Physical Disorder and Crime examines the empirical and theoretical literature on physical incivilities, broken windows, graffiti, abandoned buildings, litter, and vandalized property, and their relationship to neighborhood crime rates within broken windows theory. Distinct from social disorder, which involves observable behavior, physical disorder concerns the visible condition of the built environment itself and is theorized to signal an absence of maintenance, ownership investment, and municipal responsiveness. This article traces how physical disorder has been measured, from Wilson and Kelling’s original impressionistic descriptions through Sampson and Raudenbush’s systematic social observation methodology, and evaluates the empirical evidence linking physical disorder independently to crime outcomes, distinguishing findings that survive adjustment for underlying structural conditions from those that do not.
Physical Disorder and Crime addresses a measurement problem that has proven consequential for the broader empirical status of broken windows theory: because physical disorder and crime frequently co-occur in the same disadvantaged neighborhoods, and because both are plausibly caused by shared underlying conditions such as concentrated poverty, residential instability, and reduced municipal investment, isolating physical disorder’s independent causal contribution to crime, as opposed to its status as a co-symptom of the same structural causes, has required increasingly sophisticated research designs over the four decades since Wilson and Kelling’s foundational essay, discussed in Wilson and Kelling’s Broken Windows Article.
This article should be read alongside Social Disorder and Crime, which addresses the parallel but analytically distinct literature on behavioral incivilities, and Core Principles of Broken Windows Theory, which situates physical disorder within the theory’s broader causal sequence. Where the two disorder types diverge in their measurement, mechanisms, or empirical support, this article flags the distinction explicitly rather than treating physical and social disorder as interchangeable, a conflation that has complicated interpretation of the broader disorder-crime literature.
Defining and Operationalizing Physical Disorder
Categories of Physical Incivility
The empirical literature has generally organized physical disorder into several recurring categories: structural neglect, including abandoned or boarded buildings, broken windows, and deteriorated facades; environmental litter, including trash accumulation, discarded furniture, and abandoned vehicles; vandalism, including graffiti and broken public fixtures; and infrastructural deficiency, including broken streetlights, potholes, and poorly maintained sidewalks (Skogan, 1990; Sampson & Raudenbush, 1999). Ross and Mirowsky (1999) further distinguished disorder attributable to private property owners, such as unrepaired building facades, from disorder attributable to public or municipal responsibility, such as street lighting and sanitation, a distinction with direct implications for which actors bear responsibility for remediation and, correspondingly, which policy levers are available for intervention.
This categorical variation matters empirically because different types of physical disorder appear to carry different signaling weight; graffiti and abandoned buildings, being more visually salient and more clearly indicative of abandonment, have shown stronger associations with residents’ perceived neighborhood safety in several studies than more mundane forms of disorder such as accumulated litter, which residents in some contexts appear to normalize or discount as an ordinary feature of urban life rather than as a signal of danger (Perkins, Meeks, & Taylor, 1992).
Perkins and colleagues’ (1992) block-level research in Baltimore further found that the signaling weight of a given disorder category interacted with block architectural characteristics, such as building setback and street width, that affected how visible the disorder was to passersby and residents alike, suggesting that identical physical conditions may function as stronger or weaker signals depending on the surrounding built environment’s capacity to render that disorder visible. This architectural interaction has informed subsequent work in environmental design connecting physical disorder research to the broader crime prevention through environmental design tradition, treated in CPTED and Environmental Design Approaches to Disorder.
Ross and Mirowsky’s (1999) research further complicated the categorical picture by demonstrating that resident-reported disorder measures and independently observed disorder measures, while correlated, diverged systematically in ways tied to respondent characteristics such as length of residence and homeownership status, with longer-tenured homeowning residents generally reporting lower disorder levels than either newer residents or trained observers recorded for the identical block. This divergence suggests that habituation, the gradual normalization of familiar surroundings, may dampen some residents’ subjective disorder perception over time even as the objective physical conditions remain constant, a possibility with direct implications for how municipal disorder-tracking programs should weight resident complaint data relative to independent observational audits when prioritizing remediation resources.
From Impressionistic Description to Systematic Social Observation
Wilson and Kelling’s original essay relied on impressionistic, narrative description of physical disorder rather than any systematic measurement instrument, a methodological limitation that constrained early empirical testing of the theory to studies using resident survey reports of perceived disorder, which are themselves potentially contaminated by respondents’ independent knowledge of or exposure to actual crime, raising concerns about circularity in early disorder-crime associations (Harcourt, 2001). Sampson and Raudenbush’s (1999) development of systematic social observation, in which trained observers drove or walked through Chicago neighborhood blocks recording standardized ratings of physical conditions independent of any survey data, represented a methodological turning point, allowing researchers for the first time to measure physical disorder as an objective environmental characteristic rather than as a subjective resident perception.
This methodological advance proved consequential for the theory’s empirical status: Sampson and Raudenbush found that once systematically observed physical disorder was entered into models alongside neighborhood structural characteristics, particularly concentrated poverty and residential instability, disorder’s independent association with subsequent crime rates was substantially attenuated, suggesting that much of the disorder-crime correlation in earlier survey-based studies reflected shared structural causation rather than disorder’s independent causal contribution (Sampson & Raudenbush, 1999). This finding is examined at greater length in Criticisms of Broken Windows Theory and remains among the most frequently cited empirical challenges to the theory’s strongest causal claims.
The systematic social observation method itself has since been refined and extended by subsequent researchers seeking to reduce observer subjectivity further, including protocols employing photographic documentation reviewed by multiple independent coders and, more recently, computer-vision analysis of street-level imagery capable of classifying physical disorder indicators at city-wide scale without requiring in-person observer visits to every block (O’Brien, Sampson, & Winship, 2015). These technological extensions have expanded the geographic scope at which physical disorder can be measured but have not resolved the underlying causal-inference challenge that motivated their development, since even perfectly measured disorder remains correlated with the same unmeasured structural conditions that complicated earlier survey-based research.
Empirical Evidence on Physical Disorder and Crime
Cross-Sectional Associations
A substantial body of cross-sectional research, examining physical disorder and crime rates measured at the same point in time, has generally documented positive associations between physical disorder and both violent and property crime across a range of American cities (Skogan, 1990; Taylor, 2001). Skogan’s (1990) multi-city survey found that neighborhoods with higher resident-reported physical disorder also reported higher rates of criminal victimization, and this association held across the several cities included in his sample, lending initial cross-sectional support to the theory’s core prediction.
Cross-sectional designs, however, cannot by themselves establish the temporal sequencing the theory’s escalation hypothesis requires, since a positive correlation between physical disorder and crime observed at a single time point is equally consistent with disorder causing crime, crime causing disorder, for instance through offenders damaging property during the commission of other crimes, or both disorder and crime arising jointly from unmeasured structural causes such as poverty and vacancy rates (Harcourt, 2001; Taylor, 2001). This interpretive ambiguity has driven subsequent researchers toward longitudinal and quasi-experimental designs capable of better isolating physical disorder’s independent, temporally prior contribution to crime.
Taylor’s (2001) Baltimore panel study addressed this limitation directly by tracking the same neighborhood blocks over more than a decade, finding that early physical disorder predicted subsequent neighborhood decline, including population loss and rising crime, on some blocks but not others, with the difference substantially explained by baseline homeownership rates and residents’ informal territorial behavior, such as yard maintenance and porch use. Taylor’s findings suggested that physical disorder’s predictive power for later crime was neither uniform nor automatic but instead contingent on the same block-level social conditions that moderate the theory’s core signaling mechanism, discussed in Core Principles of Broken Windows Theory, reinforcing that a single national or even city-wide estimate of disorder’s crime effect likely obscures substantial underlying heterogeneity across block types.
Longitudinal and Quasi-Experimental Evidence
Longitudinal studies tracking physical disorder and crime over time within the same neighborhoods have produced more mixed results than the earlier cross-sectional literature. Sampson and Raudenbush’s (1999) Chicago-based longitudinal analysis, considered among the most methodologically rigorous tests of the theory, found that physical disorder’s association with subsequent violent crime became statistically weak once collective efficacy and structural neighborhood characteristics were controlled, suggesting that collective efficacy, rather than physical disorder itself, better explained variation in serious crime across Chicago neighborhoods.
A contrasting body of quasi-experimental evidence has emerged from studies evaluating targeted physical-disorder remediation interventions. Branas and colleagues’ (2011, 2018) studies of vacant lot greening and abandoned building remediation in Philadelphia, employing a difference-in-differences design comparing treated and untreated properties, found statistically significant reductions in nearby gun violence and other crime following remediation, providing some of the more methodologically credible evidence that physical disorder reduction can produce independent crime reductions, at least for specific interventions targeting the most visually salient forms of structural neglect (Branas et al., 2011; South et al., 2018). These findings are treated further in CPTED and Environmental Design Approaches to Disorder, which examines environmental remediation as a distinct crime-prevention strategy connected to but analytically separable from broken windows theory’s original policing focus.
Reconciling Divergent Findings
The apparent tension between Sampson and Raudenbush’s (1999) attenuated findings and Branas and colleagues’ (2011) positive remediation effects can be partly reconciled by attending to differences in research design and outcome specificity: Sampson and Raudenbush examined disorder’s association with aggregate neighborhood violent crime rates after controlling for collective efficacy, a test of the theory’s broadest causal claim, while Branas and colleagues examined the effect of a specific, randomly or quasi-randomly assigned physical intervention on crime occurring in close spatial proximity to the treated site, a narrower and more mechanistically direct test. This reconciliation suggests that physical disorder’s causal role may be real but geographically localized and intervention-specific, rather than operating as the broad, neighborhood-wide driver of crime trends the original 1982 essay implied.
This localized reading gains further support from Kondo and colleagues’ (2018) systematic review of neighborhood violence-reduction interventions, which found that the strongest and most consistently replicated evidence for physical-environment effects on crime came specifically from vacant land and abandoned structure remediation studies employing randomized or quasi-randomized assignment, while broader claims about disorder’s role in aggregate crime trends across entire cities or precincts remained considerably less well supported by comparably rigorous designs. The review’s authors explicitly recommended that policymakers treat physical remediation as a targeted, evidence-supported violence-reduction tool in its own right, distinct from and not requiring endorsement of the broader, more contested broken windows causal narrative from which such interventions originally drew their conceptual justification.
Policy and Practical Implications
Environmental Remediation as a Distinct Strategy
The accumulating evidence on targeted physical remediation, distinct from order-maintenance policing, has supported a policy approach sometimes termed situational or environmental crime prevention, in which municipal governments, land banks, and community organizations address vacant lots, abandoned structures, and street lighting directly rather than relying primarily on police enforcement against physical disorder (Kondo et al., 2018). This approach carries the practical advantage of avoiding the discretionary policing concerns associated with order-maintenance strategies targeting social disorder, discussed in Broken Windows Theory and Race and Stop-and-Frisk and Broken Windows Policing, since physical remediation does not require officer contact with residents and therefore does not raise the same discriminatory-enforcement concerns.
Measurement Considerations for Practitioners
Municipal agencies seeking to apply physical-disorder-reduction strategies face practical measurement challenges paralleling those researchers have confronted: identifying which forms of physical disorder carry the strongest signaling and safety effects, prioritizing limited remediation resources accordingly, and establishing baseline measurement sufficient to evaluate intervention effectiveness. The systematic social observation methodology developed by Sampson and Raudenbush (1999), adapted for practitioner use through standardized neighborhood audit instruments, has informed several municipal disorder-tracking programs seeking an evidentiary basis for remediation prioritization beyond resident complaint volume alone.
A further practical consideration concerns the durability of remediation effects over time. Branas and colleagues’ (2011) Philadelphia vacant lot study included follow-up observation confirming that greening interventions retained their crime-reduction association for at least several years following initial treatment, provided the treated lots received ongoing, low-cost maintenance such as periodic mowing, whereas comparison lots that received one-time cleanup without sustained maintenance showed disorder and associated crime measures reverting toward pre-intervention levels within a comparatively short period. This maintenance-dependency finding carries a significant practical implication for municipal budgeting: physical disorder remediation appears to function less as a one-time capital expenditure and more as an ongoing operational commitment, a distinction relevant to how cities structure land bank and vacant property programs seeking to sustain the crime-reduction benefits documented in the research reviewed above.
Disorder Perception, Race, and Neighborhood Context
Racial and Socioeconomic Variation in Disorder Perception
A significant methodological complication for physical disorder research emerged when Sampson and Raudenbush (2004) compared trained observers’ systematic ratings of physical disorder against residents’ own perceptions of disorder in the same Chicago blocks, finding that resident-perceived disorder was significantly influenced by the racial and socioeconomic composition of the neighborhood independent of the objectively observed physical conditions. Specifically, residents, including Black and white residents alike, tended to perceive predominantly Black and poor neighborhoods as more disordered than trained observers’ independent ratings of physical conditions warranted, a pattern Sampson and Raudenbush termed the social construction of disorder.
This finding carries substantial implications for the broader disorder-crime literature, since much of the earlier empirical evidence supporting the theory relied on resident survey reports of perceived disorder rather than independently observed physical conditions, raising the possibility that at least part of the documented disorder-crime association reflected racialized perception rather than objective environmental difference (Sampson & Raudenbush, 2004; Quillian & Pager, 2001). Quillian and Pager’s (2001) related research found that residents’ perceptions of neighborhood crime levels were disproportionately influenced by the percentage of young Black men in the neighborhood, independent of the neighborhood’s actual recorded crime rate, a parallel finding reinforcing concern that subjective disorder measures may partly encode racial bias rather than purely environmental observation.
Implications for Cross-Neighborhood Comparisons
These perceptual findings complicate cross-neighborhood comparisons of physical disorder research that rely on survey-based measures, since apparent disorder differences between predominantly white and predominantly Black or Latino neighborhoods may partly reflect measurement bias rather than genuine differences in physical conditions. Researchers have responded by increasingly favoring systematic, trained-observer, or administrative measures of physical disorder over resident-reported measures specifically to reduce this confound (Sampson & Raudenbush, 1999, 2004), a methodological shift with direct bearing on how future physical disorder research should be designed and interpreted, and one with clear relevance to the discriminatory-enforcement concerns examined at length in Broken Windows Theory and Race.
Physical Disorder in Comparative and Historical Perspective
Disorder Trends Across the Crime Decline of the 1990s
The steep decline in American violent and property crime rates during the 1990s occurred alongside declining physical disorder in several major cities that had adopted order-maintenance and remediation strategies, a temporal coincidence Kelling and Sousa (2001) cited as supportive evidence for the theory’s practical value. Economists and criminologists examining the broader national crime decline, however, have identified multiple concurrent explanatory factors independent of disorder reduction, including the waning of the 1980s crack cocaine epidemic, demographic shifts in the population at highest risk of offending, and increased incarceration rates, complicating any straightforward attribution of the crime decline to physical disorder reduction specifically (Levitt, 2004). Because the crime decline was national and occurred even in cities that had not adopted disorder-focused policies, single-city before-and-after comparisons of the kind Kelling and Sousa (2001) presented face significant threats to internal validity absent an appropriate comparison group.
Contemporary Digital and Administrative Measurement Tools
More recent physical disorder research has incorporated administrative data sources unavailable to earlier researchers, particularly municipal 311 non-emergency service request systems, which generate large volumes of geocoded, time-stamped reports of physical disorder conditions such as graffiti, illegal dumping, and abandoned vehicles (O’Brien, Sampson, & Winship, 2015). These administrative data sources offer researchers substantially larger sample sizes and finer geographic and temporal resolution than earlier survey or observer-based methods permitted, though O’Brien and colleagues (2015) caution that 311 reporting rates themselves vary systematically with neighborhood civic engagement and demographic composition, reintroducing a version of the same measurement-bias concern identified in resident-perception research and requiring careful methodological correction before administrative disorder data can be treated as an objective proxy for underlying physical conditions.
Analytical Tables
Table 1. Major Empirical Studies of Physical Disorder and Crime
| Study | Design | Location | Physical Disorder Measure | Key Finding |
|---|---|---|---|---|
| Skogan (1990) | Cross-sectional, multi-city survey | 40 U.S. neighborhoods | Resident-reported disorder | Positive association with victimization |
| Sampson & Raudenbush (1999) | Longitudinal, systematic observation | Chicago | Trained-observer ratings | Association attenuated after controlling for collective efficacy |
| Harcourt & Ludwig (2006) | Quasi-experimental reanalysis | New York City, five other cities | Public housing disorder reduction | No independent crime-reduction effect found |
| Branas et al. (2011) | Difference-in-differences | Philadelphia | Vacant lot greening | Significant reduction in nearby gun assaults |
| South et al. (2018) | Randomized controlled trial | Philadelphia | Abandoned building remediation | Reductions in nearby crime, particularly gun violence |
Table 2. Categories of Physical Disorder, Responsible Party, and Typical Remediation Approach
| Category | Example Conditions | Typically Responsible Party | Common Remediation Approach | Discretionary Policing Risk |
|---|---|---|---|---|
| Structural neglect | Abandoned buildings, broken windows | Private property owner or land bank | Code enforcement, demolition, rehabilitation | Low |
| Vacant land | Overgrown or debris-filled lots | Municipal government or absentee owner | Greening, lot cleanup programs | Low |
| Vandalism | Graffiti, broken fixtures | Private or public property owner | Rapid removal or repair programs | Low |
| Infrastructural deficiency | Broken streetlights, potholes | Municipal public works | Scheduled maintenance, targeted repair | Low |
| Litter accumulation | Trash, discarded furniture | Municipal sanitation or property owner | Sanitation enforcement, cleanup crews | Low to moderate |
Conclusion
The empirical literature on physical disorder and crime has moved considerably beyond Wilson and Kelling’s original impressionistic account, developing systematic measurement tools and increasingly rigorous research designs capable of distinguishing physical disorder’s independent causal contribution from its status as a co-symptom of shared structural disadvantage. That literature offers a more qualified conclusion than the original 1982 essay’s confident causal narrative: broad neighborhood-level associations between physical disorder and aggregate crime rates appear substantially attenuated once collective efficacy and structural conditions are held constant, while targeted, spatially localized remediation interventions, particularly vacant lot greening and abandoned building rehabilitation, have produced more methodologically credible evidence of independent crime-reduction effects.
This pattern suggests that physical disorder’s relationship to crime may be less a matter of a generalized neighborhood-wide signaling mechanism, as the original theory proposed, and more a matter of specific, spatially concentrated environmental conditions that create identifiable opportunities or reduce identifiable guardianship at particular locations. This more localized, opportunity-focused reading of the evidence connects physical disorder research to the broader environmental criminology tradition addressed in CPTED and Environmental Design Approaches to Disorder, and represents one of several respects in which contemporary scholarship has substantially revised the theory’s original, more sweeping causal claims.
Related Articles
- Core Principles of Broken Windows Theory
- Social Disorder and Crime
- CPTED and Environmental Design Approaches to Disorder
- Criticisms of Broken Windows Theory
- What the Research Says About Broken Windows
References
- Branas, C. C., Cheney, R. A., MacDonald, J. M., Tam, V. W., Jackson, T. D., & Ten Have, T. R. (2011). A difference-in-differences analysis of health, safety, and greening vacant urban space. American Journal of Epidemiology, 174(11), 1296–1306.
- Harcourt, B. E. (2001). Illusion of Order: The False Promise of Broken Windows Policing. Harvard University Press.
- Harcourt, B. E., & Ludwig, J. (2006). Broken windows: New evidence from New York City and a five-city social experiment. University of Chicago Law Review, 73(1), 271–320.
- 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.
- Kondo, M. C., Andreyeva, E., South, E. C., MacDonald, J. M., & Branas, C. C. (2018). Neighborhood interventions to reduce violence. Annual Review of Public Health, 39, 253–271.
- 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.
- O’Brien, D. T., Sampson, R. J., & Winship, C. (2015). Ecometrics in the age of big data: Measuring and assessing “broken windows” using large-scale administrative records. Sociological Methodology, 45(1), 101–147.
- Perkins, D. D., Meeks, J. W., & Taylor, R. B. (1992). The physical environment of street blocks and resident perceptions of crime and disorder. Journal of Environmental Psychology, 12(1), 21–34.
- Quillian, L., & Pager, D. (2001). Black neighbors, higher crime? The role of racial stereotypes in evaluations of neighborhood crime. American Journal of Sociology, 107(3), 717–767.
- Ross, C. E., & Mirowsky, J. (1999). Disorder and decay: The concept and measurement of perceived neighborhood disorder. Urban Affairs Review, 34(3), 412–432.
- Sampson, R. J., & Raudenbush, S. W. (1999). Systematic social observation of public spaces: A new look at disorder in urban neighborhoods. American Journal of Sociology, 105(3), 603–651.
- Sampson, R. J., & Raudenbush, S. W. (2004). Seeing disorder: Neighborhood stigma and the social construction of “broken windows.” Social Psychology Quarterly, 67(4), 319–342.
- Skogan, W. G. (1990). Disorder and Decline: Crime and the Spiral of Decay in American Neighborhoods. Free Press.
- South, E. C., Hohl, B. C., Kondo, M. C., MacDonald, J. M., & Branas, C. C. (2018). Effect of greening vacant land on mental health of community-dwelling adults: A cluster randomized trial. JAMA Network Open, 1(3), e180298.
- Taylor, R. B. (2001). Breaking Away from Broken Windows: Baltimore Neighborhoods and the Nationwide Fight Against Crime, Grime, Fear, and Decline. Westview Press.
- Wilson, J. Q., & Kelling, G. L. (1982). Broken windows: The police and neighborhood safety. The Atlantic Monthly, 249(3), 29–38.