et al. 2019). These bodycams sometimes led to escalations instead of the cooling-down effect that the bodycams—without screens—in Amsterdam had (Flight 2019). Despite the fact that these findings are nothing more than a confirmation of common sense, most academic publications on bodycams contain no description of the design of the bodycam. This needs to change, because technology matters.

The framework

Now imagine you are a police officer who has been tasked by her chief to decide whether bodycams should be introduced into the department. After studying the available evidence, you will soon discover how complicated that decision is. The political, social, and policing context matters, your starting point matters, mechanisms are important, policies guiding activation and discretion of officers can make or break the project, and you need to carefully consider the design of the bodycam. To help practitioners in this type of dilemma, I designed a general framework with three groups of variables which have been shown in one or more impact assessments to influence how bodycams work.

Figure 1.1 The Bodycam Framework: Three clusters of variables that determine how bodycams “work”

Technology

Bodycam

The device itself is an important influence on the effect of the bodycam on the police and public. Its design determines visibility, which impacts on the preventative effect, as has been discussed previously. But there are many more technical details with an influence on the mechanisms that can be triggered. Battery duration, for instance, determines whether some choices made in the policy are practically feasible or not. If the guidelines dictate that all interactions with the public have to be recorded, the battery will have to last an entire 10 to 12-hour shift. Some brands are able to deliver this, but for many bodycams—especially in cold outdoor working conditions—four to six hours will be the most you can squeeze out of one battery. This means officers will have to carry extra batteries or a mobile charging station in their vehicle or they will not be ab...

Table of contents

• Cover
• Half Title
• Series
• Title
• Copyright
• Contents
• List of illustrations
• List of contributors
• Preface
• Abbreviations
• Introduction: The ayes have it—Should they?: Police body-worn cameras
• SECTION 1 Setting the stage: Theory and practice
• SECTION 2 Accountability and discretion
• SECTION 3 Privacy and surveillance
• Conclusion: Body-worn cameras, surveillance, and police legitimacy
• Index