Broken Windows Thesis Research Paper Example

This sample Broken Windows Thesis Research Paper is published for educational and informational purposes only. If you need help writing your assignment, please use our research paper writing service and buy a paper on any topic at affordable price. Also check our tips on how to write a research paper, see the lists of criminal justice research paper topics, and browse research paper examples.

Few ideas in criminology have had the type of direct impact on criminal justice policy exhibited by the broken windows thesis. From its inauspicious beginnings in a nine-page article by James Q. Wilson and George Kelling in The Atlantic Monthly, the broken windows thesis has impacted policing strategies around the world. From the “quality of life policing” efforts in New York City (Kelling and Sousa 2001) to “Zero Tolerance” policing in England (Dennis and Mallon 1998), police agencies around the world have embraced Wilson and Kelling’s idea that focusing on less serious offenses can yield important benefits in terms of community safety and prevention of more serious crime. Despite this broad policy influence, research on the theory itself has been relatively weak and has produced equivocal findings as will be detailed in this research paper.

The Broken Windows Model

In a nutshell, the broken windows thesis (Wilson and Kelling 1982) suggests that police could more effectively fight crime by focusing on more minor annoyances which plague communities – hereafter referred to as disorder (some works also label these issues as “incivilities”). Disorder includes both rundown physical conditions in the form of litter, dilapidated buildings, graffiti, etc. as well as social nuisances such as panhandling, loitering, and public drinking. Their idea that crime could be prevented through targeting these issues was based on their thesis that such social ills eventually lead to community decline if left untended.

This process essentially has four key steps and begins with disorder not being dealt with in a timely manner. Trash is not picked up; loiterers are not asked to move on; people are drinking in public without being warned away. In time this invites more trash being thrown in the street, more loiterers to gather, and more people to start drinking in public. As this disorder accumulates, it sends a message to residents that things are getting out of control and that social controls have failed in their neighborhood. The key here is that residents perceive this untended disorder accumulating. It will likely have little impact if residents are not aware of the disorder in the community. In turn, the next step of the process is that residents who perceive worsening disorder problems eventually become fearful and begin to withdraw from the community. They spend less time outside, become less likely to intervene and ward off disorderly people, and, in the extreme, “good” residents may move away.

This leads to the third step of the process, which is a lowering of informal social controls. The community withdrawal results in fewer watchful eyes on the street, and the area is now less able to regulate behavior through informal social controls. As such disorder and minor crimes continue to flourish in these areas. Finally, this brings about the fourth step, in which criminals take these signs of untended disorder as a cue that such a neighborhood is a good place for them to work with relative impunity as they perceive their chances of being arrested in such areas to be slim. In Wilson and Kelling’s terms, such neighborhoods are vulnerable to criminal invasion. It is not inevitable, but such places are much more likely, in their view, to see an increase in crime than neighborhoods which exert control in regulating the occurrence of disorder. Once crime occurs, residents also notice this and the cycle of fear and withdrawal is likely to worsen as the process is essentially a feedback loop. The steps of the community decline cycle outlined by broken windows thesis can be visualized as shown in Fig. 1.

As such, a main thrust of Wilson and Kelling’s argument was that police could fight crime more effectively by dealing with disorder. If they stop disorder from accumulating and prevent neighborhoods from reaching the tipping point where they become vulnerable for criminal invasion, they can have a large impact on crime. Wilson and Kelling do not discuss what police may do in neighborhoods already past the tipping point and fully invaded by criminal behavior, but one could imply that cleaning up disorder would still play a role in restoring informal social control in such neighborhoods and helping residents take back the streets. As with any theory about crime, the broken windows thesis can be more readily understood by examining the research that influenced its formation.

Theoretical Development Of The Broken Windows Thesis

While Wilson and Kelling are credited with developing the broken windows thesis, they were not the first to examine the role disorder played in communities. In the area of criminology, concern over disorder can partly be traced to early research on fear of crime. One issue that drove interest in the topic was a body of research that consistently found that fear of crime had seemingly little to do with crime. For instance, most studies found that females and the elderly reported the highest levels of fear of crime, yet the National Crime Surveys conducted by the Bureau of Justice Statistics consistently showed young males to have the highest rates of victimization. Thus, fear of crime did not appear to be driven by actual victimization risk, and these findings naturally led criminologists to question what was driving fear of crime if not crime itself.

A number of studies turned to disorder to explain fear of crime. For example, James Q. Wilson first noted in 1975 that people were troubled not only by crime but also by:

—–The daily hassles they are confronted with on the street – street people, panhandlers, rowdy youths, or ‘hey honey’ hassles – and the deteriorated conditions that surround them – trash strewn alleys and vacant lots, graffiti, and deteriorated or abandoned housing – inspire concern. (p. 66)

Similarly, Garofalo and Laub (1978) stated that “.. .what has been measured in research as the ‘fear of crime’ is not simply fear of crime” (p. 245) and tied fear to quality of life and concern for the community. Ideas closely related to the broken windows thesis are most clearly seen in work by Hunter (1978). Hunter suggested that disorder affected both fear of crime and actual crime through a process in which disorder signaled to residents that local controls had failed and caused them to become personally at risk of victimization. He suggested that this would increase crime and further increase fear. His work can easily be seen as an early version of the broken windows thesis.

Finally, Wilson and Kelling’s ideas were greatly influenced by a social-psychological experiment conducted by Stanford psychologist Philip Zimbardo in 1969, as indicated by the detailed discussion of the experiment in their broken windows article. Zimbardo abandoned a car with its hood up in two places – the Bronx in New York City and on the Stanford Campus in Palo Alto, California. The car in the Bronx was vandalized within 10 min, and within 24 h everything of value was removed. The car in Palo Alto, however, was not touched for more than a week. Zimbardo then smashed the windshield with a sledgehammer, and from that point on, people passing by saw the activity and the damaged car and joined in the destruction. This is where the broken windows metaphor came from for Wilson and Kelling and, combined with the above work on fear of crime and disorder, formed the basis for their ideas that untended disorder is what eventually leads to a neighborhood becoming crime plagued. Just like the broken window on the car in Palo Alto invited more vandalism, untended disorder is a visual cue in a community which invites more disorder and eventually more serious crime.

While the theoretical underpinnings of the broken windows thesis can clearly be seen in this early work on fear of crime and the Zimbardo experiment, the law enforcement portion of Wilson and Kelling’s ideas was directly influenced by earlier research they had conducted on policing. Most notably, this is seen in the work of Wilson and Boland (1978) who noted that aggressive policing can reduce crime. Their main point was that the police may reduce crime not by how many are on patrol but rather by what they do while on patrol. They suggested that if police are aggressive in arresting criminals, they can have more of an impact on crime. Their study used traffic citations for as a proxy for aggressive policing and found a negative relationship between police aggressiveness and crime rates.

George Kelling was also greatly influenced by his earlier work on policing. In particular, the broken windows article repeatedly makes reference to his experience working on an evaluation of foot patrol in Newark, New Jersey (Kelling et al. 1981). In the Broken Windows article, Kelling elaborates on how his experiences on this study showed him that by being active in the community, the police could maintain order and make residents feel better, which could have positive impacts even if the police strategy was not directly reducing crime.

The above discussion lays out the theoretical foundation for Wilson and Kelling’s broken windows thesis. It is clear how work both on causes of fear of crime and studies of specific police practices laid the groundwork for their ideas that police could fight crime by tackling smaller problems – the disorder in a community that made residents fearful and uneasy. In the quarter century since its inception, the broken windows thesis has had a profound impact on policing and continues to be a subject of theoretical debate in scholarly journals. The sections below provide an overview of evidence on the validity of the broken windows thesis itself.

Testing The Broken Windows Thesis

For a theory that has had such a large impact on police practice, the broken windows thesis has received relatively little research attention in terms of testing the theoretical propositions the broken windows thesis itself. As this section will show, there are a number of studies which have examined individual propositions of the broken windows thesis (often with no direct connection to or mention of broken windows), but few studies have set out to explicitly and directly test the entire broken windows thesis.

Looking at the broken windows thesis outlined above, one can see that there are three main theoretical propositions that make up the path from disorder to more serious crime. First, disorder must increase fear of crime among residents. In particular, perceptions of disorder are most relevant here. If residents are unaware of disorder on their block, it is unlikely to impact their levels of fear. Secondly, residents who become fearful must withdraw from the community, thus lowering informal social controls/ collective efficacy. Third, crime must increase in the area in response to this withdrawal as criminals perceive the area to be a place to commit crimes with relative impunity.

Does Disorder Cause Fear Of Crime?

Wilson and Kelling (1982) suggested that disorder’s first effect was increasing resident fear of crime. A number of studies have explored this issue by testing the connection between disorder and fear of crime, and a selection of these studies are reviewed here. A neighborhood-level study by Covington and Taylor (1991) found that both objectively measured disorder, as well as survey measured perceptions of disorder, were related to fear of crime. Moreover, their perceived disorder measure was the dominant effect in the model. Another study involving a panel interview of residents in Baltimore found that between the two surveys, those whose perceptions of disorder increased reported less satisfaction with the block they lived on, as well as showing greater increases in feelings of vulnerability and fear of crime (Robinson et at. 2003). Also of note, some research has suggested that social disorder has a stronger relationship with fear of crime than physical disorder (LaGrange et al. 1992).

More recent research has also supported the link between disorder and fear. One study tested a variation of the broken windows/decline hypothesis similar to the model in Fig. 1 using a sample of neighborhoods from three waves of the British Crime Survey (Markowitz at al. 2001). Their findings suggested the model was correct. Of particular interest to the current research, they found that “the dominant effect in the cycle is the effect of disorder on fear” (p. 310). The authors thus concluded that their findings were consistent with the broken windows/decline hypothesis as they show that disorder may increase crime indirectly by increasing levels of fear which in turn reduce the level of social cohesion which may then lead to crime.

On the other hand, some research has suggested that the relationship between disorder and fear may not be straightforward. Taylor and Shumaker (1990) found a relationship between disorder and fear of crime, but noted that it may not be linear. Their findings suggested a quadratic relationship where disorder does increase fear, but as disorder gets higher, the strength of the relationship weakens. They note that perhaps people in high disorder areas get somewhat inoculated to disorder, just as people living in areas plagued by natural disasters may fear earthquakes or hurricanes less than people from other areas. In short, some people may be accustomed to disorder and not bothered by it as much. Additionally, a few studies have not found clear evidence of a link between disorder and fear of crime. For instance, Taylor’s (2001) well-known study in Baltimore found that incivilities measured in 1981 and 1982 were not related to changes in fear of crime in these neighborhoods between 1982 and 1994.

Overall, the studies reviewed above illustrate that the link between disorder and fear of crime has generally been supported in past research, with a few exceptions. Thus, there is reason to believe that Wilson and Kelling (1982) and those before them (e.g., Garafalo and Laub 1978; Hunter 1978; Wilson 1975) were correct in asserting that signs of disorder make people feel uneasy and even fearful of being victimized by crime in their neighborhood. Moving on to the next proposition, what does the research say about the link between fear of crime and withdrawal from the community?

Does Fear Of Crime Lead To Withdrawal?

As with the link between disorder and fear of crime, there has been work examining whether fear of crime leads to community withdrawal. Again, most of this work was done without explicit reference to broken windows, but there are studies of the topic in other areas such as tests of variables related to social control or collective efficacy (Sampson et al. 1997). A review of these studies is relevant to understanding the evidence on the broken windows thesis as informal social control/collective efficacy can be viewed as analogous to Wilson and Kelling’s (1982) discussion of the ability (or inability) of a community to regulate behavior.

A study by Garafalo (1981) found that fear of crime was related to a host of social outcomes related to collective efficacy such as heightened interpersonal distrust, withdrawal of support for formal control agencies and decreased social interaction. The Markowitz et al. (2001) study cited above which found a link between disorder and fear also found that fear in turn reduced social cohesion. Thus, as with the link between disorder and fear, a review of the research on the link between fear and withdrawal/collective efficacy (and related constructs) is generally supported in past research. The majority of the studies reviewed find that increases in fear of crime are tied to reductions in things such as neighborhood cohesion, social involvement, and interpersonal trust. Moving to the final theoretical proposition, what does the research say about the impacts of withdrawal and reductions in social control/collective efficacy on crime?

Does Withdrawal/Weakening Of Social Controls Lead To Crime?

This link of the broken windows thesis really gets at the notion that reductions in informal social lead to increases in crime. This topic has received a fair amount of research attention in work examining the impact of social control/ collective efficacy on crime, as well as work on social disorganization theory. This brief review will focus on research on collective efficacy, as that is the specific theory of informal social control that is at the center of current debates over the broken windows thesis. Collective efficacy refers to the notion that crime can be controlled in a community where there is a high level of social cohesion/social ties and willingness to intervene for the common good (Sampson et al. 1997). Places with higher levels of collective efficacy are more able to effectively regulate behavior and thus have lower crime rates.

In general, the collective efficacy literature has been supportive of this inverse relationship between collective efficacy and crime rates. For instance, in the work first advancing the concept, Sampson et al. (1997) found that collective efficacy was negatively associated with violence after controlling for individual factors and prior violence. A study by Sabol et al. (2004) found that collective efficacy reduced youth violence, child maltreatment, and intimate partner violence. Ford and Beveridge (2004), in looking at illegal drug sales, noted that collective efficacy can prevent drug sales not only by resisting undesirable factors (like drug dealers) but also through increasing the capacity to attract positive factors (i.e., legal businesses).

However, a study by Morenoff et al. (2001) found that collective action for social control was negatively related to homicide, but found little support that social ties mattered. Thus, it appears to be important that communities have shared goals and that residents are willing to intervene for the common good, but that social cohesion/social ties may not necessarily be a key component of collective efficacy. What matters is whether a community can work together to solve crime and disorder problems. Residents may not have to have strong social ties/cohesion to share mutual goals for the community and be willing to intervene to achieve these goals. These findings fit with the broken windows thesis as Wilson and Kelling’s (1982) work focused on the notion that disorder led people to become fearful and less willing to intervene in the community. Social cohesion was not explicitly focused on in their work, as they focused more on the notion of people being too afraid to intervene and/or even moving away in response to rising untended disorder in their neighborhood.

In short, work looking at the relationship between collective efficacy and crime has generally been supportive and builds on the social disorganization literature which has a long history of showing that a community’s ability to exert informal social control is a powerful predictor of crime rates. As such, the final link of broken windows has been supported with a long history of criminological work which suggests that withdrawal and other factors that reduce informal social control, collective efficacy, etc. are likely to lead to increases in crime.

The Current Empirical Status Of The Broken Windows Thesis

The above review shows is that there is a fair amount of support for the individual theoretical propositions behind the broken windows thesis. However, this support comes largely from a body of unrelated studies, few of which examine more than one step of the broken windows model, and even for the individual propositions, there are still mixed findings at times. In turn, there is still a great deal of debate about the validity of the broken windows thesis. While some authors have strongly advocated the model and/or broken windows policing (e.g., Bratton and Kelling 2006; Kelling and Coles 1996; Kelling and Sousa 2001), some recent research has challenged the broken windows thesis by suggesting that the relationship between disorder and crime is spurious and explained by collective efficacy (see Sampson and Raudenbush 1999).

A Direct Or Indirect Relationship Between Disorder And Crime?

A study by Sampson and Raudenbush (1999) sparked considerable debate over the broken windows thesis. Sampson and Raudenbush set out to test the broken windows thesis and collective efficacy theory using data collected as part of the Project on Human Development in Chicago Neighborhoods (PHDCN). These data included systematic social observations of physical and social disorder which were collected by videotaping segments with video cameras mounted to a vehicle which drove through study areas at 5 mph recording both sides of the streets. The tapes were then coded by members of the PHDCN research team to record levels of social and physical disorder on the study blocks. Collective efficacy and other individual-level variables were obtained through a resident survey, while crime was measured through official police data (and survey measured victimization in some models).

Sampson and Raudenbush used weighted regression analysis and measured variable path analysis in their study and found that disorder was positively related to crime. However, when they added collective efficacy to the model, they found that the relationship between disorder and crime disappeared – except for robbery where it remained significant. As such they concluded that the broken windows thesis was not supported as disorder and crime were only spuriously related – their results showed both to be a result of low collective efficacy. They argued that disorder and crime were simply different degrees of the same problem with the same underlying cause, rather than being causally related as the broken windows thesis suggests. As such, they concluded “.. ..that neighborhoods high in disorder do not have higher crime rates in general than neighborhoods low on disorder once collective efficacy and structural antecedents are held constant” (p. 638).

Other studies have also examined this notion of a direct link between disorder and crime and produced findings that challenged the broken windows thesis. A study by Taylor (2001) found that disorder measured in 1981 was not strongly related to crime in 1994 after controlling for initial neighborhood structure. A more recent study found that neither collective efficacy nor disorder was a sufficient explanation for crime (St. Jean 2007). While high collective efficacy and low social disorder (physical disorder did not matter) explained low crime rates, places with low levels of collective efficacy and/or high levels of social disorder were found to be about equally likely to have high or low crime rates. Finally, in another study using 16 years of census block-level data collected in Seattle, Yang (2007, 2010) found that the trends between violent crime and disorder were correlated. The direction of causation, however, was opposed to what was suggested by the broken windows thesis. The results from Granger causality tests generally showed no causal relationship between disorder and violence, and in a few places, the causality appeared to run from violent crime to disorder.

On the other hand, a recent social-psychology experiment conducted in the Netherlands did find some support for a direct link between the presence of disorder and minor criminal behavior (Keizer et al. 2008). The study manipulated the levels of disorder in places (graffiti, trash, etc.) and tested whether it had an impact on behavior among passersby. Most of the tests looked at the spread of disorder – for instance, one of the experiments which was part of this study found that the presence of graffiti in a researcher-manipulated area led to a higher likelihood of people littering (throwing a flyer attached to their bicycles on the ground) compared with another area where the researches did not place graffiti (but also attached flyers to bicycles). However, one part other experiment examined minor theft. This experiment involved placing an envelope with a small amount of money clearly visible sticking of a mailbox. In the control setting, there was no graffiti or trash around the mailbox, and in the two experimental conditions, there was graffiti on the mailbox in one setting and trash strewn around it in the other. In both cases, passersby in the conditions with disorder present were significantly more likely to steal the envelope than those in the control condition. Of course, it is impossible to say from this study if the presence of disorder could lead to more serious crimes as the broken windows thesis suggests, or if it is just another example of how disorder spreads (which was the main focus of the study) given the very minor nature of the crime in question.

While the methodologies of these studies are sound, and the negative results of all but the Netherlands study seem to be a large challenge to the tenets of the broken windows thesis, this body of work has been criticized as an unfair test of broken windows. The challenge to this work comes from its assertion that crime and disorder are directly related – that disorder directly leads to crime. Some scholars, including Kelling himself, have denied that broken windows ever implied a direct relationship between disorder and crime (Bratton and Kelling 2006; Gault and Silver 2008; Xu et al. 2005), arguing that it has always posited an indirect relationship between disorder and crime.

A reading of Wilson and Kelling’s (1982) article would seem to support Bratton and Kelling’s (2006) assertion that they have always posited an indirect relationship between crime and disorder. For instance, Wilson and Kelling stated “.. .at the community level, disorder and crime are usually inextricably linked, in a kind of developmental sequence” (p. 31). A developmental sequence does not imply a direct relationship. In fact, going back as far as Zimbardo’s 1969 experiment, the broken windows thesis has been a social-psychological theory. It has never suggested that disorder in a community directly causes crime. Rather it has always posited that perceptions of disorder, created through visual cues of untended disorder in a community, increase fear and lead to residential withdrawal which leaves communities vulnerable to criminal invasion. Studies that look for a direct relationship between observed disorder and crime ignore the socialpsychological foundation of the broken windows thesis and thus are not complete tests of the theory.

Given this flaw, the impact of Sampson and Raudenbush’s study, and other work testing for a direct relationship between disorder and crime, is challenged. If Sampson and Raudenbush had specified their model in accordance with the propositions and social-psychological foundation of the broken windows thesis, their data may have actually supported the broken windows thesis. A correct specification would test whether perceptions of disorder were positively related to levels of fear and then test whether fear in turn was negatively related to collective efficacy (the informal social control portion of the broken windows thesis) and whether collective efficacy was related to crime. Given that they found collective efficacy inversely related to crime, a finding that disorder, through increased fear of crime, reduced collective efficacy would be supportive of the broken windows’ notion that disorder erodes informal social controls and leads to increases in crime.

Sampson and Raudenbush (1999) themselves even made some statements that are supportive of a true reading of broken windows. For instance, in their conclusion, they state that “[e]radicating disorder may indirectly reduce crime by stabilizing neighborhoods.. ..” (p. 648). This is likely based on an analysis presented in passing on page 636 in which they found that “[t]he results indicated that observed disorder increases perceived disorder, which in turn reduces collective efficacy. The significant reciprocal relationship between violence and collective efficacy nonetheless remained intact.. …” Thus, if Sampson and Raudenbush had set up their study to test the indirect link between perceived disorder and crime as outlined above, their conclusions likely would have been very different, as others have noted (Gault and Silver 2008; Xu et al. 2005).

Some recent research has aimed to address this issue by testing models that are more faithful to Wilson and Kelling’s conceptualization of the broken windows thesis. Xu et al. (2005), in their study of community policing, found that perceived disorder had strong direct and indirect impacts on perceived crime after controlling for collective efficacy (Xu et al. 2005) and thus challenge the assertion made by Sampson and Raudenbush. However, Xu et al. also did not model the relations in the specific order suggested by the broken windows thesis and thus limited the ability of their study to test the theoretical propositions behind the broken windows thesis. Specifically, collective efficacy is only included as an exogenous variable in their structural equation models. Collective efficacy (along with community policing variables) is said to affect disorder and crime, which in turn affect fear of crime and perceptions of quality of life which then affect satisfaction with the police. As outlined above, a true test of the broken windows thesis as outlined by Wilson and Kelling (1982) would test the impact of disorder on fear of crime, which should in turn affect collective efficacy and crime.

A recent study aimed to shed light on this issue by using structural equation modeling to test the relationships between perceived disorder, fear of crime, collective efficacy, and perceived crime specified by the broken windows thesis (see Fig. 1 above) using structural equation modeling (Hinkle 2009). This study used perceptual data measured through resident surveys and found support for both direct and indirect pathways between perceived disorder and perceived crime in the best fitting model (see Fig. 2 below). While caution is needed in interpreting these findings due to the study using perceptual measures of disorder and crime from the same survey, the results nonetheless suggest that theories dealing with the relationship between disorder and crime may be overly simplistic if they only consider either direct or indirect pathways between these two variables. Additionally, the final model from this study challenged the centrality of fear of crime to the broken windows thesis. Once disorder was allowed to have direct impacts on collective efficacy and crime, the relationship between fear and collective efficacy hypothesized by the broken windows thesis was no longer statistically significant. The implications of these findings are elaborated on in the concluding section below.

The Future Of The Broken Window Thesis

This review of the broken windows thesis essentially points out two key issues about the topic. First, it is an idea that has had a very large impact on policing policy and practice. Second, despite this impact, the thesis itself has not received much direct empirical research attention. While there is a good deal of evidence on the individual relationships between disorder, fear of crime, collective efficacy/informal social control, and crime, there are hardly any studies which examine all of these variables simultaneously in one model with the relationships tested in a manner consistent with Wilson and Kelling’s (1982) hypotheses.

As such, we still know little about the validity of the idea behind the policing model, and thus, debates over the thesis have been a hot topic in academic circles in recent years as outlined above. Given this dearth of empirical research explicitly testing the broken windows thesis, the future of the idea is very much up in the air. What is clear is that we need a substantial amount of empirical research testing the broken windows thesis. The study by Hinkle (2009) outlined above is one example and can serve as a starting point for a discussion of where future research on the broken windows thesis should begin.

A first takeaway point from that study is that future research should ideally consider both direct and indirect pathways from disorder to crime when testing the broken windows thesis. Future studies would be aided by including hard measures of crime in addition to perceptual measures. A second takeaway point was that fear of crime did not have the expected impact on collective efficacy once disorder was allowed to have direct impacts on efficacy and crime.

While this is only one study, and this finding should thus be interpreted with caution, it does challenge the centrality of fear in the broken windows framework and suggests the need for theoretical elaboration on Wilson and Kelling’s model.

For instance, while the results from this study challenge the role of fear of crime, it may also be unlikely that disorder and collective efficacy are truly directly related as the model in Fig. 2 shows. It may be more likely that there are other intervening variable besides fear of crime which mediate the impact of disorder on levels of informal social control. Maybe instead of untended disorder increasing fear and scaring off residents who care, it is more a matter of residents getting annoyed or frustrated and thus they give up intervening or even move away. Or perhaps over time some residents simply become used to the disorder and are not bothered by it as Taylor and Shumaker (1990) argued – in which case residents may become apathetic over time and thus stop intervening for the common good. Future theoretical work should consider these types of issues and attempt to elaborate on the broken windows thesis. It is up to such work to put forth more elaborate models based on theoretical propositions which clearly outline the relationships between disorder, crime, and whatever mediating variables are hypothesized to lie between these two constructs. These models can then be tested with SEM and other techniques to gauge whether they are supported empirically.

Finally, a key in sorting out these debates over the validity of the broken windows thesis will be testing the idea with longitudinal data. Most of the studies discussed above have tested the model with cross-sectional data. While such studies offer valuable insight into the broken windows model, it is nonetheless a theory of neighborhood decline that occurs over time. As such, it is best tested using data that covers a number of years. There have been some longitudinal tests of the theory by researchers such as Taylor (2001) and Yang (2007, 2010), and these tests tended to challenge the relationship between disorder and crime. However, these studies only tested for direct relationships between disorder and crime and thus were not full tests of the broken windows thesis.

Bibliography:

Bratton WJ, Kelling GL (2006) There are no cracks in the broken windows. Natl Rev Online, February 28. Retrieved from http://www.nationalreview.com/comment/ bratton_kelling200602281015.asp
Covington J, Taylor RB (1991) Fear of crime in urban residential neighborhoods: implication of betweenand within-neighborhood sources for current models. Sociol Q 32(2):231–249
Dennis N, Mallon R (1998) Confident policing in Hartlepool. In: Dennis N (ed) Zero tolerance: policing a free society. Institute of Economic Affairs Health and Welfare Unit, London, pp 62–87
Ford JM, Beveridge AA (2004) “Bad” neighborhoods, fast food, “sleazy” businesses, and drug dealers: relationships between the location of licit and illicit businesses in the urban environment. J Drug Issues 34(1):51–76
Garafalo J (1981) The fear of crime: causes and consequences. J Crim Law Criminol 72(2):839–857
Garofalo J, Laub JH (1978) The fear of crime: broadening our perspective. Victimology 3(3–4):242–253
Gault M, Silver E (2008) Spuriousness or mediation? Broken windows according to Sampson and
Raudenbush (1999). J Crim Just 36(3):240–243
Hinkle JC (2009) Making sense of broken windows: the relationship between perceptions of disorder, fear of crime, collective efficacy and perceptions of crime. Unpublished doctoral dissertation, University of Maryland, College Park
Hunter A (1978) Symbols of incivility: Social disorder and fear of crime in urban neighborhoods. Paper presented at the annual meeting of the American Society of Criminology, Dallas, TX
Keizer K, Lindenberg S, Steg L (2008) The spreading of disorder. Science 12:1681–1685
Kelling GL, Coles C (1996) Fixing broken windows: restoring order and reducing crime in American cities. Free Press, New York
Kelling GL, Sousa WH (2001) Do police matter? An analysis of the impact of New York City’s police reforms (Civic report No., 22). Manhattan Institute for Policy Research, New York
Kelling GL, Pate T, Ferrara A, Utne M, Brown CE (1981) The newark foot patrol experiment. The Police Foundation, Washington, DC
LaGrange RL, Ferraro KP, Supancic M (1992) Perceived risk and fear of crime: role of social and physical incivilities. J Res Crime Delinq 29(3):311–334
Markowitz FE, Bellair PE, Liska AE, Liu J (2001) Extending social disorganization theory: modeling the relationships between cohesion, disorder and fear. Criminology 39(2):293–320
Morenoff JD, Sampson RJ, Raudenbush SW (2001) Neighborhood inequality, collective efficacy, and the spatial dynamics of urban violence. Criminology 39(3):517–560
Robinson JB, Lawton BA, Taylor RB, Perkins DD (2003) Multilevel longitudinal impacts of incivilities: fear of crime, expected safety, and block satisfaction. J Quant Criminol 19(3):237–274
Sabol WJ, Coulton CJ, Korbin JE (2004) Building community capacity for violence prevention. J Interpers Violence 19(3):322–340
Sampson RJ, Raudenbush SW (1999) Systematic social observation of public spaces: a new look at disorder in urban neighborhoods. Am J Sociol 105(3):603–651
Sampson RJ, Raudenbush SW, Earls F (1997) Neighborhoods and violent crime: a multilevel study of collective efficacy. Science 277(5328):918–924
St. Jean PKB (2007) Pockets of crime: broken windows, collective efficacy, and the criminal point of view. The University of Chicago Press, Chicago
Taylor RB (2001) Breaking away from broken windows: Baltimore neighborhoods and the nationwide fight against crime, grime, fear, and decline. Westview Press, Boulder
Taylor RB, Shumaker SA (1990) Local crime as a natural hazard: implications for understanding the relationship between disorder and fear of crime. Am J Community Psychol 18(5):619–641
Wilson JQ (1975) Thinking about crime. Basic Books, New York
Wilson JQ, Boland B (1978) The effect of the police on crime. Law Soc Rev 12:367–390
Wilson JQ, Kelling GL (1982) Broken windows: the police and neighborhood safety. Atl Mon 211:29–38
Xu Y, Fiedler ML, Flaming KH (2005) Discovering the impact of community policing: the broken windows thesis, collective efficacy and citizens’ judgment. J Res Crime Delinq 42(2):147–186
Yang S (2007) Causal or merely co-existing: a longitudinal study of disorder and violence at places. Unpublished doctoral dissertation, University of Maryland, College Park
Yang S (2010) Assessing the spatial-temporal relationship between disorder and violence. J Quant Criminol 26(1):139–163
Zimbardo PG (1969) The human choice: Individuation, reason, and order versus deindividuation, impulse, and chaos. Nebr Sym Motiv 17:237–307