CRIMINAL JUSTICE

R E S E A RC H A RT I C L E

I M P A C T O F P O L I C E O N C J R E F O R M

The Impact of Police on Criminal Justice Reform Evidence from Cincinnati, Ohio

Robin S. Engel U n i v e r s i t y o f C i n c i n n a t i , T h e I n t e r n a t i o n a l A s s o c i a t i o n o f C h i e f s o f P o l i c e

Nicholas Corsaro M. Murat Ozer U n i v e r s i t y o f C i n c i n n a t i

Research Summary Despite significant national reductions in crime during the past three decades, a comparable reduction in adult arrest rates has not occurred. In addition, scant attention has been paid to the role of the police in pretrial justice and other criminal justice reform efforts, despite their role as gatekeepers to the criminal justice system. A key inquiry that must be addressed by both academics and practitioners is whether it is possible to reduce crime and the number of arrests simultaneously. Cincinnati (Hamilton County), Ohio, provided a unique opportunity to examine this unanswered question when it closed the Queensgate Correctional Facility in 2008, thereby reducing the available jail space in the county by 36%. By relying on an interrupted time-series analysis, our findings show that contrary to public concern, both crime and arrests were reduced in Cincinnati even after the jail closure. Specifically, the Cincinnati Police Department reported a statistically significant decrease in felony arrests, and a nonsignificant decline in misdemeanor arrests, while maintaining a continued (nonsignificant) decline in violence and property crimes. Importantly, our findings demonstrate that the previous

This article was originally presented at the John Jay College Roundtable on Pretrial Practices on March 18, 2015, sponsored by the Laura and John Arnold Foundation. The thoughts and opinions in this article are of the authors and do not necessarily represent those of the City of Cincinnati, Hamilton County Sheriff’s Office, or Cincinnati Police Department. The authors thank members of the Cincinnati Police Department, participants of the John Jay College Roundtable, and colleagues from the University of Cincinnati School of Criminal Justice for their helpful comments. Direct correspondence to Robin S. Engel, Vice President for Safety and Reform, University of Cincinnati, P.O. Box 210632, Cincinnati, OH 45221 (e-mail: [email protected]).

DOI:10.1111/1745-9133.12299 C© 2017 American Society of Criminology 375 Criminology & Public Policy � Volume 16 � Issue 2

Research Article Impact of Police on CJ Reform

existent downward trend in Cincinnati reported crimes was not interrupted with the loss of more than one third of the available jail space in Hamilton County.

Policy Implications Policy makers and practitioners are concerned with balancing the individual rights of the accused with public safety; reducing incarceration; and promoting a more efficient, effective, and fair criminal justice system. The Cincinnati Police Department addressed these fundamental concerns by changing how officers viewed the use of arrest: as a limited commodity rather than as a standard response. By using strategies such as problem- oriented policing, place-based policing, and focused deterrence, Cincinnati Police were able to narrow their focus on the repeat places, problems, and groups of individuals that were driving crime within the city. The evidence suggests that the police can have a significant impact on pretrial justice and other criminal justice reform efforts through the implementation of evidence-based policing strategies that seek to reduce crime and reduce the use of arrest simultaneously.

P olice agencies in the United States balance myriad concerns that, if improperly

handled, have the potential to undermine their perceived legitimacy and overall

effectiveness. While balancing constrained resources, police are tasked with better addressing root causes of negative public attitudes related to police use of force, policy shifts

that focus on reducing mass incarceration, and the need to sustain record-level declines in

crime rates (Clear, 2011; President’s Task Force on 21st Century Policing, 2015). In this article, we argue that the important role of the police is often overlooked by those interested

in broad reform efforts and pretrial justice at other decision points within the criminal

justice system. Too frequently reforms are implemented that target one specific area of the

criminal justice system without considering how changes at other stages could significantly enhance those reform efforts. The police have long been recognized as the gatekeepers of

the criminal justice system with near-exclusive control over arrest decisions (Bittner, 1967).

Nevertheless, the police are rarely included in discussions regarding pretrial, court, and jail

reforms. Rather, the flow of clients through the criminal justice system is often accepted as an invariable “condition of work” for those engaging in reform efforts at various decision

points within the system (Lipsky, 1980). This limited consideration of the role of police is

a critical impediment to successful and sustainable reform efforts throughout the criminal

justice system. We contend that police strategies designed specifically to reduce the use of arrest, while

reducing crime, can have substantial short- and long-term impacts on the entire criminal

justice system. To demonstrate this impact, we consider work conducted in Cincinnati,

Ohio. Faced with significant budget cuts in December 2008, the Hamilton County Sheriff ’s Office (HCSO) closed its ancillary jail facility, which reduced the available jail bed space in

376 Criminology & Public Policy

Engel, Corsaro, and Ozer

the county by more than one third. Many civic leaders, law enforcement officials, and mem-

bers of the public predicted the worst. With a severely reduced capacity to house offenders awaiting trial or serving sentences for misdemeanor offenses, they forecasted that crime

would increase dramatically and that conditions in Cincinnati could become dangerous.

We examine the specific impact of the jail closure in Hamilton County on reported

crime and arrests in Cincinnati. By using an interrupted time-series design, we measure the differences in the number of Part I reported property and violent crimes and the monthly

number of misdemeanor and felony arrests by the Cincinnati Police Department (CPD)

during a 14-year period that includes time periods before and after the December 2008 jail

closure. Contrary to the initial concerns of many leaders in Cincinnati, the findings show reductions in violent crime and property crime, along with reductions in misdemeanor and

felony arrests in the years after the jail closure. The policing community, and in particular

the CPD, was forced to adopt policing strategies that focused less on arrest, while striving to

reduce crime. They accomplished this by viewing the use of arrest as a limited commodity and by incorporating a variety of evidence-based policing practices into their operational

strategies, which reduced their reliance on the use of arrest to reduce crime. Study findings

are viewed in light of public concerns regarding unjust police practices as well as with policy changes concerning incarceration declines.

Criminal Justice Reform and the Role of Police for Pretrial Justice The “get tough” crime control approach used over the past 40 years across all branches of

the criminal justice system can be classified as punitive and costly in terms of direction,

approach, and resource exhaustion. In 1980, there were 520,000 people in prison and jail in

the United States; comparatively, in the peak year of 2007, there were more than 2.3 million people in jails and prisons (Sourcebook of Criminal Justice Statistics, 2011). Support for this

approach has been linked to strong public attitudes toward crime control and punishment

(Cullen, Clark, and Wozniak, 1985) coupled with the observed crime drop of the 1990s

(see Blumstein and Rosenfeld, 1998), ensuring that the “net widening” method of criminal justice policy would continue for another two decades, or until sustained declines in crime

made it easier to portray the U.S. prison system as unnecessarily large (Austin et al., 2007).

In 2015, The New York Times asked experts to debate the following question: Will crime rise if more people are kept out of prison? The real-world experiences in Cincinnati provide evidence that the answer to this posed hypothetical question can be “no.” We conclude with

a series of recommendations for changing the way police view and use arrest. Specifically,

we argue that the evidence-based policing practices used by the CPD, including hot-spots

policing, focused deterrence, and problem-solving crime prevention strategies, are the most promising for reducing crime while reducing arrests and, furthermore, that these practices

will impact pretrial and other criminal justice reforms efforts simply by reducing the size

and type of the arrestee population.

Volume 16 � Issue 2 377

Research Article Impact of Police on CJ Reform

Pretrial justice has been defined as “the honoring of the presumption of innocence,

the right to bail that is not excessive, and all other legal and constitutional rights afforded to accused persons awaiting trial while balancing these individual rights with the need

to protect community, maintain the integrity of the judicial process, and assure court

appearance” (VanNostrand and Keebler, 2007: 20). Achieving pretrial justice, however, has

remained an elusive goal for more than 50 years, in part because of the delicate balance needed between individual rights, liberties, and freedoms and the protection of society. As

such, additional research is needed to examine the pretrial stages of the criminal justice

system, including bail decision making, rights of accused persons, and the role of pretrial

services. Although the academic literature has mainly been focused on the factors that lead to pretrial outcomes and how these outcomes influence sentencing decisions (e.g., Sacks

and Ackerman, 2014; Williams, 2003), more recent work has been focused on developing

evidence-based practices to predict pretrial outcomes better while balancing the rights of the

defendant (e.g., Bechtel, Lowenkamp, and Holsinger, 2011; Mamalian, 2011). Likewise, empirical examinations of the impact of bail reform and pretrial services are available

(e.g., Goldkamp and White, 2006; Lowenkamp and VanNostrand, 2013). Yet despite the

growing body of research aimed at documenting the issues surrounding pretrial justice and reform, the critical role of the police in this process has been almost ignored by the research

community and, until recently, by the practitioner community. The number and type of

arrestees entering the pretrial stage of the criminal justice system are determined almost

exclusively by the police, yet the role of the police is perhaps the largest gap in available research on pretrial justice, and on criminal justice system reform more generally.

Scholars have noted for decades the important role of police and the near monopoly they

have on the decision to invoke the criminal justice process (Bittner, 1967). Furthermore,

a large body of research has been focused on understanding police discretion, and the most frequently examined decision historically is whether to arrest (Kochel, Wilson, and

Mastrofski, 2011; Riksheim and Chermak, 1993; Sherman, 1980). The results of this

body of research generally show that legal factors (e.g., seriousness of the offense and

strength of the evidence) have the strongest influence over police arrest decisions. Other factors have also consistently demonstrated a significant impact on the likelihood of arrest,

including victim preference, victim–suspect relationship, and suspect demeanor, although

there is mixed evidence regarding the impact of suspect, victim, and officer demographic

factors (Engel and Swartz, 2014). The impact of suspect demographic factors on arrest decisions, however, has been hypothesized to be more influential for less serious offenses and

misdemeanor arrests—the same types of cases that are of greatest concern for pretrial reform.

At the pretrial stage, the results of research on extralegal factors have shown that minority

males are more likely to be detained, have higher bond amounts, and have higher odds of imprisonment relative to White males, even when including rigorous controls for legally

relevant information (Wooldredge, 2012). Furthermore, research findings demonstrate that

defendants jailed until time of trial are more than four times likely to be sentenced to jail and

378 Criminology & Public Policy

Engel, Corsaro, and Ozer

three times more likely to be sentenced to prison, even when controlling for demographics,

risk level, and offense type (Lowenkamp, VanNostrand, and Holsinger, 2013). As a result, racial/ethnic disparities that exist at the arrest decision are compounded at subsequent stages

of the criminal justice system. Pretrial reformers advocate for the increased development

and use of evidence-based methods for determining pretrial release decisions (VanNostrand

and Keebler, 2009). And although we agree with this approach, we also argue that pretrial reform efforts need to begin sooner—before an arrestee ever stands before a court official

in the pretrial process.

Despite the amount of scholarly attention given to police arrest decisions, almost no

systematic research has been dedicated to better understanding the impact of arrest decisions on pretrial and jail reform efforts. This gap in research seems especially problematic because

of all the potential reform efforts in the area of pretrial justice (e.g., risk assessments

and pretrial services). Police strategies designed to reduce the use of arrest significantly,

particularly for minor offenses, have the potential for the greatest long-term impacts on the entire criminal justice system. Durlauf and Nagin (2011) made a compelling argument

based on deterrence theory that imprisonment and crime can be simultaneously reduced

if we concentrate more resources on policing and less on prisons. Their argument was focused on three basic propositions: (1) Evidence suggests that increasing the length of

prison terms has a modest impact on crime; (2) evidence suggests that increasing police

visibility by hiring more officers or strategically deploying current officers increases offenders’

perceived risks of apprehension and has deterrent effects; and (3) evidence suggests that the experiences of incarceration have a criminogenic effect and do not prevent reoffending.

They concluded that “the marginal deterrent effects of certainty and severity suggest that

crime prevention would be enhanced by shifting resources from imprisonment to policing”

(Durlauf and Nagin, 2011: 38). Nagin, Solow, and Lum (2015) contended that policies that shift perceptions of apprehension certainty can substantially impact crime. The authors

also contended that the concept of probability of apprehension is ill-posed without a

specification of the characteristics of the targets of victimization and that no summary

statistic (e.g., clearance rates) describing probability of apprehension risk provides useful information on whether a specific target will be victimized and, more generally, on crime

rates (i.e., the proportion of targets victimized). By logical extension, arrest figures also

likely provide no useful information on changes in crime rates, which has been established

in prior research (see Greenberg and Kessler, 1982). Thus, conclusions regarding a needed shift in police resources have been based specifically on an underlying assumption that an

increase in expenditures for police could deter crime without increasing arrests depending

on the policing strategies employed.

Although the role of law enforcement in pretrial and larger jail reform efforts has been generally overlooked by academic researchers, practitioners have increased their recognition

regarding the important role that the police can (and should) play in this regard. For

example, in 2010, the International Association of Chiefs of Police (IACP) in collaboration

Volume 16 � Issue 2 379

Research Article Impact of Police on CJ Reform

with the Bureau of Justice Assistance (BJA) and the Pretrial Justice Institute (PJI) convened

a focus group of police executives, prosecutors, judges, defense attorneys, and researchers to discuss the role of law enforcement in pretrial release decisions and their appropriate role for

improving the system. This group noted that “surprisingly, in only a handful of jurisdictions

does law enforcement play a role in helping a judge determine who should and should not be

released back into society” and further posited that “leaders in the justice field [feel] that the time has come for law enforcement to become a part of the dialogue on pretrial release and

eventually to take its appropriate leadership role in improving the pretrial system” (IACP,

2011: 3). The report summarizing these discussions was focused on the problems with the

financial bail system, and on the concerns about pretrial release of dangerous and violent criminals that are not being appropriately supervised in the community (IACP, 2011).1 Even

though the concerns noted are clearly important to the criminal justice field and the public

more generally, what is particularly striking is an issue that was not mentioned by the authors of the report. None of the key concerns identified by this panel included a consideration of the role of police as the gatekeepers of the criminal justice system. Specifically, there is no

mention of the importance of the initial decision to arrest as a key factor for pretrial and jail reform efforts. This, we believe, is a significant oversight. The sheer number of cases sent into the criminal justice system, the strength of the evidence related to arrests, and the type

and severity of the offenses resulting in arrest are all within the direct control of the police.

These same factors have the strongest impact on pretrial detention and release decisions.

The easiest and most direct route for pretrial reform is through the intake process, which is nearly singularly controlled by the police.

Others have echoed the sentiment of the vital role that the police will play in pretrial

justice reform. At the National Symposium on Pretrial Justice, sponsored by the U.S De- partment of Justice, Office of Justice Programs, and the PJI (2011), speaker James Austin noted that despite a significant reduction in crime across the last two decades, there has not

been a comparable reduction in the adult arrest rate. As Austin reportedly described: “Our

national ‘numbness’ to the experience of imprisonment, and its somewhat casual use for so

many offenses must change” and that “finding a solution other than needlessly incarcerating persons for whom jail is an improper response will be the ‘ultimate solution to bringing that

jail population back to where it is supposed to be’” (PJI, 2011: 13).

1. Specifically documented in this report were the following issues: (a) the balancing of release versus detention decisions and what set of criteria should be used to make these determinations; (b) the variation in size, budget, and resources of police agencies that might gather information pertinent to pretrial release decisions; (c) development of regional, unified central processing centers where all arrestees can be evaluated for release or detention; (d) determining additional consequences for those who fail to return to court when released; (e) the need for shared wisdom from other professional organizations representing the law enforcement community; (f) the impact of pretrial release on public and officer safety and witness intimidation; and (g) balancing bond decisions with dangerousness assessments (IACP, 2011).

380 Criminology & Public Policy

Engel, Corsaro, and Ozer

Most recently, in October 2015, more than 150 current and former police chiefs

from across the country united under the banner of “Law Enforcement Leaders to Reduce Crime and Incarceration” to announce their support of core principles to reduce crime

while reducing incarceration. These leaders noted: “We believe the country can reduce

incarceration while keeping down crime. We believe unnecessary incarceration does not

work to reduce crime, wastes taxpayer dollars, damages families and divides communities. We aim to build a smarter, stronger, and fairer criminal justice system by replacing ineffective

policies with new solutions that reduce both crime and incarceration” (Law Enforcement

Leaders to Reduce Crime and Incarceration, n.d.: para. 1). How these changes will be

accomplished is now part of the national conversation in policing, and it should be part of the conversations and research about pretrial justice.

Outputs Versus Outcomes: How Police View Arrest For decades, police have routinely captured and reported the number of arrests; these summary statistics are included in annual reports and other official documents (Moore,

2002; Sparrow, Moore, and Kennedy, 1990). But arrest counts have also been routinely

used by the police as a measure of productivity for specialized units, individual squads, shifts, and even individual officers. This view of arrests as work outputs—or as a demonstration

of productivity—was enhanced by the proliferation of Compstat, a management process

originally developed and implemented by the New York City Police Department (NYPD)

that used crime analysis to hold commanders accountable for crime reduction efforts (Weisburd, Mastrofski, McNally, Greenspan, and Willis, 2003). This process quickly spread

to other police agencies across the country, where the four principles of Compstat—accurate

and timely intelligence, effective tactics, rapid deployment, and relentless follow-up and

assessment—were adopted and modified to fit local agency needs (Godown, 2009). In many police agencies across the country, the number of arrests are tracked and reported by

staff to demonstrate productivity levels, with increases in arrests emphasized and praised by

police executives.

Measuring and rewarding the number of arrests was initially based on an underlying, yet untested, assumption that increased arrests were positively correlated with reductions

in crime. Over time, arrests became viewed by members in many police organizations as

“outputs” rather than as “outcomes” (McCarthy, 2015). This view is problematic for several

reasons, but most importantly because it encourages field officers to generate measurable “activity” (i.e., arrests) to demonstrate they are working during their shifts. These arrests

are often for minor, nonviolent crimes or violations, and they have contributed to the

unprecedented number of individuals entering the criminal justice system.

Similar problems were observed with the proliferation of aggressive zero-tolerance en- forcement strategies that were loosely based on broken windows theory, which suggested

that major crimes could be reduced by focusing on the enforcement of minor crimes and

incivilities (Harcourt and Ludwig, 2006; Wilson and Kelling, 1982). Again, the use of

Volume 16 � Issue 2 381

Research Article Impact of Police on CJ Reform

these strategies inevitably resulted in dramatic increases in arrests for minor, nonviolent

offenses, yet the correlation to reductions in crime has been disputed (e.g., Fagan and Davies, 2000; Harcourt and Ludwig, 2006). Additionally, at the height of the prolifer-

ation of the zero-tolerance movement, Greene (1999) compared police approaches and

changes in crime in New York City (aggressive enforcement) with those in San Diego

(problem-oriented enforcement), where she concluded that effective crime control could be achieved with problem-oriented policing while producing fewer negative impacts and citizen

complaints.

Thus, what if the police embraced a culture where the use of arrest was consistently

viewed as a limited commodity—a tool only to be used when no other alternatives for resolution were readily available, rather than as a gauge of effectiveness or the primary

mechanism to control crime? In this scenario, the impact on the entire criminal justice

system, including pretrial justice, could be profound. Some current policing strategies have

demonstrated success in reducing crime and violence, while reducing incarceration (Braga, 2016). The proliferation of these evidence-based policing strategies—in combination with

enhanced pretrial services, use of risk assessments, and other innovations—is critical for

pretrial justice and reform. A natural experiment from Cincinnati clearly demonstrates the potential impact on the criminal justice system when police view the use of arrest

differently.

Research Setting: Cincinnati, Ohio The history of police–community relations and implementation of policing strategies and

tactics in Cincinnati is complicated, with significant changes occurring over the last 15 years.

In April 2001, Cincinnati experienced racial tension and civil unrest based on a series of

police killings of African American suspects. After an extensive U.S. Department of Justice (USDOJ) investigation, the city entered into a memorandum of agreement (MOA) to im-

plement numerous reforms within the police department, including changes in use-of-force

reporting and training, and implementation of a risk management system. In addition, the

settlement to a racial profiling lawsuit resulted in the historic Collaborative Agreement de- signed to address police–community relations. The Collaborative also included provisions

to adopt the MOA with the USDOJ in its entirety. As a result, the city created the Citizens’

Complaint Authority, the Community Police Partnering Center, as well as other reforms

to improve community outreach and engagement. City and police leaders aggressively ad- dressed issues of racial tension, civil unrest, and the need for police reform by establishing

proactive problem-solving approaches, as well as collaborative efforts among police, com-

munity, and businesses to promote a better quality of life for residents (Eck and Rothman,

2006). There was an ongoing commitment to a variety of efforts designed specifically to improve effectiveness, transparency, and legitimacy of the police (Sabel and Simon, 2016).

In addition, strong partnerships were forged with local academic researchers to implement

proactive and collaborative policing approaches (Eck, 2014; Engel and Whalen, 2010).

382 Criminology & Public Policy

Engel, Corsaro, and Ozer

Of course these changes did not happen overnight, nor did they occur in a vacuum;

rather they evolved over years and required continuous managerial oversight and attention. There have been numerous empirical articles, case studies, descriptive reports, and media

accounts about the implementation of police reform efforts in the 15 years after the civil

unrest in Cincinnati.2 Across these sources is a consistent description of a police agency

undergoing tremendous organizational change. When framed in a general timeline, these sources collectively document a shift from initial disengagement and “de-policing” in

the first few years (2001–2004) immediately after the civil unrest and the Collaborative

Agreement (Shi, 2009), followed by a rise in violent crime and shooting offenses from 2004

to 2006 that led to the creation of a high-impact, arrest-focused, specialized unit (Operation Vortex) that used traditional law enforcement tactics and implemented unfocused and

indiscriminant “sweeps” to reduce crime (Sabel and Simon, 2016; Seabrook, 2009). This

aggressive policing strategy was later abandoned, followed by the slow embrace of problem-

solving and focused policing strategies (Eck, 2014). These efforts began in mid-2007 with a focused deterrence strategy (Cincinnati Initiative to Reduce Violence, CIRV), followed

by several additional strategies after 2008 including hot-spots policing, problem-oriented

policing, and place-based policing strategies (Eck, 2014; Engel, Tillyer, and Corsaro, 2013). Although policing strategies in Cincinnati were evolving, operations within the

county were not keeping pace. In 2006, Hamilton County, Ohio, officials commissioned

a $190,000 study of the county’s criminal justice system (Leis and Heimlich, 2006). This

180-page study, conducted by Voorhis Associates (a corrections planning firm), projected that the jail population would grow to more than 3,000 prisoners by the year 2020 but

that the four separate jail facilities in Hamilton County only had a total capacity of 2,272

and were functionally obsolete. The study therefore recommended that the county replace

one of its current jail facilities with a new 1,800-bed jail, raising its overall total capacity to meet the growing need. Multiple proposals to fund the new jail through sales tax

increases were placed on the ballot; nevertheless, these proposals were routinely defeated by

voters.

By 2007, Hamilton County was in economic crisis. Rather than building a new jail, in 2008, the Hamilton County Commissioners enforced dramatic budget cuts for all

county services, including the Hamilton County Sheriff ’s Office (HCSO), which had

jurisdiction over the four county jail facilities. As a result of these mandatory budget cuts,

the HCSO was forced to close the Queensgate Correctional Facility, which was the second largest of the four county facilities that housed inmates. Queensgate was an 822-bed jail

facility that initially housed primarily minimum-security offenders. Nonetheless, in 2006,

HCSO officials claimed that 60% of the inmates were classified as medium or maximum

security (Leis, 2007). On December 20, 2008, more than one third (36%) of the available

2. For example, see Eck (2014), Eck and Rothman (2006), Engel et al. (2013), Green (2008), Ridgeway et al. (2009), Riley et al. (2005), Sabel and Simon (2016), Schatmeier (2012), Shi (2009).

Volume 16 � Issue 2 383

Research Article Impact of Police on CJ Reform

jail beds in Hamilton County were eliminated when the HCSO closed the Queensgate

facility. The political and public debate leading up to the closure of the jail facility was intense,

emotionally charged, and at times malevolent. Many politicians and law enforcement

officials—and in particular the outspoken sheriff—predicted unprecedented increases in

crime. “It’s a frustrating, dangerous situation. We’re going to get to a point where we’re not going to be able to provide the necessary services. We’re at that point” (Sheriff Simon

Leis, Jr. as quoted by Brown, 2008: para. 7). Members of the judiciary also predicted the

worst. For example, Judge William Mallory of the Hamilton County Municipal Court was

quoted in the Cincinnati Enquirer as saying: “I don’t put people in jail if they don’t belong there . . . We’re looking at a powder keg” (Bronson, 2009: para. 10).

Although attempts at culling specific jail data from the HMCO were unsuccessful

as a result of lack of data collection and availability, it is well recognized that the largest

contributor to the Hamilton County jail population are arrestees from the CPD. The CPD is the largest police agency in Hamilton County by a significant margin, with 1,008 sworn

officers and 124 nonsworn employees (cincinnati-oh.gov/police/). The predicted impact of

the jail closure on Cincinnati, and on the CPD specifically, was expected to be enormous. Yet, after the first year of the jail closure, crime in Cincinnati did not increase as predicted.

Rather, Part I Violent crime decreased 5.8% and Part I Property crime remained stable, with

reductions of less than a half percent compared with the year prior. The CPD simultaneously

reduced the number of in-custody arrests in 2009 compared with 2008 by 15.1% for felony offenses and by 11.7% for misdemeanor offenses.

Yet angst continued to grow among the public, politicians, and some law enforcement

officials regarding the lack of jail space, based on the perceived inevitability that crime

would necessarily increase at some point in the near future. One strategy adopted by the HCSO to handle the reduced jail space was to cite and release selected offenders.

Many perceived the cite-and-release practices for arrestees that the HCSO instituted as

largely unsystematic, lacking in principles of efficiency and effectiveness. Furthermore,

the screening tool used for arrestees that was created internally by HCSO to determine cite-and-release and early release decisions was not a validated risk assessment instrument,

and it lacked many of the most common indicators widely accepted in correctional research

(e.g., Latessa and Lovins, 2010; Lowenkamp, Latessa, and Holsinger, 2006). Stories of

offenders who were arrested, processed at the jail, immediately released, and rearrested during the same 8-hour work shift became folklore and were repeatedly shared at law

enforcement meetings and with the media. Critically described as a “revolving door,” the

HCSO adopted a policy of arrest and release of nonviolent offenders with no detainment

prior to making bail or bond, or appearing before a judge for arraignment. These practices continued for 5 years until a new sheriff was elected and discontinued their use (Elgazzar,

2013). Throughout this time, the following question remained: What would be the

long-term impact of the jail closure? This question is the subject of our analyses.

384 Criminology & Public Policy

Engel, Corsaro, and Ozer

Data and Method The data for this study include reported crime incident data as well as official arrest measures from the CPD, collected between January 1, 2001 and December 31, 2014.3 The data were

culled into a monthly format (i.e., each count represents the total number of crime incidents

or arrests that occurred between the first and the last day of each month-year). A total of four

outcomes were modeled: (1) Uniform Crime Reports (UCR) Part I violent crime incidents,4

(2) UCR Part I property crime incidents,5 (3) felony arrests, and (4) misdemeanor arrests.

Interrupted time-series analysis is an analytical technique that we used to examine

whether there was a statistically significant shift in crime incidents and arrests (both felony

and misdemeanor) in Cincinnati between pre- and postjail closure periods (December 2008) while controlling for systemic shifts in the longitudinal data. Time-series analysis can be

a useful analytical tool for isolating programmatic or immediate social impact (Cook and

Campbell, 1979). We followed the Box and Jenkins (1976) autoregressive integrated moving

average (ARIMA) approach by using model identification, estimation, and diagnostic testing to find the most appropriate stochastic model for each time series prior to estimating the

impact of the jail closure. The use of ARIMA models allows us to examine the impact

on time-series outcomes while testing and controlling for the likelihood that the repeated

observations were independent. Dependence among the observations is often problematic in repeated count data, and it would violate the assumptions of typical inferential techniques

that test the intervention effect.

Box and Jenkins (1976) outlined how ARIMA models can remove dependence from

the data, although it should be noted that there are concerns with the model specifica- tion process inherent to ARIMA modeling (see Kleck, 1997). To determine whether these

special models are needed, the autocorrelations and partial autocorrelations for each time

series were inspected.6 If the intervention (closing of Queensgate) had the hypothesized

3. Unfortunately, reliable data (including arrest, crime, jail population, and jail transfer data) were not available from the HCSO during the time periods examined. Therefore, this analysis only considers crime and arrest data from the CPD.

4. UCR Part I Violent offenses include murder and nonnegligent homicide, forcible rape, robbery, and aggravated assault.

5. UCR Part I Property offenses include burglary, motor vehicle theft, and larceny-theft. Arson is collected by the Cincinnati Fire Department, not the Cincinnati Police Department, and therefore is not included in these analyses.

6. As a sensitivity check, and in an effort to address concerns inherent to ARIMA model specification, we also conducted ordinary least-squares (OLS) regression analyses on each time-series model presented herein. Specifically, we followed steps outlined by Darlington (1990) in that we estimated the impact of time, time-squared, and time-cubed (to reduce the presence of functional changes within the time series over time) as well as seasonal effects (through monthly and annual dummy variables), and the results were stable across different methodologies. Findings from the ARIMA models (instead of the regression-based models) were presented in the Results section because there was no evidence of “powerful but erratic trends and seasonal components” (see Gardner, 1979) that might make OLS regressions more precise than ARIMA estimates.

Volume 16 � Issue 2 385

Research Article Impact of Police on CJ Reform

T A B L E 1

Changes in Cincinnati Structural Characteristics (2000–2014)

City Structural Characteristics 2000 2010 2014

Resource Deprivation Indicators Percent Black 42.9 44.8 43.5 Percent persons in poverty 18.2 14.0 17.4 Median family income $29,493 $33,681 $34,002 Percent single household families with children 12.4 11.8 12.0

Population Size 331,285 296,943 297,117 Percent Divorced 12.1 11.7 11.9 Percent 15–24 Years Old 16.7 17.9 17.5 Percent Unemployed 7.3 6.8 8.5 Police Per Capita 30.84 36.70 36.45

effect of increasing the total number of arrests or offenses, the result would be evident

as a change in level of the time series at the point of intervention. Thus, trends in ar-

rests and crimes would remain steady, but a constant would be added to all points after the intervention. As it is feasible that the direction of the series could also change at the

time of intervention, a general multiple regression model was employed that could test

for changes in slope as well as in level. Finally, we examined potential changes in the

long-term trends in our estimation methods. After diagnosis and model selection, the inter- vention parameter (an abrupt, permanent transfer function given the theorized immediate

and sustained impact) was next incorporated within each ARIMA model. SAS version

9.1 (SAS Institute, Inc., Cary, NC) was used for all ARIMA analyses. Alternative func-

tional forms (i.e., first-order, gradual, or pulse transfer functions) were also examined and yielded no substantive differences, and thus, they were excluded in the analyses presented

herein.

Findings To interpret changes in arrests and crime within Cincinnati over time, we first provide a

series of summary statistics via census estimates that document the changes in structural

conditions, established within the scholarly literature, that are associated with varying

levels of crime across cities over time (see Land, McCall, and Cohen, 1990; McCall, Land, and Parker, 2011). Traditional indicators of resource deprivation (i.e., percent Black,

percent in poverty, median income, and single-family households with children younger than

18 years of age) remained stable in Cincinnati between 2000 and 2014. The city population

maintained its downward trajectory (that began in the 1960s) with a decline from 331,285 in 2000 to 296,943 in 2010 (–10.4%). The estimated city population remained steady

between 2010 and 2014 (297,117). Table 1 also shows that additional correlates of city

crime changes (percent divorced, percent 15–24 years old, and percent unemployed) as well

386 Criminology & Public Policy

Engel, Corsaro, and Ozer

as the number of uniformed police officers per capita were also largely unwavering during

the period of examination. When we move to the interrupted time-series analyses, the best-fitting model for each

outcome was selected based on reviewing the time plots, the autocorrelation function (ACF),

partial autocorrelation function (PACF), and model fit statistics (AIC and BIC). For violent

crime incidents, an AR1, 12-month seasonal model (1,0,0)(1,0,0)12 that controlled for seasonal trends was most appropriate for modeling changes in violence within Cincinnati.

There were no statistically significant spikes in the Box-Ljung Q residuals at key lags (24 months), which indicates additional support for appropriate model fit. As observed

in Table 2, the results showed no significant evidence of change in violence in that the postclosure parameter was not statistically significant for violent crimes. Additionally, the

direction of the estimate was negative (b = –25.77, standard error [SE] = 19.30), indicating violent incidents experienced a modest (and nonsignificant) decline in the postclosure period

(December 2008–December 2014) relative to the preclosure period. Likewise, property crimes in Cincinnati did not experience a statistically significant shift after December 2008,

and again the postclosure estimate was negative in direction (b = –.03, SE = .02), indicating a slight decline in property crimes, net of controls.

We next examined whether arrests within the city changed in the postclosure period

relative to the preclosure period. The results shown in Table 2 indicate that there was a

statistically significant decline in felony arrests (b = –.13, SE = .06), which equates to a 13% reduction in felony arrests, net of controls. Finally, misdemeanor arrests did not experience any significant changes in the postclosure period, and again, the estimate was

negative in direction (b = –.08, SE = .04). In sum, the results here indicate that not only was there no empirical evidence of

an increase in crime incidents or arrests at the time of the closure of the jail facility but a consistent (and nonsignificant) decline in violence, property crimes, and misdemeanor

arrests, and a statistically significant decrease in felony arrests (displayed in Figure 1). These

estimates account for the general decline in Cincinnati offenses and arrests that took place

between 2001 and 2014 (displayed in Figure 2) while controlling for seasonal fluctuations in the time-series data.

Given that felony arrests were the only empirically examined outcome that experienced

a statistically significant shift in the postintervention (i.e., post jail closure) period, we

conducted a series of sensitivity tests. Specifically we followed procedures used by Cook and MacDonald (2011) and by Piehl, Cooper, Braga, and Kennedy (2003) to assess whether

model fit parameters and the intervention point estimate coefficients had the largest

estimated effect sizes (and optimal model fit indices) at the time of the intervention

(To), or in the postintervention period (i.e., potential lagged effects) relative to potential preintervention effects (i.e., lead effects). The results presented in the Appendix illustrate

that for felony arrests (the lone outcome with a statistically significant shift in the time-series

Volume 16 � Issue 2 387

Research Article Impact of Police on CJ Reform

T A B L E 2

Impact of Hamilton County Jail Closure on Crime Incidents and Police Arrests (January 1, 2001–December 31, 2014)

ViolentOffenses(1,0,0) × (12,0,0) ARIMA Parameters Estimate SE t Value

Postclosure ω −25.77 19.30 −1.34 AR (1) φ .48 .06 8.43 AR (12) φ .45 .05 8.44 Box-Ljung test (24 lags) Q = 26.61 (p = .320) (Logged)PropertyOffenses(2,0,0) × (12,0,0) ARIMA Parameters Estimate SE t Value

Postclosure ω −0.03 .02 −0.14 AR (2) φ .15 .04 3.38 AR (12) φ .79 .04 18.01 Box-Ljung test (24 lags) Q = 31.64 (p = .136) (Logged)FelonyArrests(1,0,0) × (12,0,0) ARIMA Parameters Estimate SE t Value

Postclosure ω −0.13 .06 −2.12 AR (1) φ .34 .07 4.78 AR (12) φ .28 .08 3.59 Box-Ljung test (24 lags) Q = 27.86 (p = .265) (Logged)MisdemeanorArrests(2,0,0) × (12,0,0) ARIMA Parameters Estimate SE t Value

Postclosure ω −0.08 .04 −1.61 AR (2) φ .15 .06 3.10 AR (12) φ .69 .05 12.79 Box-Ljung test (24 lags) Q = 33.18 (p = .100) Notes. AR = autoregressive term; ARIMA = autoregressive integrated moving average; MA = moving-average term; SAR = sea- sonal autoregressive term; SE = standard error. *p < .05. **p < .01.

models) the Wald chi-square statistic was smallest at the point of the true intervention

(December 2008 and later). A retrospective examination of the impact of closing the Queensgate Correction Facility

in Hamilton County, Ohio, shows the opposite of what many city leaders predicted for

Cincinnati: reductions in property and violent crime, accompanied by reductions in felony

and misdemeanor arrests. Note that these findings are modest when examined as part of the interrupted time-series analyses. Only one model (reduction in felony arrests) reached a level

of statistical significance when compared before and after the jail closure. Crime and arrests

were already trending downward before the jail closure (and were consistent with broader

388 Criminology & Public Policy

Engel, Corsaro, and Ozer

F I G U R E 1

UCR Part I Violent and Property Crimes (January 2001–December 2014)

F I G U R E 2

Felony and Misdemeanor Arrests (January 2001–December 2014)

national declines that were occurring simultaneously); nevertheless, the continuation of

these downward trends uninterrupted by the jail closure is powerful. As Cincinnati civic

leaders and criminal justice officials braced for the worst, crimes and arrests decreased year after year. A simple comparison of 2008 to 2014 shows felony arrests decreased 41.3%

and misdemeanor arrests decreased 32.7%; simultaneously, Part I Violent crime decreased

38.5% and Part I Property crime decreased 18.9%.

Volume 16 � Issue 2 389

Research Article Impact of Police on CJ Reform

Discussion: Viewing Arrest as a Limited Commodity It is instructive to consider why crime rates in Cincinnati continued on their preexisting downward trajectory rather than being interrupted by the elimination of more than one

third of the available jail space in Hamilton County. We contend that three interrelated

factors help to explain the continued decline of crime within the city coupled with

the statistically significant reduction in felony arrests. First, city and police leaders were committed to addressing civil unrest and citizen concerns regarding discriminatory police

practices by entering into the unprecedented (at the time) Collaborative Agreement,

adopting the MOA with the USDOJ. In many ways, this effort to improve effectiveness,

transparency, and legitimacy of the police marked the beginning of the adoption of the problem-solving model in Cincinnati. Second, a restriction in criminal justice resources

(e.g., the jail closure) constrained the local police to function from a more tightly coupled

framework of operations. The police had to be more circumspect in their use of arrests, and

as a result, it became a less widely used tool to control local crime problems. Third, as the city grew in its network of capacity to control crime, including the development of strong

partnerships with local researchers, police officials began to adopt emerging and promising

evidence-based police practices that drew largely (but not solely) from the deterrence

framework; evidence of impact from such proactive policing efforts (i.e., hot-spots policing and focused deterrence policing) also seemingly created tighter partnerships among the

various city stakeholders who were focused on reducing crime.

To address concerns regarding police–community relationships while focusing on

crime control, the city signed the historic Collaborative Agreement in 2002 with the parties agreeing that problem solving would be adopted as the principal strategy for addressing

crime and disorder problems by the CPD. The full implementation of problem solving,

however, required an additional 1-year extension (after a full 5-year external monitoring

period) because of a failure to implement this requirement of the agreement in full. In 2008, the external monitor indicated that the CPD had finally met this requirement (Green,

2008). The implementation process and specific problem-solving activities engaged by the

CPD were documented in a database that tracked implementation (Eck, 2014). A review

of these data revealed that 261 distinct problems were documented in the problem-solving tracking system from 2007 through early 2014, with the largest number of problem-solving

projects occurring during 2008 and 2009. Eck (2014: 2) concluded that “interviews with

police officials and community members, and a review of problem solving efforts show that

police have addressed a wide variety of problems, collaborated with numerous non-police organizations, and have done so successfully.”

The initial budgetary discussions and ultimate closing of the Queensgate Correctional

Facility in Hamilton County roughly correspond with a renaissance in policing strategies adopted by the CPD. The reduced jail space immediately forced CPD commanders

to think strategically about how and when their officers should arrest individuals. This

shift toward strategic planning follows the problem-solving model, which is based on a

390 Criminology & Public Policy

Engel, Corsaro, and Ozer

systematic analysis of the local crime problems (Braga, 2008; Goldstein, 1990). Specifically,

crime analysis helps identify the geographic and social patterns of crime problems across a variety of contexts. The analysis is designed to uncover patterns such as linkages to

drug sales and distribution; gangs; chronic offenders; domestic violence; illegal gun sales;

and related patterns of people, places, and contexts. As a result, the power to arrest was

increasingly regarded as a limited commodity by the CPD—a resource to be reserved for the most high-risk, serious, or chronic offenders.

Changes in CPD operational strategies were developed and implemented to address

crime patterns on the basis of understanding the underlying conditions identified by anal-

ysis (Eck and Spelman, 1987). The CPD specifically used a combination of crime analysis and hot-spots policing, problem solving, focused deterrence, and other evidence-based

approaches. From this model, as strategies became implemented, researchers monitored

the level of intervention (dosage) as well as assessed evidence of impact for each strategy

that was implemented. It is important to note that some of these strategies were imple- mented in 2007, prior to the jail closure in late December 2008. Nevertheless, it is clear

through field notes, qualitative ethnographic research, and media accounts that the re-

duced jail space increased the relevance, prioritization, implementation, and use of these strategies.7

In January 2009, crime analysis became a focal point for the CPD Patrol Bureau, and

weekly meetings were established to review crime patterns and trends systematically, discuss

alternative approaches to reduce crime, and hold district commanders accountable for crime in their areas. From a broad perspective, the strategic process is intended to be dynamic

and ongoing, allowing for continual revision with the ultimate goal of reducing targeted

crime problems with focused interventions. In the CPD, police managers were encouraged

to focus their limited patrol resources on problem-solving efforts to address repeat crime locations, victims, and suspects. Although hot-spots policing deployment strategies were

used by department leaders, officers were not encouraged to make large numbers of arrests

in these areas. Rather, combined with other problem-solving efforts, Chief Thomas H.

Streicher, Jr. articulated the need for “surgical precision” in the use of arrest in these high-crime neighborhoods (Seabrook, 2009). This surgical precision was developed in

many different ways. As described by Patrol Bureau Commander, Assistant Chief James L.

Whalen:

The practical impact of reduced jail space underscored the need for alterna- tive strategies, especially as it related to misdemeanor offenders. We further

7. The implementation and managerial oversight of various policing strategies were routinely discussed by CPD command staff during meetings that included participation by the lead author. These discussions and the accompanying adjustments to policing strategies were captured as field notes for a larger qualitative study of the CPD. These field notes, along with other published materials, were reviewed and used for this article.

Volume 16 � Issue 2 391

Research Article Impact of Police on CJ Reform

emphasized problem solving strategies that reduced arrest as a likely outcome.

Use of repeat databases [identification of repeat offenders, victims, suspects, and crime times] was increased and strategies that focused on specific offend-

ers were enhanced, such as direct communication with prostitutes and petty

thieves, to deter them from future offenses. Partnerships with social service

and mental health agencies were solidified and extended. (J. Whalen, personal communication, March 6, 2015)

All of these initiatives were implemented in an effort to reduce crime and the use of arrest that was contributing to the problems with jail overcrowding. Much of the analytical

work was supported by researchers at the University of Cincinnati, through a strong partner-

ship where research and strategic planning were combined with intimate street knowledge (Coolidge, 2009). As described by Assistant Chief Vincent Demasi at the time, “this is about

how to deploy units when you have limited resources” and that “it is more strategic than

anything we’ve done in the past” (Coolidge, 2009: para. 1). In retrospect, Colonel Demasi further noted that:

A combination of funding, state reductions, and federal regulations caused

county facilities to start reducing housing of misdemeanor offenders except

for violent recidivists. I asked my District Captains to make sure we were not using the limited jail resources for minor violations that could be handled in

alternative ways, and to be more strategic about who we were arresting and

why. (V. Demasi, personal communication, March 6, 2015)

Although some persons in the law enforcement community and the public voiced con-

cerns with this approach and warned of an impending surge in crime as a result of the limited use of arrest, it simply never happened. The lesson for police departments in other jurisdic-

tions is that analysis is the key to understanding and changing the underlying conditions

(i.e., opportunities for crime), which can reduce crime without an overreliance on arrests.

Use of Focused Deterrence in Cincinnati During this period of strategic reform, the CPD adopted an emerging policing strategy that

was designed to reduce crime and arrests: focused deterrence policing. Focused deterrence

strategies involve collaborative partnerships among federal and local government agencies, social service providers, and community members with a focus on (a) deterring violent

crime through targeted enforcement, deterrence, and social services and (b) enhancing

the perception of police legitimacy in marginalized communities (Kennedy, 2009). In

2007, the CPD led the development and implementation of the Cincinnati Initiative to Reduce Violence (CIRV) to combat the increasing levels of gun violence in the city (Engel

et al., 2013). With an emphasis on a data-driven and problem-oriented approach with

multi-agency collaboration, CIRV brought together a variety of stakeholders to engage in a

392 Criminology & Public Policy

Engel, Corsaro, and Ozer

focused deterrence approach similar to that of the Boston Gun Project’s Operation Ceasefire

(Braga, Kennedy, Waring, and Piehl, 2001). A key feature to focused deterrence strategies involves channeling criminal justice

partnerships to focus on specific criminal behavior committed by a small number of chronic

offenders and high-risk groups that are vulnerable to sanctions and punishment. Beyond

deterring violent behavior, the strategies also reward compliance and nonviolent behavior among targeted offenders by providing positive incentives, such as access to social services

and job opportunities. Indeed, the focused deterrence approach illustrates that net-

widening and enhanced sanctions for general lower risk offenders is unnecessary to impact

violence and that greater crime prevention gains occur when the brunt of criminal justice efforts are directed at the highest risk groups and individuals responsible for driving gun,

gang, and drug market violence (Braga and Weisburd, 2012). Complementary crime control

mechanisms are presumably linked to offender notification sessions, including normative

change in offending behavior as well as increased views on police legitimacy through the inclusion of the “community moral voice,” where members of local communities, including

reformed offenders, appeal to young men to seek alternative lifestyles away from gang

violence and, thus, increase informal social control mechanisms at the community level (Kennedy, 2009; Papachristos, Meares, and Fagan, 2007). The outcomes of a 42-month

pre–post evaluation demonstrated significant reductions in shootings and gang member

homicides in Cincinnati after the implementation of CIRV (Engel et al., 2013).

Of importance for this discussion is that CPD officers were instructed to write “CIRV” at the top of arrests paperwork for those individuals who were identified as group/gang

members and were arrested as part of the larger focused deterrence effort. This marking

on their paperwork allowed criminal justice officials throughout the system to recognize

individuals who were identified by CPD as a priority for enforcement action. The informal expectation was that those identified would not be released early as a result of lack of jail space

or given low bonds, and they would be considered by the prosecutor’s office as a priority for

prosecution. This informal mechanism allowed the courtroom workgroup to understand

CPD’s priorities better and provided information that functioned as a rudimentary risk assessment by police. The overall purpose, however, was to reduce incarceration over time

with arrest as a tool for those specifically and narrowly identified as a priority for enforcement.

Promise of Problem-Solving to Control Crime and Reduce Arrests Study findings within Cincinnati are also consistent with recent developments in scholarly

research as well as with policy conversations and recommendations. The results from several

studies show the most recent evolution in place-based policing holds promise for reducing

both crime and the use of arrest. There is ample evidence that most crime is concentrated at a few places (Sherman, Gartin, and Buerger, 1989); that overall crime (Weisburd, Bushway,

Lum, and Yang, 2004) and, more specifically, gun assault incidents (Braga, Papachristos,

and Hureau, 2010) tend to be stable at such places over time; and that police attention to

Volume 16 � Issue 2 393

Research Article Impact of Police on CJ Reform

these places has been shown to have a significant impact on crime across several different

research settings (Braga, Papachristos, and Hureau, 2014). Also, the outcomes of evaluations of place-based interventions reveal that most interventions are successful across a variety of

place types (e.g., residential, public sidewalks, retail, and transport) through the use of

various techniques that target different types of crimes (Eck, 2002; Eck and Guerette,

2012). In an updated systematic review of the effects of hot-spots policing, Braga et al. (2014) found that police interventions designed to alter place characteristics and dynamics

that create opportunities for crime (i.e., problem-oriented police interventions) produced

larger crime prevention benefits than did simply increasing police visibility and making

additional arrests. Thus, comprehensive place-based strategies can address persistent crime problems without an overreliance on arrests.

In their consideration of how we can reduce both crime and incarceration simulta-

neously, Durlauf and Nagin (2011: 40) concluded that the types of police deployment

strategies that could achieve this goal are those that “result in large and visible shifts in apprehension risk” because these are “the mostly likely to have deterrent effects that are

large enough to reduce not only crime but also apprehensions.” They specifically called

attention to hot-spots policing strategies, problem-oriented policing strategies that target enforcement on selected high-crime people or places, and strategies that combine deterrence

and reintegration tactics for high-crime groups as examples of policing strategies that can

reduce both crime and incarceration. As Braga noted (2016: 234), “such strategies can

simultaneously reduce crime and imprisonment by increasing potential offenders’ perceived risk of apprehension and by reducing their opportunities to commit crimes.”

The evolution in policing strategies in Cincinnati has been previously documented (e.g.,

see Eck, 2014; Eck and Rothman, 2006; Engel and Eck, 2015; Sabel and Simon, 2016), and

even though none of these descriptions rise to the level of causal evidence that the strategies and tactics implemented reduced crime or the use of arrest, the summary descriptions and

limited empirical evidence across researchers collectively provides a compelling case that

these strategies were at a minimum correlated with reductions in crime and arrests. Even

in the absence of more rigorous testing, it is clear that the CPD significantly changed its core operational strategies and practices, and that these efforts were accelerated as a

direct result of the jail closure. CPD commanders were routinely quoted (both at the

time and in retrospect) saying they could not “arrest their way out of the problem,”

which is a reference to the ineffective use of aggressive arrest-based strategies to suppress crime. Rather, the common mantra among CPD leaders was to encourage officers to

do more with less, work smarter, consider alternatives to arrest, and promote evidence-

based and data-driven strategies (Coolidge, 2009; Seabook, 2009). These messages were

routinely echoed to mid-level managers and street-level supervisors. In particular, after the jail closure in December 2008, the goal of the agency was to continue to reduce crime

but without the use of aggressive arrest strategies that had contributed to community

distrust.

394 Criminology & Public Policy

Engel, Corsaro, and Ozer

Although the results of prior research indicate that proactive or “aggressive” types

of policing strategies may increase arrests or further strain police–community relations (Brunson, 2007; Fagan, Geller, Davies, and West, 2010; Gau and Brunson, 2010; Weitzer,

1999), Braga (2016: 234) suggested that focused policing can avoid indiscriminate and

overly aggressive policing tactics by engaging a community problem-solving orientation. As

he noted, how police respond to high-crime places matters and “a growing body of research evidence . . . suggests that ‘high risk places, high-risk people’ strategies can actually improve

police legitimacy when they are coupled with a strong commitment to community part-

nership.” Engel and Eck (2015) made a similar argument while debunking the perception

that a trade-off between effective and equitable policing is inevitable. Indeed, hot-spots policing, problem-oriented policing, and focused deterrence with social service and com-

munity components were the specific strategies used by the CPD during the time period

examined and were associated with decreases in both arrests and crime in the face of a

significant reduction in available jail space. Even though other agencies across the country continued to review arrests and citations as part of Compstat-type measures of productivity

or outputs, in Cincinnati, arrests were viewed as specific outcomes for individual cases.8 As

documented by Sabel and Simon (2016: 195), “while arrests are part of a problem-oriented strategy [in Cincinnati], they are used [by CPD] as a last resort and applied in as precisely

targeted a way as possible.”

Conclusion In their review of federal and court system data for the prior 10–15 years, VanNostrand and

Keebler (2007: 23) concluded that “in our society liberty is not the norm and detention prior to trial or without trial is not the carefully limited exception.” How we got here may be widely debated, but certain pathways forward should be clear. The role of police

in guiding changes in the pretrial stages and other criminal justice reform efforts can

be accomplished through the implementation of evidence-based policing strategies and

tactics that seek to reduce crime while reducing the use of arrest. When arrest becomes systematically viewed by police as a limited and precious commodity, to be used sparingly

and for the most chronic or serious offenders, change throughout the criminal justice system

will likely result. When police use focused strategies that enhance deterrence, address the

root causes of problems, and address specific problem places, rather than entire high-crime neighborhoods, both reductions in crime and incarceration are possible. And when data

analytics are emphasized, including social network analysis and other promising innovative

techniques, strategic precision in the use of arrest can be achieved (Papachristos, Hureau,

8. For example, although crimes were routinely reported and discussed at weekly crime analysis meetings, aggregate numbers of arrests were not; it was not until a new chief was hired in August 2011 that arrest counts were even reported in aggregate form or included as part of the discussion of crime analysis meetings.

Volume 16 � Issue 2 395

Research Article Impact of Police on CJ Reform

and Braga, 2013). Most recently, Engel and Eck (2015) argued that policing strategies

should be considered evidence based only if they are effective at reducing crime, efficiently make use of limited resources, and are perceived as equitable by the community. Based on the findings from this natural experiment in Cincinnati, our recommendation for the most

promising police strategies to reduce crime while reducing the number of arrests includes

focused deterrence, place-based policing, and problem-solving crime prevention efforts. Note that our study has several constraints that limit our interpretation of the findings.

First, data from the HCSO were not available; we therefore relied on an examination of

reported crimes and arrests conducted by the CPD, which is the largest police agency in

the county. As such, we have not specifically measured the ways in which the HCSO made jail retention decisions or other areas related to case processing of arrestees. We also cannot

determine causal inference regarding the crime–arrest relationship; rather, we can simply

note that both reported crimes and arrests continued to decrease after the jail closure. Finally,

we cannot directly link the number of arrests to particular policing strategies. We describe what strategies were used by the CPD, and we infer that these strategies had a direct impact

on the number of arrests made by officers, yet this remains an untested assumption.

Collectively, the experiences in Cincinnati demonstrate the important role that police can, and should, play in criminal justice system reform efforts. Through the forced imme-

diate change in the availability of jail space, law enforcement officials in Hamilton County

learned that public safety can still be enhanced when arrests are limited and strategically

focused. Limiting the flow of arrestees into local jails will naturally impact the issues sur- rounding pretrial justice and, ultimately, the entire criminal justice system. Furthermore,

at every stage in the criminal justice system, strategic use of limited resources guided by

research could enhance public safety. More does not mean better.

During this time of growing controversy regarding police practices and community relations, concerns continue to be raised about the disproportionate use of aggressive polic-

ing tactics and use of force against minority citizens. Yet as anxieties regarding police bias

dominate discussions about crime and justice, concerns about police safety are also at the

forefront of our collective attention. The combined political, media, and public attention focused on policing right now provides an important opportunity to reinvest in reform

efforts across the criminal justice system. This reinvestment should emphasize a change in

the way police view arrest—as an outcome, not as an output—and further as an outcome

that should be limited when possible. After innovative strategies are developed and imple- mented, it is critical that police seek to improve their communication with other criminal

justice officials. Information shared by the police with officials involved in pretrial deci-

sions will enhance pretrial risk assessment processes. If the ultimate goal for pretrial justice

is to focus limited resources on the most dangerous offenders, then the adoption of fo- cused, evidence-based policing strategies, coupled with enhanced communication between

police and other criminal justice officials during the pretrial process, will help achieve that

goal.

396 Criminology & Public Policy

Engel, Corsaro, and Ozer

Appendix: Sensitivity Tests on Felony Arrests After Hamilton County Jail Closure

Parameter Lead + 2 Lead + 1 Intervention To Lag – 1 Lag – 2 b SE b SE b SE b SE b SE

Postclosure −.11* .05 −.11* .05 −.13* .06 −.14* .06 −.15* .04 AR (1) .57* .07 .56* .07 .34* .07 .55* .07 .52* .07 AR (12) .35* .07 .36* .07 .28* .08 .35* .07 .37* .07 Wald X2 283.19 313.85 266.07 256.50 273.76

Note. AR = autoregressive term. *p < .05.

References Austin, James, Todd R. Clear, Troy Duster, David F. Greenberg, John Irwin, Candace

McCoy, Alan Mobley, Barbara Owen, and Joshua Page. 2007. Unlocking Amer- ica: Why and How to Reduce America’s Prison Population. Washington, DC: JFA Institute.

Bechtel, Kristen, Christopher T. Lowenkamp, and Alex Holsinger. 2011. Identifying the predictors of pretrial failure: A meta-analysis. Federal Probation, 75: 78–87.

Bittner, Egon. 1967. The police on skid-row: A study of peace keeping. American Journal of Sociology, 32: 699–715.

Blumstein, Alfred and Richard Rosenfeld. 1998. Explaining recent trends in US homicide rates. Journal of Criminal Law and Criminology, 88: 1175–1216.

Box, George E. and Gwilym M. Jenkins. 1976. Time Series Analysis: Forecasting and Control. San Francisco, CA: Holden-Day.

Braga, Anthony A. 2008. Problem-Oriented Policing and Crime Prevention, 2nd Edition. Boulder, CO: Lynne Rienner.

Braga, Anthony A. 2016. Better policing can improve legitimacy and reduce mass incarcer- ation. Harvard Law Review Forum, 129: 233–241.

Braga, Anthony A., David M. Kennedy, Elin J. Waring, and Anne Morrison Piehl. 2001. Problem-oriented policing, deterrence, and youth violence: An evaluation of Boston’s Operation Ceasefire. Journal of Research in Crime & Delinquency, 38: 195–226.

Braga, Anthony A., Andrew V. Papachristos, and David M. Hureau. 2010. The concen- tration and stability of gun violence at micro-places in Boston, 1980–2008. Journal of Quantitative Criminology, 26: 33–53.

Braga, Anthony A., Andrew V. Papachristos, and David M. Hureau. 2014. The effects of hot spots policing on crime: An updated systematic review and meta-analysis. Justice Quarterly, 31: 633–663.

Braga, Anthony A. and David L. Weisburd. 2012. The effects of focused deterrence strategies on crime: A systematic review and meta-analysis of the empirical evidence. Journal of Research in Crime and Delinquency, 49: 323–358.

Volume 16 � Issue 2 397

Research Article Impact of Police on CJ Reform

Bronson, Peter. 2009. Jail space crisis raises specter of years-ago violence returning. Cincin- nati Enquirer. January 9.

Brown, Jessica. 2008. County shuts doors to Queensgate jail. Cincinnati Enquirer. December 20.

Brunson, Rod K. 2007. “Police don’t like Black people”: African-American young men’s accumulated police experiences. Criminology & Public Policy, 6: 71–101.

Clear, Todd R. 2011. A private-sector, incentives-based model for justice reinvest- ment: Justice reinvestment. Criminology & Public Policy, 10: 585–608. doi: 10.1111/j.1745-9133.2011.00729.x.

Cook, Philip J. and John MacDonald. 2011. Public safety through private action: An economic assessment of BIDS. The Economic Journal, 121: 445–462.

Cook, Thomas D. and Donald T. Campbell. 1979. Quasi-experimentation: Design and Analysis Issues for Field Settings. Boston, MA: Houghton Mifflin.

Coolidge, Sharon. 2009. Jail beds reserved for worst. Cincinnati Enquirer. May 24.

Cullen, Francis T., Gregory A. Clark, and John F. Wozniak. 1985. Explaining the get tough movement: Can the public be blamed? Federal Probation, 49: 16–24.

Darlington, Richard. 1990. Regression and Linear Models. New York: McGraw-Hill.

Durlauf, Steven N. and Daniel S. Nagin. 2011. Imprisonment and crime: Can both be reduced? Criminology & Public Policy, 10: 13–54. Retrieved from doi.org/10.1111/j.1745-9133.2010.00680.x.

Eck, John E. 2002. Preventing crime at places. In (Lawrence W. Sherman, David P. Farrington, and Brandon Welsh, eds.), Evidence-Based Crime Prevention. New York: Routledge.

Eck, John E. 2014. The Status of Collaborative Problem Solving and Community Problem-Oriented Policing in Cincinnati. Cincinnati, OH: University of Cincinnati. Retrieved from uc.edu/content/dam/uc/ccjr/docs/reports/The%20 Status%20of%20Collaborative%20Problem%20Solving%206.pdf.

Eck, John E. and Rob T. Guerette. 2012. Place-based crime prevention: Theory, evidence, and policy. In (Brandon C. Welsh and David P. Farrington, eds.), The Oxford Handbook of Crime Prevention. New York: Oxford University Press.

Eck, John E. and Jay Rothman. 2006. Police-community conflict and crime prevention in Cincinnati, Ohio. In (John Bailey and Lucia Dammert, eds.), Public Security and Police Reform in the Americas. Pittsburgh, PA: University of Pittsburgh Press.

Eck, John E. and William Spelman. 1987. Newport News Tests Problem-Oriented Policing. Washington, DC: U.S. Department of Justice, National Institute of Justice.

Elgazzar, Kareem. 2013. Hamilton County Sheriff Jim Neil ends arrest-and-release practice at jail [TV news report]. WCPO Cincinnati. March 20. Retrieved from wcpo.com/news/region-central-cincinnati/hamilton-county-sheriff-jim-neil-ends- arrest-and-realease-practice-at-jail.

Engel, Robin S. and John E. Eck. 2015. Effectiveness vs. Equity in Policing: Is a Trade- off Inevitable? Ideas in American Policing Lecture Series. Washington, DC: Police Foundation.

398 Criminology & Public Policy

Engel, Corsaro, and Ozer

Engel, Robin S. and Kristen Swartz. 2014. Race, crime, and policing. In (Sandra M. Bucerius and Michael Tonry, eds.), Oxford Handbook of Ethnicity, Crime, and Immigration. New York: Oxford University Press.

Engel, Robin S., Marie S. Tillyer and Nicholas Corsaro. 2013. Reducing gang violence using focused deterrence: Evaluating the Cincinnati Initiative to Reduce Violence (CIRV). Justice Quarterly, 30: 403–439.

Engel, Robin S. and James L. Whalen. 2010. Police-academic partnerships: Ending the dialogue of the deaf, the Cincinnati Experience. Police Practice and Research, 11: 105– 116.

Fagan, Jeffrey and Garth Davies. 2000. Street stops and broken windows: Terry, race and disorder in New York City. Fordham Urban Law Journal, 28: 457–504.

Fagan, Jeffrey, Amanda Geller, Garth Davies, and Valerie West. 2010. Street stops and broken windows revisited: The demography and logic of proactive policing in a safe and changing city. In (Stephen K. Rice and Michael D. White, eds.), Race, Ethnicity, and Policing. New York: New York University Press.

Gardner, Everette S., Jr. 1979. Box-Jenkins vs multiple regression: Some adventures in forecasting the demand for blood tests. Interfaces, 9: 49–54.

Gau, Jacinta M. and Rod K. Brunson. 2010. Procedural justice and order maintenance policing: A study of inner-city young men’s perceptions of police legitimacy. Justice Quarterly, 27: 255–279.

Godown, Jeff. 2009. The CompStat process: Four principles for managing crime reduction. The Police Chief, 76.

Goldkamp, John S. and Michael D. White. 2006. Restoring accountability in pretrial re- lease: The Philadelphia pretrial release supervision experiments. Journal of Experimental Criminology, 2: 143–181.

Goldstein, Herman. 1990. Problem-Oriented Policing. Philadelphia, PA: Temple University Press.

Green, Saul. 2008. City of Cincinnati Independent Monitor’s Final Report. Retrieved from cincinnati-oh.gov/police/linkservid/97D9709F-F1C1-4A75-804C07D9873DC 70F/showMeta/0/.

Greenberg, David F. and Ronald C. Kessler. 1982. The effect of arrests on crime: A multivariate panel analysis. Social Forces, 60: 771–790.

Greene, Judith A. 1999. Zero tolerance: A case study of police policies and practices in New York City. Crime & Delinquency, 45: 171–187.

Harcourt, Bernard and Jens Ludwig. 2006. Broken windows: New evidence from New York City and a five-city experiment. University of Chicago Law Review, 73: 271–320.

International Association of Chiefs of Police (IACP). 2011. Law Enforcement’s Leadership Role in the Pretrial Release and Detention Process. Washington, DC: Author.

Kennedy, David M. 2009. Deterrence and Crime Prevention: Reconsidering the Prospect of Sanction, vol. 2. London, U.K.: Routledge.

Kleck, Gary. 1997. Targeting Guns: Firearms and Their Control. New York: Aldine.

Volume 16 � Issue 2 399

Research Article Impact of Police on CJ Reform

Kochel, Tammy R., David B. Wilson, and Stephen D. Mastrofski. 2011. Effect of suspect race on officers’ arrest decisions. Criminology, 49: 473–512.

Land, Kenneth C., Patricia L. McCall, and Lawrence E. Cohen. 1990. Structural covariates of homicide rates: Are there any invariances across time and social space? American Journal of Sociology, 95: 922–963.

Latessa, Edward J. and Brian Lovins. 2010. The role of offender risk assessment: A policy maker guide. Victims and Offenders, 5: 203–219.

Law Enforcement Leaders to Reduce Crime and Incarceration. n.d. We can reduce unneces- sary incarceration and crime [Home page]. Retrieved from lawenforcementleaders.org/.

Leis, Simon. 2007. Multiple overcrowding crises make passing anti-crime plan imperative. Cincinnati Enquirer. June 9.

Leis, Simon and Phil Heimlich. 2006. Experts support sheriff’s plan for jail. Cincinnati Enquirer. July 22.

Lipsky, Michael. 1980. Street-level Bureaucracy: Dilemmas of the Individual in Public Services. New York: Russell Sage Foundation.

Lowenkamp, Christopher T., Edward J. Latessa, and Alexander M. Holsinger. 2006. The risk principle in action: What have we learned from 13,676 offenders and 97 correctional programs? Crime & Delinquency, 52: 77–93.

Lowenkamp, Christopher T. and Marie VanNostrand. 2013. Exploring the Impact of Supervision on Pretrial Outcomes. Houston, TX: Laura and John Arnold Foundation.

Lowenkamp, Christopher T., Marie VanNostrand, and Alexander M. Holsinger. 2013. The Hidden Costs of Pretrial Detention. New York: Laura and John Arnold Foundation.

Mamalian, Cynthia A. 2011. State of the Science of Pretrial Risk Assessment. Gaithersburg, MD: Pretrial Justice Institute.

McCall, Patricia L., Kenneth C. Land, and Karen F. Parker. 2011. Heterogeneity in the rise and decline of city-level homicide rates, 1976–2005: A latent trajectory analysis. Social Science Research, 40: 363–378.

McCarthy, Garry F. 2015. Ohio task force on community-police relations. In (Charles H. Ramsey and Laurie O. Robinson, co-chairs), Taskforce on 21st Century Policing. Cincinnati, OH.

Moore, Mark H. 2002. Recognizing Value in Policing: The Challenge of Measuring Police Performance. Washington, DC: Police Executive Research Forum.

Nagin, Daniel S., Robert M. Solow, and Cynthia Lum. 2015. Deterrence, criminal oppor- tunities, and the police. Criminology, 53: 74–100.

Papachristos, Andrew V., David M. Hureau, and Anthony A. Braga. 2013. The corner and the crew: The influence of geography and social networks on gang violence. American Sociological Review, 78: 417–447.

Papachristos, Andrew V., Tracey L. Meares, and Jeffery Fagan. 2007. Attention felons: Evaluating Project Safe Neighborhoods in Chicago. Journal of Empirical Legal Studies, 4: 223–272.

400 Criminology & Public Policy

Engel, Corsaro, and Ozer

Piehl, Anne M., Suzanne J. Cooper, Anthony A. Braga, and David M. Kennedy. 2003. Testing for structural breaks in the evaluation of programs. Review of Economics and Statistics, 85: 550–558. doi: 10.1162/003465303322369713.

President’s Task Force on 21st Century Policing. 2015. Final Report of the President’s Task Force on 21st Century Policing. Washington, DC: Office of Community Oriented Policing Services.

Pretrial Justice Institute (PTI). 2011. National Symposium on Pretrial Justice: Summary Report of Proceedings (Grant No. 2010-DB-BX-K034). Washington, DC: U.S. Department of Justice, Office of Justice Programs, Pretrial Justice Institute.

Ridgeway, Greg, Terry L. Schell, Brian Gifford, Jessica Saunders, et al. 2009. Police Com- munity Relations in Cincinnati. Santa Monica, CA: RAND.

Riley, K. Jack, Susan Turner, John MacDonald, Greg Ridgeway et al. 2005. Police- Community Relations in Cincinnati. Santa Monica, CA: RAND.

Riksheim, Eric and Steven M. Chermak. 1993. Causes of police behavior revisited. Journal of Criminal Justice, 21: 353–382.

Sabel, Charles F. and William H. Simon. 2016. The duty of responsible administration and the problem of police accountability. Yale Journal on Regulation, 33: 165–211.

Sacks, Meghan and Alissa R. Ackerman. 2014. Bail and sentencing: Does pretrial detention lead to harsher punishment? Criminal Justice Policy Review, 25: 59–77.

Schatmeier, Elliot H. 2012. Reforming police use-of-force practices: A case study of the Cincinnati police department. Columbia Journal of Law and Social Problems, 46: 539– 586.

Seabrook, John. 2009. Don’t shoot: A radical approach to the problem of gang violence. The New Yorker. June 22.

Sherman, Lawrence W. 1980. Causes of police behavior: The current state of quantitative research. Journal of Research in Crime and Delinquency, 17: 69–100.

Sherman, Lawrence W., Patrick R. Gartin, and Michael E. Buerger. 1989. Hot spots of predatory crime: Routine activities and the criminology of place. Criminology, 27: 27–56.

Shi, Lan. 2009. The limit of oversight in policing: Evidence from the 2001 Cincinnati riot. Journal of Public Economics, 93: 99–113.

Sourcebook of Criminal Justice Statistics. 2011. Table 6.1.2011: Adults on probation, in jail or prison, and on parole. Retrieved from albany.edu/sourcebook.

Sparrow, Malcolm, Mark H. Moore, and David M. Kennedy. 1990. Beyond 911: A New Era for Policing. New York: Basic Books.

The New York Times. 2015. Will crime rise if more people are kept out of prison? [Opinion page]. October 29. Retrieved from nytimes.com/roomfordebate/ 2015/10/29/will-crime-rise-if-more-people-are-kept-out-of-prison.

VanNostrand, Marie and Gena Keebler. 2007. Our journey toward pretrial justice. Federal Probation, 71: 20–25.

VanNostrand, Marie and Gena Keebler. 2009. Pretrial risk assessment in the federal court. Federal Probation, 73: 3–29.

Volume 16 � Issue 2 401

Research Article Impact of Police on CJ Reform

Weisburd, David, Shawn Bushway, Cynthia Lum, and Sue-Ming Yang. 2004. Trajectories of crime at places: A longitudinal study of street segments in the city of Seattle. Criminology, 42: 283–321.

Weisburd, David, Stephen D. Mastrofski, Ann Marie McNally, Rosann Greenspan, and James J. Willis. 2003. Reforming to preserve: Compstat and strategic problem solving in American policing, Criminology & Public Policy, 2: 421–456.

Weitzer, Ronald. 1999. Citizens’ perceptions of police misconduct: Race and neighborhood context. Justice Quarterly, 16: 819–846.

Williams, Marian R. 2003. The effect of pretrial detention on imprisonment decisions. Criminal Justice Review, 28: 299–316.

Wilson, James Q. and George L. Kelling. 1982. Broken windows: The police and neigh- borhood safety. The Atlantic Magazine. March, 29–38.

Wooldredge, John. 2012. Distinguishing race effects on pre-trial release and sentencing decisions. Justice Quarterly, 29: 41–75.

Robin S. Engel is vice president for safety reform at the University of Cincinnati and director of the International Association of Chiefs of Police (IACP)/University of Cincinnati

Center for Police Research and Policy. Her research includes empirical assessments of police

behavior, police/minority relations, police supervision and management, police agency reform, criminal justice policies, criminal gangs, and violence reduction strategies. Her

previous research has appeared in Criminology, Justice Quarterly, Journal of Research in Crime and Delinquency, Journal of Criminal Justice, Crime & Delinquency, and Criminology & Public Policy.

Nicholas Corsaro is an associate professor in the School of Criminal Justice and Director of the Institute of Crime Science at the University of Cincinnati. His research focuses on

working with police agencies to develop efficient and effective approaches to crime preven- tion, problem analysis, and urban violence and drug markets. His previous research appears

in Crime & Delinquency, Criminology & Public Policy, Journal of Criminal Justice, Journal of Experimental Criminology, Journal of Quantitative Criminology, and Justice Quarterly. He received his Ph.D. from the School of Criminal Justice at Michigan State University.

M. Murat Ozer is a senior research associate and director of data analytics in the Institute of Crime Science at the University of Cincinnati. His research interests include developing

early warning systems comprising data analytics, violence reduction strategies, social net- work analysis, anti-terrorism policies, and data-driven crime prevention efforts in policing

agencies. His previous research has appeared in Justice Quarterly and Journal of Criminal Justice.

402 Criminology & Public Policy

Copyright of Criminology & Public Policy is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.