Broken Windows Theory and Policing traces how Wilson and Kelling’s 1982 essay was translated from a magazine argument into an operational policing doctrine, examining the specific institutional mechanisms, patrol strategies, and administrative reforms through which American police departments implemented disorder-focused policing between the late 1980s and the present. Where other articles in this silo examine the theory’s causal claims or its application to specific tactics such as zero-tolerance enforcement and stop-and-frisk, this article addresses the broader policing doctrine itself: how departments reorganized patrol assignments, retrained officers, and restructured performance metrics around order maintenance, and what the accumulated evaluation evidence says about the resulting crime and disorder effects.
Broken Windows Theory and Policing sits between the theoretical foundation established in Wilson and Kelling’s Broken Windows Article and Core Principles of Broken Windows Theory and the more specific tactical treatments found in Order Maintenance Policing, Zero Tolerance Policing and Broken Windows, and Stop-and-Frisk and Broken Windows Policing. The distinction matters because broken windows policing has never been a single, uniform practice; departments implementing the same underlying theoretical premise adopted markedly different operational approaches, ranging from community-relationship-based order maintenance to aggressive, arrest-heavy quality-of-life enforcement, with correspondingly different consequences for crime, civil liberties, and police-community relations.
This article traces that institutional translation chronologically, from the doctrine’s earliest adoption in transit and municipal policing through its national diffusion, before turning to the accumulated body of evaluation evidence, which, as in the broader disorder-crime literature examined in Physical Disorder and Crime and Social Disorder and Crime, offers a considerably more qualified picture of effectiveness than the doctrine’s early proponents claimed.
From Order Maintenance to Policing Doctrine
Institutionalizing Wilson and Kelling’s Prescriptions
Wilson and Kelling’s essay proposed a policing philosophy rather than a specific operational protocol, leaving individual departments to translate its order-maintenance premise into concrete patrol assignments, training curricula, and supervisory expectations. George Kelling himself played a direct role in this institutional translation, serving as a consultant to the New York City Transit Police in the late 1980s and early 1990s, where he worked with Chief William Bratton to redesign subway patrol strategy around aggressive enforcement of fare evasion and platform disorder, a collaboration Bratton later described as directly informed by the broken windows framework (Bratton & Knobler, 1998; Kelling & Coles, 1996).
This early transit application proved influential because it offered a bounded, measurable proving ground, a closed transit system with clearly defined disorder indicators, fare evasion, graffiti, and aggressive panhandling, before the same personnel and philosophy moved to the broader NYPD under Bratton’s subsequent tenure as Police Commissioner. The transit experience shaped the specific administrative tools, including systematic disorder tracking and precinct-level accountability meetings, that would later become central to broken windows policing’s institutional character across the full department (Kelling & Coles, 1996).
Kelling and Coles’s (1996) Fixing Broken Windows extended this institutional translation into a more explicit administrative blueprint, proposing specific organizational reforms including revised patrol officer job descriptions emphasizing order maintenance alongside crime response, retrained supervisory expectations rewarding proactive disorder enforcement rather than solely reactive arrest statistics, and community advisory structures intended to calibrate enforcement priorities to locally negotiated disorder standards, as the original 1982 essay had proposed in its discussion of community standards versus uniform legal enforcement.
This administrative blueprint proved considerably more difficult to implement uniformly than its clear prose suggested, since translating community-standards-based discretion into consistent, defensible departmental policy required resolving genuine tensions the original essay had left unaddressed, particularly how to calibrate enforcement thresholds across neighborhoods with differing resident preferences without producing the appearance, or reality, of unequal application of law across a city’s different communities. Individual departments resolved this tension differently, with some establishing formal community advisory boards empowered to set local enforcement priorities and others retaining centralized command discretion while soliciting less formal resident input, a divergence that produced the community-oriented and centrally directed variants examined later in this article.
The Shift from Reactive to Proactive Patrol
Broken windows policing’s institutional adoption coincided with, and substantially accelerated, a broader shift in American policing away from the reactive, 911-response-centered model that had dominated since the professionalization era Kelling and Moore (1988) described, toward proactive patrol strategies in which officers actively sought out and addressed disorder rather than waiting for citizen-initiated calls for service. This shift required substantial retraining, since officers accustomed to responding primarily to dispatched calls needed new skills in identifying disorder conditions, exercising discretion about appropriate intervention, and documenting disorder-related activity for supervisory review.
Departments adopting this proactive model generally restructured patrol assignments around fixed geographic beats rather than citywide rotation, on the theory, consistent with the broken windows causal chain discussed in Core Principles of Broken Windows Theory, that officers with sustained, consistent assignment to a specific area would develop the local knowledge necessary to distinguish genuine disorder from normal neighborhood activity and to build the resident relationships Wilson and Kelling’s essay had identified as central to effective order maintenance (Kelling & Coles, 1996).
Core Elements of Broken-Windows-Informed Policing
Discretionary Enforcement and Officer Judgment
Broken windows policing, across its various implementations, has consistently relied on expanded officer discretion to identify and respond to disorder conditions that do not always constitute clear-cut criminal violations, a design feature directly inherited from Wilson and Kelling’s original argument for community-standards-based, rather than purely statutory, order maintenance. This discretionary structure has proven to be simultaneously the doctrine’s most operationally distinctive feature and its most persistent source of controversy, since discretion that allows officers to address genuine neighborhood-specific disorder also creates the conditions for the disparate enforcement patterns examined in Broken Windows Theory and Race.
Departments have managed this discretion through varying combinations of officer training, supervisory review, and community input mechanisms, with more community-oriented implementations generally emphasizing officer judgment calibrated through sustained resident relationships, and more centralized, metrics-driven implementations emphasizing standardized enforcement thresholds and quantified performance targets, a distinction elaborated further in Order Maintenance Policing and Zero Tolerance Policing and Broken Windows.
Geographic Targeting and Hot Spot Identification
A second core element of broken windows policing involves geographic targeting, concentrating enforcement resources on specific blocks, transit stations, or commercial corridors identified as disorder hot spots rather than distributing patrol resources evenly across an entire precinct or city. This targeting approach draws on the broader hot-spots policing research tradition, which found that crime and disorder are typically concentrated in a small proportion of urban micro-locations rather than distributed evenly across neighborhoods (Sherman, Gartin, & Buerger, 1989; Weisburd & Eck, 2004), a finding with direct implications for how limited police resources should be allocated under a disorder-focused strategy.
Braga and Bond’s (2008) randomized controlled trial in Lowell, Massachusetts, directly tested this geographic targeting logic, assigning matched disorder hot spots to either intensified order-maintenance policing or standard patrol, and finding statistically significant reductions in both disorder and crime calls for service at treated hot spots relative to controls. This study is frequently cited as among the more methodologically rigorous pieces of evidence supporting geographically targeted disorder policing specifically, as distinct from broader claims about citywide order-maintenance strategy, a distinction examined further in What the Research Says About Broken Windows.
The Lowell study’s design also allowed the researchers to disaggregate treated hot spots by intervention type, distinguishing locations receiving primarily social-service and situational-prevention interventions, such as coordination with property owners and community groups, from locations receiving primarily traditional police enforcement activity such as increased arrests and citations. Braga and Bond (2008) found that the social-service and situational-prevention interventions produced larger disorder reductions than pure enforcement activity, an early empirical signal, later reinforced by the meta-analytic findings discussed below, that broken windows policing’s crime-control benefits may derive disproportionately from its problem-solving and environmental components rather than from enforcement volume per se.
Variants of Broken Windows Policing
Community-Oriented Order Maintenance
One major variant of broken windows policing emphasizes sustained officer-resident relationships, problem-solving partnerships with community organizations, and discretionary, warning-based enforcement calibrated to locally negotiated standards rather than centrally mandated enforcement quotas. This community-oriented variant, sometimes explicitly integrated with broader community policing philosophy, treats order maintenance as one component of a collaborative relationship between police and residents rather than as a primarily enforcement-driven activity, and is examined in greater detail in Order Maintenance Policing.
Skogan’s (2006) evaluation of Chicago’s Alternative Policing Strategy, which combined community-oriented order maintenance with problem-solving partnerships, found improvements in resident-reported disorder and modest reductions in fear of crime, though the program’s decentralized, officer-discretion-dependent structure made consistent implementation across Chicago’s many police districts difficult to sustain, illustrating a recurring operational challenge for community-oriented variants relative to more centrally directed enforcement models.
Aggressive Quality-of-Life Enforcement
A contrasting variant, most closely associated with the NYPD’s implementation under Commissioners Bratton and Howard Safir during the 1990s, emphasizes centrally mandated enforcement targets, systematic arrest and summons activity for disorder offenses, and data-driven accountability structures holding precinct commanders responsible for measurable disorder and crime reductions within their commands (Bratton & Knobler, 1998; Silverman, 1999). This variant, examined in greater detail in Broken Windows Theory in New York City and Zero Tolerance Policing and Broken Windows, prioritizes measurable enforcement output and rapid, citywide implementation over the slower, relationship-building emphasis of community-oriented variants.
The aggressive quality-of-life variant’s reliance on centralized performance metrics, most notably the CompStat accountability system Bratton introduced alongside broken windows policing, created strong institutional incentives for high-volume enforcement activity, incentives that critics have argued contributed directly to the disparate stop-and-frisk patterns examined in Stop-and-Frisk and Broken Windows Policing, since precinct commanders under pressure to demonstrate measurable enforcement activity had strong incentives to generate high volumes of low-level stops and summonses regardless of their individualized evidentiary basis (Fagan & Davies, 2000).
Evaluating Broken Windows Policing’s Crime Effects
City-Level Before-After Studies
Early evaluation evidence supporting broken windows policing’s effectiveness relied substantially on city-level, pre-post comparisons, most prominently New York City’s documented crime decline during the 1990s following the NYPD’s adoption of order-maintenance and CompStat-driven policing. Kelling and Sousa (2001) analyzed this decline, finding that precincts implementing more intensive order-maintenance enforcement showed correspondingly larger reductions in serious crime, and interpreted this pattern as direct evidence of broken windows policing’s causal effectiveness.
This city-level, pre-post design faces significant methodological limitations examined at length in What the Research Says About Broken Windows, most importantly the absence of an appropriate comparison group unaffected by the citywide crime decline that many American cities, including those that had not adopted comparable order-maintenance strategies, experienced simultaneously during the 1990s (Levitt, 2004; Zimring, 2011). Zimring’s (2011) comprehensive analysis of New York’s crime decline concluded that while order-maintenance policing likely contributed to the city’s crime reduction, its independent contribution relative to other concurrent factors, including increased police staffing levels, targeted gun enforcement, and broader national crime-decline trends, could not be isolated with confidence using the available city-level data.
Rosenfeld, Fornango, and Rengifo’s (2007) precinct-level analysis attempted to address this comparison-group limitation by examining variation in order-maintenance enforcement intensity across New York’s 75 precincts rather than relying on a single citywide before-after comparison, finding a modest but statistically significant association between precinct-level misdemeanor arrest intensity and subsequent robbery decline, though the association proved considerably weaker for homicide and other serious violent crimes than the strongest early claims for broken windows policing’s effectiveness had suggested. This precinct-level approach represented a methodological improvement over pure citywide comparison, since it could exploit within-city variation in enforcement intensity rather than relying entirely on a single citywide time trend, though it remained vulnerable to the possibility that precincts choosing more intensive enforcement differed systematically from other precincts in ways that independently affected their crime trajectories.
Quasi-Experimental and Randomized Evidence
More recent evaluation research has moved toward quasi-experimental and randomized designs capable of better isolating broken windows policing’s independent effects from confounding citywide trends. Braga, Welsh, and Schnell’s (2015) systematic review and meta-analysis of disorder policing evaluations, synthesizing findings across numerous studies employing varied methodologies, found a statistically significant but modest overall crime-reduction effect associated with disorder policing interventions, with substantially stronger effects for community-oriented and problem-solving variants than for aggressive, arrest-focused zero-tolerance implementations.
This meta-analytic finding, that disorder policing’s effectiveness varies considerably by implementation style rather than representing a uniform effect of the underlying theory, has become an influential conclusion within the contemporary evaluation literature, suggesting that debates framed simply as “does broken windows policing work” obscure more policy-relevant questions about which specific operational variants produce measurable benefits at acceptable civil liberties and community-relations costs (Braga, Welsh, & Schnell, 2015).
Departmental Adoption Beyond New York
Boston, Chicago, and Other Early Adopters
Beyond New York City, several major American police departments adopted broken windows-informed policing strategies during the 1990s and 2000s, often with meaningfully different operational emphases reflecting local political and community contexts. Boston’s Operation Ceasefire, while primarily a focused-deterrence gang violence intervention rather than a pure disorder-policing strategy, incorporated order-maintenance elements and became an influential model demonstrating that disorder-focused and violence-focused policing strategies could be integrated within a single coordinated framework (Braga & Weisburd, 2010).
Chicago’s implementation, examined briefly above through Skogan’s (2006) evaluation, illustrated the community-oriented variant’s characteristic strengths and limitations, achieving meaningful resident satisfaction improvements in districts with sustained implementation while facing significant consistency challenges across a large, decentralized department, a pattern that has recurred in numerous other cities attempting community-oriented order-maintenance strategies at scale.
International Diffusion of the Model
Broken windows-informed policing also diffused beyond the United States, with British, Dutch, and other European police services adopting disorder-focused patrol strategies drawing explicitly on the American broken windows literature, though generally adapted to different legal frameworks governing police discretion and, in the British case, integrated with the distinct signal crimes framework discussed in Fear of Crime and the Incivilities Thesis (Innes, 2004). These international adaptations have generally emphasized the community-oriented and reassurance-focused variants of the doctrine over the more aggressive, enforcement-heavy American implementations, reflecting both differing legal constraints on police discretion and differing political tolerance for the civil liberties costs associated with zero-tolerance enforcement.
Institutional and Professional Consequences
Effects on Police Officer Discretion and Training
Broken windows policing’s institutional adoption produced lasting changes to American police training and supervisory practice beyond its direct crime-control effects, embedding disorder-recognition skills and community-standards-based discretion training into standard police academy curricula in numerous departments (Kelling & Coles, 1996). These training changes have persisted in many departments even as specific enforcement tactics, particularly aggressive stop-and-frisk practices, have faced substantial legal and political retrenchment following litigation examined in Stop-and-Frisk and Broken Windows Policing.
Effects on Police-Community Relations
Broken windows policing’s institutional legacy for police-community relations has proven considerably more contested than its crime-control legacy, with aggressive quality-of-life enforcement variants implicated in the erosion of police legitimacy and community trust documented in numerous studies of over-policed neighborhoods, a consequence examined at length in Broken Windows Theory and Race and Broken Windows Theory and Community Relations. This tension between the doctrine’s crime-control aspirations and its community-relations costs has become a central organizing theme for the contemporary policy debates examined in Alternatives to Broken Windows Policing, which surveys strategies seeking to preserve whatever genuine crime-control benefits disorder-focused policing offers while substantially reducing its documented civil liberties and legitimacy costs.
Several departments that initially adopted aggressive quality-of-life variants during the 1990s and 2000s have since undertaken formal strategic reviews explicitly reconsidering their order-maintenance enforcement posture in light of accumulated legitimacy costs, often prompted by federal consent decrees, civil litigation, or documented community trust deterioration, and generally moving toward the community-oriented variant’s greater emphasis on relationship-building and discretionary, non-arrest-based disorder response. This institutional recalibration, still ongoing in numerous jurisdictions, represents one of the clearest practical consequences of the accumulated evaluation evidence reviewed in this article: departments have increasingly treated the choice between broken windows policing’s operational variants, rather than the underlying disorder-focused premise itself, as the primary lever available for reconciling crime-control goals with community-relations and civil liberties considerations.
Analytical Tables
Table 1. Variants of Broken Windows Policing Compared
| Dimension | Community-Oriented Variant | Aggressive Quality-of-Life Variant |
|---|---|---|
| Primary mechanism | Officer-resident relationship building | Centralized enforcement metrics |
| Enforcement style | Discretionary, warning-based | Systematic arrest and summons activity |
| Accountability structure | Community advisory input | CompStat-style command accountability |
| Representative example | Chicago Alternative Policing Strategy | NYPD under Bratton and Safir |
| Meta-analytic effectiveness | Stronger crime-reduction effect | Weaker, more contested effect |
| Civil liberties risk | Lower | Higher |
Table 2. Evidentiary Basis for Broken Windows Policing Across Study Designs
| Study Design | Example Study | Key Strength | Key Limitation |
|---|---|---|---|
| City-level pre-post comparison | Kelling & Sousa (2001) | Captures large-scale real-world implementation | No comparison group; confounded by national trends |
| Comprehensive historical analysis | Zimring (2011) | Weighs multiple concurrent explanatory factors | Cannot fully isolate independent policing effect |
| Randomized controlled trial | Braga & Bond (2008) | Strong causal inference at hot-spot level | Limited generalizability beyond studied locations |
| Systematic review and meta-analysis | Braga, Welsh & Schnell (2015) | Synthesizes across many studies and designs | Effect estimates sensitive to included study quality |
Conclusion
The institutional history of broken windows policing reveals a doctrine considerably more variable in practice than its unified theoretical origin in Wilson and Kelling’s 1982 essay might suggest, encompassing community-oriented, relationship-based implementations and aggressive, metrics-driven enforcement regimes that have produced meaningfully different crime-control, civil liberties, and community-relations outcomes. The accumulated evaluation evidence, moving from early, methodologically limited city-level comparisons toward more rigorous quasi-experimental and randomized designs, has converged on a qualified conclusion: disorder-focused policing can produce modest, geographically concentrated crime reductions, particularly in its community-oriented and hot-spot-targeted forms, while the more aggressive, enforcement-heavy variants historically associated with the doctrine’s highest-profile implementation carry crime-control benefits that appear smaller and civil liberties costs that appear considerably larger than early proponents claimed.
This variable record has reshaped contemporary policy discussion away from a simple binary question of whether broken windows policing works and toward the more granular, operationally specific questions examined throughout the remainder of this silo: which variants, implemented under which conditions and subject to which accountability structures, achieve genuine public safety benefits without imposing the civil liberties and community-trust costs that the doctrine’s more aggressive historical implementations have generated.
Related Articles
- Order Maintenance Policing
- Zero Tolerance Policing and Broken Windows
- Stop-and-Frisk and Broken Windows Policing
- Broken Windows Theory in New York City
- What the Research Says About Broken Windows
References
- Braga, A. A., & Bond, B. J. (2008). Policing crime and disorder hot spots: A randomized controlled trial. Criminology, 46(3), 577–607.
- Braga, A. A., & Weisburd, D. L. (2010). Policing Problem Places: Crime Hot Spots and Effective Prevention. Oxford University Press.
- Braga, A. A., Welsh, B. C., & Schnell, C. (2015). Can policing disorder reduce crime? A systematic review and meta-analysis. Journal of Research in Crime and Delinquency, 52(4), 567–588.
- Bratton, W., & Knobler, P. (1998). Turnaround: How America’s Top Cop Reversed the Crime Epidemic. Random House.
- Fagan, J., & Davies, G. (2000). Street stops and broken windows: Terry, race, and disorder in New York City. Fordham Urban Law Journal, 28(2), 457–504.
- Innes, M. (2004). Signal crimes and signal disorders: Notes on deviance as communicative action. British Journal of Sociology, 55(3), 335–355.
- Kelling, G. L., & Coles, C. M. (1996). Fixing Broken Windows: Restoring Order and Reducing Crime in Our Communities. Free Press.
- Kelling, G. L., & Moore, M. H. (1988). The evolving strategy of policing. Perspectives on Policing, 4, 1–15.
- Kelling, G. L., & Sousa, W. H. (2001). Do police matter? An analysis of the impact of New York City’s police reforms. Civic Report, 22, Manhattan Institute.
- Levitt, S. D. (2004). Understanding why crime fell in the 1990s: Four factors that explain the decline and six that do not. Journal of Economic Perspectives, 18(1), 163–190.
- Rosenfeld, R., Fornango, R., & Rengifo, A. F. (2007). The impact of order-maintenance policing on New York City homicide and robbery rates: 1988–2001. Criminology, 45(2), 355–384.
- Sherman, L. W., Gartin, P. R., & Buerger, M. E. (1989). Hot spots of predatory crime: Routine activities and the criminology of place. Criminology, 27(1), 27–56.
- Silverman, E. B. (1999). NYPD Battles Crime: Innovative Strategies in Policing. Northeastern University Press.
- Skogan, W. G. (2006). Police and Community in Chicago: A Tale of Three Cities. Oxford University Press.
- Weisburd, D., & Eck, J. E. (2004). What can police do to reduce crime, disorder, and fear? Annals of the American Academy of Political and Social Science, 593(1), 42–65.
- Wilson, J. Q., & Kelling, G. L. (1982). Broken windows: The police and neighborhood safety. The Atlantic Monthly, 249(3), 29–38.
- Zimring, F. E. (2011). The City That Became Safe: New York’s Lessons for Urban Crime and Its Control. Oxford University Press.