An Evaluation of Compstat’s Effect on Crime: The Fort Worth Experience

• Author: Larry T. Hoover,Hyunseok Jang,Hee-Jong Joo
• Published date: 01 December 2010
• Date: 01 December 2010
• DOI: 10.1177/1098611110384085
• Subject Matter: Articles

13410.1177/109861
1110384085Jang et al.Police Quarterly
© 2010 SAGE Publications
Reprints and permission: http://www.
sagepub.com/journalsPermissions.nav
Police Quarterly
13(4) 387 –412
An Evaluation of
© 2010 SAGE Publications
Reprints and permission: http://www.
Compstat’s Effect
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1098611110384085
http://pqx.sagepub.com
on Crime:  The Fort
Worth Experience
Hyunseok Jang, PhD1, Larry T. Hoover, PhD2,
and Hee-Jong Joo, PhD2
Abstract
Compstat as a policing strategy became popular fol owing the significant crime reduction
in New York City during the 1990s. As an innovative management strategy in policing,
Compstat attracted considerable attention from scholars and police practitioners. Despite
its popularity, little empirical research has scientifical y evaluated the effectiveness of the
Compstat strategy. In addition, few studies have concentrated on Compstat strategies
implemented during the 2000s outside New York City. This study examines the effective-
ness of Compstat as implemented by the Fort Worth (Texas) Police Department (FWPD).
Using monthly time-series arrest and crime data over a multiyear period, the study
examines whether Compstat engendered a significant increase in “broken windows”
arrests (minor nuisance offenses) and, using multivariate time-series analysis, the role of
the Compstat strategy in explaining changes in violent, property, and total index crimes.
Findings indicate that the implementation of Compstat significantly increased some types
of broken windows arrests in the FWPD whereas others decreased. Analysis indicates
significant decreases in property and total index crime rates after control ing for rival
factors, but fails to show a significant change in violent crime rates. If the Fort Worth
strategic approach to Compstat had to be described with a single word, it would be
focusing. The Queensland study of Compstat also reported using a problem-oriented
intervention model—focusing—in lieu of a broken windows approach (Mazerolle,
Rombouts, & McBroom, 2007). Property crime was significantly reduced in both settings.
Paral el findings from two differently constituted Compstat programs on two different
continents provides evidence that the primary component of the Compstat model is
focusing, not broken windows enforcement, and the primary impact is on property crime.
1Missouri Western State University, Saint Joseph, MO
2Sam Houston State University, Huntsville, TX
Corresponding Author:
Larry T. Hoover, College of Criminal Justice, Sam Houston State University, Huntsville, TX 77341
Email: larryhoover@shsu.edu

---

388
Police Quarterly 13(4)
Keywords
Compstat policing, broken windows enforcement, time-series analysis, Fort Worth
Police Department
Introduction
Following community-oriented or problem-solving strategies, Compstat, an alternative
policing management strategy, has drawn significant attention from scholars and police
practitioners. The crime decline in New York City following the introduction of Compstat
fueled heated debates regarding the influence of this crime-focused strategy on crime
rates since the 1990s (Eck & Maguire, 2000; Harcourt, 1998, 2001; Harcourt & Ludwig,
2006; Kelling & Sousa, 2001; Rosenfeld, Fornango, & Baumer, 2005; Willis, Mastrofski,
& Weisburd, 2004). Partially due to the well-known success stories in New York City
and subsequently several other metropolitan jurisdictions, the Compstat strategy quickly
spread throughout the United States and has been characterized by some as the most
popular police management and crime control strategy in American history (Weisburd,
Mastrofski, McNally, Greenspan, & Willis, 2003).
In a nationwide survey conducted in 1999, for example, one third of municipal police
departments with more than 100 sworn officers reported that they had implemented a
“Compstat-like program” (Weisburd et al., 2003, p. 430). As such, the Compstat strategy
became one of “the most quickly diffused forms of innovation” (Weisburd et al., 2003,
p. 433) and was described as an organizational paradigm shift in policing due to the
swiftness of its effect on crime prevention or reduction (Henry, 2002; Walsh, 2001).
Despite Compstat’s popularity, little empirical research has scientifically evaluated
the effectiveness of the Compstat strategy. A number of the reviews of Compstat in
New York City have been in the form of “advocacy” books which assume effectiveness
(see, for example, Bratton & Knobler, 1998; Henry, 2002; Maple & Mitchell, 1999).
A majority of the existing research has been anecdotal and qualitative, thus lacking
generalizability. In addition, a few studies that measured the effectiveness of Compstat
in New York City found conflicting results concerning the causal relationship between
the Compstat strategy and the decline in crime rates (Harcourt, 1998, 2001; Harcourt
& Ludwig, 2006; Kelling & Sousa, 2001; Rosenfeld et al., 2005; Willis et al., 2004).
It is also important to recognize the potential role of other social structural variables
in the reduction of crime rates in New York. Fagan, Zimring, and Kim (1998) reported
that nonfirearm homicides began to decline in New York City prior to the implementa-
tion of the Compstat strategy. In relation to firearm homicides, however, they found
consistent timing between Compstat and the homicide decline, leading them to conclude
that the Compstat strategy may have contributed to the lower firearm homicide rate.
Karmen (2000) examined the crime drop in New York City and concluded that the
Compstat policing strategy may have influenced the reduction of crime. Karmen, how-
ever, argued that other factors such as an economic boom, escalated incarceration, change
in drug markets, and the change in demographic composition may have played greater

---

Jang et al.
389
roles than the police strategy. Rosenfeld et al. (2005) also evaluated the influences of
Compstat on the reduction of homicide rates in New York City compared with other
large U.S. cities and found no significant differences. They argued that the reduction of
homicide rates could be better explained by other social structural variables as most U.S.
cities experienced similar reductions in homicide. Due to the inconsistent results of
previous research, it has been difficult to develop a conclusion about the effectiveness
of the New York City Compstat strategy in reducing crime rates. Further research is
necessary to scientifically assess Compstat’s impact on crime rates.
Considering the widespread adoption of Compstat strategies throughout the nation,
it is important to examine the effectiveness of strategies implemented outside of the
New York City Police Department (NYPD). The Fort Worth Police Department (FWPD)
in Texas implemented a Compstat strategy to address crime and quality of life issues in
the early 2000s. While nationwide crime rates began to decline during the 1990s, crime
rates reached a plateau in the early 2000s. Therefore, the evaluation of a Compstat strategy
implemented during the 2000s provides important information to validate the effective-
ness of Compstat in a different time period as well as in a different place. Considering
the differences in geographic location, cultural background, organizational subculture,
and study period, this study attempted to determine if any significant differences existed
in effects as well as strategic approaches between the NYPD’s Compstat and the FWPD’s
“All Staff Meeting” (ASM) version of Compstat.
Using monthly time-series data and employing a time-series intervention analysis
and a multivariate time-series analysis (regression with autoregressive error modeling),
this study examines the relationship between the implementation of Compstat in the
FWPD and misdemeanor arrest rates as well as changes in index crime rates. There are
two major research questions: (a) Was there a significant increase in the arrest rate for
broken windows offenses (i.e., minor nuisance offenses) in the FWPD following the
implementation of the Compstat strategy? and (b) Was there a significant change in crime
rates (i.e., Part I violent, property, and total crime) after Fort Worth implemented the
Compstat strategy?
Literature Review
Compstat Strategy
The term Compstat originated from “Compare Stats,” a computer file name (Silverman,
1999, p. 98). Similarly, according to Safir (1997), Compstat was short for computer
comparison statistics. The term was abbreviated to the eight-character file name limita-
tion in DOS programming at the time, “Compstat.” Although the term appeared to
emphasize a technical aspect of operational strategies, Compstat became the shorthand
descriptor of a broad management strategy involving the entire police agency’s operation
and consisting of several elements (Bratton & Knobler, 1998). NYPD’s commissioner,
Bratton, and his executive team established four basic principles consisting of (a) accurate
and timely intelligence, (b) rapid deployment of personnel and resources, (c) effective

---

390
Police Quarterly 13(4)
tactics, and (d) relentless follow-up and assessment (Bratton & Knobler, 1998; Silverman,
1999; Walsh & Vito, 2004). In conclusion, Hoover (2004a) defined Compstat as “a com-
bination of a strategy and a management style” (p. 1) that emphasized the specifica-
tion of problems in a certain area and time, requiring tailor-made solutions for specific
problems. As the impact of Compstat was so broad and swift in law enforcement...