Introduction: the Founding Fathers

Tony Bouza was a champion of democracy and knowledge. He loved democracy for the same reason his parents moved his family from Spain to the USA: a hatred of the dictatorship of General Francisco Franco, who was born in the same city in Galicia as Tony Bouza (El Ferrol). Bouza loved knowledge because it was so freely available to him in New York City, to which he moved at the age of 10. He was 43 by the time I met him, by which time he had not only earned a BA and MPA from the City University of New York; he had also participated for years in the NY Public Library evening seminar program on “Great Books.’’Footnote 1 He was so taken with the great phrases of great writers that he could not resist reciting them at appropriate moments in his policing career.

Like Benjamin Franklin—another champion of both democracy and knowledge—Tony Bouza became a founding father of a new institution. Both Bouza and Franklin each lived in an age of revolution; both of them believed in rebuilding institutions. Franklin founded an Ivy League university, a fire department, a public library, and a learned society—not to mention a new nation that had democracy as its core principle. Franklin also worked with other leaders—the “founding fathers’’—to build the biggest new institution: the United States of America). The institution that Bouza’s generation of ‘’founding fathers’’ built was something they called ‘’professional’’ policing—the core element of which was knowledge. What Bouza and his colleagues—including NYPD Commissioner Patrick V. Murphy and FBI Director Clarence M. Kelley—did for policing was to promote scientific research as an indispensable tool for effective policing.

From the 1961 advent of experimental criminology in the Ford Foundation’s support for a randomized trial called the Manhattan Bail Project (Roberts, 2009), the idea of ‘’clinical’’ field testing of crime and justice policies infused the radical spirit of the 1960s. The Ford Foundation’s 1970 creation of the Police Foundation led to even more support for rigorous evidence on what works in policing.Footnote 2 It also created a national platform for Tony Bouza’s boss in the NYPD, Patrick V. Murphy, who promoted Bouza from Inspector to Assistant Chief before Murphy became President of the Police Foundation in 1973 The stated purpose of the Police Foundation--renamed the National Policing Institute-- was ‘’supporting innovation and improvement in policing.’’.

In that new role at the Police Foundation, Murphy oversaw the Kansas City Preventive Patrol Experiment, which found that neither increasing nor reducing police patrol presence had a measurable effect on crime in patrol beats (Kelling, Pate, Dieckman, and Brown, 1974). While a recent re-analysis of the Kansas City data suggested that more patrol did, in fact, reduce crime (Weisburd et al., 2023), the effect of the original experiment was to show that experiments in policing were both possible and important. It opened up a vast horizon of untested principles of policing, most of which had never been tested.

While many police chiefs were horrified by the Kansas City experiment, Bouza was delighted. He cited it repeatedly to show the importance of empirical research on policing—and to reject the idea that personal ‘’experience’’ is the only source of knowledge needed for good police practice. When he was appointed Minneapolis Police Chief in 1980, he immediately made his own police department available to the Police Foundation as a site for experimental research.

Reforming the NYPD with New Kinds of Evidence

My own relationship with Tony Bouza began before the Kansas City Preventive Patrol Experiment was in view, while Patrick Murphy was still NYPD Commissioner. I was a newly installed program analyst in Planning and Research when Murphy put Bouza in command. Our first project was to spend a large grant from the Police Foundation to hire six independent experts in various fields on 2-year contracts. We hired systems analysts, management scientists, a training expert, and a lawyer, after interviewing some 50 candidates. Bouza farmed them out across the NYPD to places in greatest need of help, including a major anti-corruption initiative. He pushed for solutions and implementation, some of which actually happened.

Murphy promoted Bouza again to an even bigger job: inspecting NYPD compliance with major regulations, from taking crime reports to complaints against police. I was transferred to the Inspections Division with him, which allowed me to see him using social science principles for data collection. His first step was sampling crime reports on thefts to re-interview the victims by phone. What many of them described was not theft (without violence), but actually robbery (with violence or threats). This proven ‘’downgrading’’ of offence seriousness was a longstanding and widespread practice, until Bouza started issuing reports aimed at precinct commanders with the highest detected rates of downgrades. Similar methods were used to telephone precinct stations and count the number of rings before the phone was answered—sometimes up to 25 rings and many minutes (with rotary phone ringers).

His next step was to send me out incognito to every police precinct in Brooklyn at slow times of day. My job was to ask the desk sergeant a simple question: “how do I file a complaint against a police officer?” In some half of the precincts, all hell broke loose upon my arrival. I was interrogated intensely about why I was asking, and rarely got an answer to my question. My reports were detailed summaries of the encounters, which provoked fury from several precinct commanders.

Yet the Commissioner strongly approved of this approach, which was later expanded to “wallet drops” observed covertly to see whether a wallet on a sidewalk left in the path of a foot patrol wound up registered in the police property room, or in the pocket of a beat officer. The results of these tests were widely publicized by Murphy, making the front page of the New York Times. All of this took place in a wider context of reforming police culture through evidence-based tracking, which led to a transformation of what had been a systematically corrupt police force (Sherman, 1978). It also led to the development of ‘’tracking’’ of police conduct as a major strand of the Triple-T of evidence-based policing: Targeting, Testing, and Tracking (Sherman, 2013).

By the time Tony Bouza was appointed Police Chief in Minneapolis, I had finished my Ph.D. and become Director of Research at the Police Foundation. On the day of his appointment, I was writing a grant proposal to the National Institute of Justice, the research arm of the US Department of Justice. The proposal responded to requests for experiments in criminal sanctions; my plan was to test arrests for misdemeanor domestic violence. While such arrests were not generally lawful for any misdemeanor across the USA unless officers had witnessed the crimes or had a warrant for the arrest, Minnesota had just changed the law in 1979. The law gave police powers to arrest—but only for domestic assaults—even if they had not seen the offense, as long as they had ‘’probable cause’’—the same evidentiary standard as for felonies.

Randomly Assigning Arrest: Minneapolis as a "Crime Laboratory'' 1980-88

Before Bouza even started packing for Minneapolis, I called to ask him if he could endorse the proposal. He had just been appointed by the new Mayor of Minneapolis, former Democratic Congressman Donald Fraser, who wanted to police to use the arrest powers of the new domestic violence law. Yet police were reluctant to do so, for many reasons. One reason was a sincere belief that arrest might backfire, and cause even more violence in the relationship. Victims’ advocates, who had lobbied for the law, firmly believed the opposite: that arrests would create a deterrent effect against domestic abuse. Tony Bouza agreed with them and with the Mayor. But he also knew how hard it would be to change police practice.

Meanwhile, the academic community believed we had no way of knowing what effect the arrests would have without empirical research. There is no such thing as a free lunch, nor a conclusion without a careful empirical test. A year later, the US Supreme Court’s Federal Judicial Center (1981) published the same conclusion—and encouraged the use of randomized controlled trials to test alternative ways of responding to crime. But even without that endorsement, Bouza went right to the Mayor to request his support for an experiment.

In the context of randomly assigning a new way of dealing with domestic abuse, rather than comparing existing methods, Bouza was able to present the idea to the City Council as an innovation. He did not put it as bluntly as saying “I want to arrest people by lottery.’’ What he did do was to offer the proposal as a way to take an incremental step towards using the new police arrest powers.

The City Council voted unanimously to approve the proposal—random assignment and all.

Bouza then called for volunteer police officers who would be willing to attend a 3-day planning meeting at an elegant corporate retreat on a lakeside out of Minneapolis, including overnight accommodations and good meals. A total of 41 officers attended the planning sessions—almost 10% of the patrol officers answering 911 calls. The discussions were robust but fruitful. By the end of day 3, the group had agreed to an experimental protocol. On St. Patrick’s Day, 1981, they met at Police Headquarters for a final review and went out on patrol to do their normal job as before—except in cases of misdemeanor domestic assault.

By 1983, University of California (Santa Barbara) Professor Richard Berk of the Police Foundation research team had analyzed the repeat offence reports 6 months after each case had been arrested (or not). The preliminary findings were provided to the New York Times, which published them in the Science Times section (Boffey, 1983). The pre-Google era news clipping service of the Police Foundation found that over 300 US papers had re-published the NY Times story around the USA, as well as major overseas papers, including The Times of London and the Sydney Morning Herald. A year later, the details were published in a major social science journal (Sherman & Berk, 1984a), with Bouza’s commentary on the experiment published in the Police Foundation’s publication of the full operational report for police agencies (Sherman & Berk, 1984b). This is what Bouza wrote:

“Police handling of chronic, thorny problems such as domestic violence cases usually has been characterized by seat-of-the-pants adoption of remedies thought to work. But little lay behind such cures except an untested belief in their efficacy. Domestic violence provided a fine example of the way police approached difficult problems. Clearly productive answers based on hard evidence were needed.

The Minneapolis domestic violence experiment not only provides new insights into the spouse assault problem and its solutions, but it highlights the general need for analysis, experimentation, and evaluation in law enforcement.” (Bouza, 1984).

One can also say, in retrospect, that what Bouza had led the Minneapolis Police to do was to lay the foundation for what has become the globally-discussed concept of evidence-based policing (Sherman, 1998).

Where Crime is Most Predictable

By 1985, Bouza accepted a voluntary appointment as Board Chair of the Washington-based Crime Control Institute, of which I was the founding President. With further grants from the National Institute of Justice, we were able to accelerate the pace of research in Minneapolis. One major project was to test the hypothesis that crime does not occur randomly in space, but is heavily concentrated in a small number of locations. Given the difficulties of analyzing data kept on large mainframe computers, few had been able to examine patterns of repeated crime at the same locations. Even more important, no one could tell where crime had not been reported, which turned out to be most of the city. Once again, Chief Bouza put his full institutional authority forward to allow the Crime Control Institute to download the mainframe data on 911 calls and dispatches onto an early personal computer.

When University of Maryland PhD student Patrick Gartin finished waiting many hours for the computers to transfer and analyze all of the data for1986, he reported the startling result: 64% of police calls came from just 5% of addresses and intersections in Minneapolis (Sherman et al., 1989). The New York Times (1987) report of this finding led police agencies around the world to start examining their own data to see if their crime was also concentrated in “hot spots.” By 2015, Professor David Weisburd (2015) had demonstrated that the pattern applied to large cities and small, in the Western and Middle Eastern worlds. Moreover, the pattern was stable, with most hot spots in Birmingham, UK, for example, remaining hot for 2 years or more (Weinborn et al., 2017).

But is it Most Preventable?

Bouza was not content to simply report the major discovery of extreme crime concentration in space; he was eager to do something with that discovery. In continued partnership with the Crime Control Institute, Minneapolis Police supported two new grants for research from the National Institute of Justice. One (with Rutgers University Professor David Weisburd as the co-investigator) was to test the deterrent effects of increasing patrol presence at crime hot spots. The other (with Rutgers PhD student Michael Buerger as the co-investigator) was to test the effects of problem-oriented policing (Goldstein, 1979) in two randomized trials: one for residential addresses (multiple dwellings) and one for commercial addresses.

The hot spots patrol experiment (Sherman and Weisburd 1995) found substantial benefits in reducing citizen-generated reports of crime and disorder, with a carefully measured increase in police time spent in hot spots from 7 to 15% of the high-crime time periods. This finding has since been largely replicated in over 80 repetitions of the experiment around the world (Braga et al., 2019). But the precedent for it would never have been set without Bouza’s leadership. Once again, the challenge was approval by the City Council.

In order to add patrol time to high-crime/high-disorder locations, Bouza knew that he had to reduce patrols in other areas. Doing that could not be accomplished sub rosa; it had to be transparent and legitimate. The best way to do that was to take the research design to the City Council, after speaking with Council members representing low-crime areas. Those areas had been patrolled for years, with almost no crime reported. But if patrols were removed, would crime then start to go up? No one knew. However, the question could be answered as the proposed experiment unfolded.

What Bouza proposed to the Council was that the research team would keep close track of the burglary and car crime trends in the low-crime areas where patrols were to be reduced. If there was any indication of an increase, he agreed to restore the patrol patterns in place prior to the experiment. Bouza was a very persuasive leader, and he won the Council’s hearts and minds. When the vote was cast, the approval was unanimous. The rest is history—including, some three decades later, national funding of some 100 million pounds for hot spots policing by the government for England and Wales in 2021–2022.

What About Problem-Solving?

Perhaps the most intensive use of resources Tony Bouza approved for evidence-based policing was for the ReCAP experiments in repeat call address policing. Herman Goldstein (1979) had proposed a model for reducing patterns of crime which was attracting widespread support world-wide. But as of 1985, there had been no randomized controlled tests of the strategy. I asked Chief Bouza how many police officers he could devote to 250 of the hottest addresses in the City, collectively generating thousands of calls and taking up police time to answer them. He was able to offer four patrol officers (constables) and one sergeant, working together as a team, but each with a caseload of some 50 addresses each. The experiments found that the problem-solving approach produced a small reduction in calls for police to high-demand residential addresses, but not to commercial addresses (Weisburd, Telep, Hinkle, J. C. and Eck, 2008).

Since Chief Bouza pioneered the use of RCTs to test problem-oriented policing (POP), a total of 34 rigorous tests (including 7 new RCTS) have been reported (Hinkle et al., 2020). Thousands of police professionals review this research every year, and apply it to their work.

Science and Policing

Beyond the historic contributions of the Minneapolis “crime lab,” as Minneapolis reporter Dennis McGrath (1988) called it, Tony Bouza contributed something far broader. His biggest experiment was in making so much room for science in the operations of a major city police force. Many chiefs, facing similar options, have decided that their force was “too busy’’ to engage in experiments or any other kind of science. Yet what they could never learn without science was whether it could make them ‘’less busy.’’ That could happen in two ways. One is that better strategies could reduce crime, and hence demands for police service; this is what happened in Minneapolis in the first hot spots experiment.

The other way that science could help make police ‘’less busy’’ is that better strategies could respond appropriately with less police time. The use of rapid video response in the Kent (UK) Police, for example, has saved substantial amounts of police time, while increasing detection rates—all by not sending a police car when a caller is willing to have an immediate call with a police officer (Rothwell, McFadzien, Strang, Hooper and Pughsley, 2022). That kind of experiment would have been unimaginable before Bouza set so many precedents in Minneapolis. 

Chief Bouza thus fostered science in policing by setting a strong example to other police chiefs. He was not the only leader of his generation to embrace research and development, and especially to work with social scientists as partners (Wexler, 2023). Yet he inspired many others who had not yet done that. By the mid-1990s, when the National Institute of Justice offered seed grants to encourage partnerships between police agencies and their local universities or research institutions, there was no shortage of chiefs willing to sign up. Before Tony Bouza came to Minneapolis in 1980, there would have been a shortage indeed.

As the Executive Director of the Police Executive Research Forum, Dr. Chuck Wexler, wrote shortly after Bouza’s death:

The Minneapolis study didn’t tell us everything we needed to know about domestic violence policy, but it advanced our knowledge of the subject. Bouza deserves credit for welcoming researchers into his agency to develop evidence-based strategies that were used nationwide. I encourage you all to do what you can in your agencies to advance our collective knowledge. Because the alternative, as Bouza put it, is to be stuck with “seat-of-the-pants adoption of remedies thought to work.” (Wexler, 2023).

Bouza’s encouragement has had sustained effects, well beyond US borders. Even the position I hold as I write this editorial in the UK is inspired by Tony Bouza. It was Commissioner Sir Mark Rowley who boldly embraced the precision of evidence-based policing to create the role of the first Chief Scientific Officer at New Scotland Yard, serving the London Metropolitan Police. Yet that tradition and body of research would not have developed absent Chief Bouza’s earlier leadership.

Democracy against racism

One final point must be made about Minneapolis and its long-ago chief, Tony Bouza. Over three decades after he retired, his former police force became infamous for a tragic failure—one that has undermined police legitimacy around the world. This awful fact requires some comment on Bouza’s defense of democracy, and not just his pursuit of knowledge. That defense starts with his bedrock commitment to social equality, including his pioneering success in hiring minorities and women in substantial numbers as Minneapolis police officers. But that defense ends with the challenge of creating lasting change in the cultures of democratic police agencies, even after 9 years as the chief.

The murder of George Floyd by Minneapolis police officers in 2020 was tragic in many ways. One of them was collective amnesia about Bouza’s legacy of fighting racism in policing and society. Yet the historical record is clear on his outspoken views. As the New York Times (Roberts, 2023) quoted him as saying in the 1970s about high crime in American areas of concentrated minority populations,

“The crime and violence that result,” he wrote, “are the consequences of the racist policies and the poverty that the white overclass has visited on the Black underclass.’’

Even his most controversial statement was an attempt to draw attention to the discrimination against young minority males in New York City, some of whom committed visible robberies during a police protest at Yankee Stadium in 1976 where a major prize fight was taking place (Roberts, 2023):

“The kids impinged on the consciousness of more prominent Americans,” Mr. Bouza said in response to the mostly wealthy white ticketholders’ wrath at the teenagers who ran amok that night. “If I failed, it’s because I didn’t continue to make these feral children invisible to middle- and upper-class Americans who aren’t used to seeing them.”

As for Derek Chauvin, the Minneapolis officer who suffocated George Floyd, Bouza (2022) told me he would have dismissed Chauvin very early in that Chauvin’s career, which attracted 18 complaints—including three shooting incidents, one fatal (Ailworth, Kesling, Gurman, and Barrett, 2020). As Bouza was about to retire in 1988 (long before Chauvin was even hired), Bouza told me that he feared his successors would not be as tough on police misconduct as he had been. “The cops who hate me,” he said, ‘’want to thump people, and they know I won’t let them’’ (Bouza, 1988). Yet when I walked down the streets of central Minneapolis with him, passers-by stopped us to shake his hand. “Great job, Chief,” I heard them say, many, many times over his nine years as chief.

Despite the challenge of any legacy for a single police agency, Tony Bouza’s legacy for global knowledge remains unchallenged. Evidence-based policing is alive and well and supports democracy. It is informing choices about policing that can be made by voters and the people they elect. It is demonstrating what many voters knew already: that policing can make communities safer.

Yet the biggest policing experiment ever conducted in Minneapolis was in Tony Bouza’s last 2 years of life when he remained as outspoken as ever in advising the voters what to do. The experiment was a 2021 referendum on whether to abolish the Minneapolis Police Department. When National Public Radio (2021a) asked for his views on the referendum, this is what he told the national audience—and tens of thousands of local voters:

The Minneapolis Police Department has to crack down on that culture [of brutality], but …eliminating the department would go too far. “It’s crazy,” he says. “They don't know what they're talking about. It’s not an institution you want to eradicate. It's an institution you need to reform and lead and reshape.”

The result of the referendum was a clear defeat of abolishing the police. While many white voters were willing to support abolition (with an unspecified alternative), the majority of voters from the African-American community was not (NPR, 2021b). The Minneapolis Police remains a department of city government.

As for evidence-based policing, over 10,000 police professionals now belong to research societies around the world. Since the first Society of Evidence-Based Policing was founded in the UK in 2010, affiliated organizations have recruited members in Australia-New Zealand, the American Society of Evidence-based Policing, the Canadian SEBP, and as of September 2023, the Dutch Society of EBP (Netherlands 2023). These institutions are thriving, with increasing amounts of research publications, conferences, and influence on police reform.

The democratic world still faces a crisis of police legitimacy. However, the institutions of evidence-based policing continue to grow. From all the evidence, they will do what Tony Bouza would want them to do: carry on and keep thinking.