A Test of the Routine Activities and Neighborhood Attachment Explanations for Bias in Disorder Perceptions

Abstract

“Neighborhood disorder” refers to how people perceive neighborhoods as unsafe and disorganized. However, certain disorder cues may indicate disorder to some residents but not to others. There are many explanations for disorder perception bias, though few have been tested. This article uses data on 4,721 residents in 100 neighborhoods in Seattle to assess two explanations for biases: neighborhood attachment and routine activities. Using fixed-effect models, this article shows that neighborhood attachment and routine activities provide additional insight into disorder perceptions. Hanging out with teens and engaging in protective neighborhood activities, like watching neighbors’ property, have a strong positive influence on disorder perceptions. This study concludes by discussing alternative explanations for disorder perception bias and their impact on disorder theory as a whole.

Keywords

disorder perceptions neighborhoods routine activities neighborhood attachment

Introduction

A question in criminology has been lurking for some time: Is there a universal definition of disorder? Put another way, is disorder truly a “you’ll know it when you see it” phenomenon or are there biases in how people perceive disorder that have not been illuminated? Though there is not a cohesive definition of disorder (Kubrin, 2008), most theories of disorder begin with individuals and end with neighborhood dynamics. As individuals perceive disorder, they take it to mean that “no one cares” about their neighborhood or they are at risk for criminal victimization, and finally, they begin to withdraw into their homes, not engaging in neighborhood life (Wilson & Kelling, 1982). The reactions and subsequent behaviors of residents then reinforce that the neighborhood is out of control, enabling crime and further decay to occur (Wilson & Kelling, 1982). Many theories of disorder invoke this process while simultaneously overlooking the pivotal role that individuals’ perceptions of disorder have in disorder theory. The residents of a neighborhood need to perceive, interpret, and react to disorder the same way in order for the neighborhood dynamic that disorder generates to emerge.

Recent scholarship suggests disorder perceptions are not uniform. Hipp (2010) finds that in highly constrained ecological areas, individuals perceive disorder and crime differently. Sampson and Raudenbush (2004) find that Blacks and older adults perceive disorder at lower levels than their counterparts, and also that the racial composition of a neighborhood influences disorder perceptions, even when objective disorder is controlled for. Said differently, individuals do not simply report objective disorder; instead, their perceptions are tainted by a number of personal and situational biases. As a result, there are a host of explanations to account for these biases, many of which will be detailed in the coming section. Two explanations are perhaps the most powerful and simple: routine activities and neighborhood attachment. The routine activities theory suggests perceptions of one’s neighborhood is dependent on how and when an individual interacts with the public spaces of their neighborhood (Cohen & Felson, 1979); differential neighborhood exposure will generate different disorder perceptions. Neighborhood attachment suggests that people who are more attached to and satisfied with their community are less likely to see disorder as a problem (Taub, Taylor, & Dunham, 1984; Woldoff, 2002). While routine activities and neighborhood attachment are both simple and powerful explanations, neither has been used to test variation in disorder perceptions.

Scholars have long called for conceptual clarity surrounding disorder (Gau & Pratt, 2008; Harcourt, 2001; Kubrin, 2008). Investigating why disorder perceptions vary adds to research on neighborhoods and disorder in several ways. First, while there is a fairly consistent set of individual characteristics that are known to influence disorder perceptions, they are not exhaustive (Hipp, 2010; Sampson & Raudenbush, 2004)—these include age, race, and sex. Second, knowing the social and neighborhood contexts under which disorder perceptions are impacted will allow researchers and practitioners to anticipate the needs of a heterogeneous neighborhood. Third, any variation in disorder perceptions looms large over the current ways of theorizing disorder. Without durable explanations for variation, the causal processes disorder perceptions are related to at the individual and neighborhood levels are disrupted. Individuals’ perceptions are typically the first step in the neighborhood processes that disorder impacts. It is easy to assume that when individuals perceive disorder, it conveys to them that the neighborhood is out of control, no one cares, and they are at risk of being a victim of crime. But, if disorder perceptions are to impact individuals’ behavior and then neighborhood dynamics, homogeneity is a must, and there is evidence that this is not the case (Hipp, 2010; Latkin, German, Hua, & Curry, 2009; Sampson, 2009; Sampson & Raudenbush, 2004. If people do not perceive disorder at the same level, if different people perceive disorder differently, or if individuals interpret disorder differently, then it becomes much harder to link disorder to neighborhood processes. The most significant reason to test explanations for bias in perceptions is as follows: What if the variation in perceptions is not attributable to these causes? Should dominant explanations for biases in perceptions not account for variation in disorder perceptions, then the operationalization, conceptualization, and how residents interpret disorder can be questioned.

Thus, the primary goal of this study is to bring individuals to the forefront when discussing disorder theory and to test dominant explanations of bias in disorder perceptions. I build on an emerging literature on disorder perceptions (Hipp, 2010; Latkin et al., 2009; Sampson, 2009; Sampson & Raudenbush, 2004) by tracing the plethora of current explanations of variation in disorder perceptions. This study will replicate the findings of these studies by developing a core set of individual characteristics that produce variation in perception. To date, research in disorder perceptions has only begun to hypothesize, but not test, why some individuals perceive disorder differently. This study tests two explanations for differing disorder perceptions: neighborhood attachment and the routine activities explanations. I conclude with a detailed examination of why variation in disorder perceptions is of concern.

What Do We Know About Disorder Perceptions?

Individuals’ disorder perceptions are very rarely explicitly studied, even though they provide the foundation to disorder theory. What we currently know about individuals’ disorder perceptions and why they vary is limited and often tied together with perceptions of crime (Hipp, 2010). However, there are a few theoretical ways of understanding disorder perceptions (see Hipp, 2010, for a comprehensive detailing of relevant explanations). Below are some of the dominant explanations behind variation in disorder perceptions, beginning with explanations related to individuals’ socioeconomic and demographic characteristics, then moving to more theoretical explanations, such as neighborhood attachment and the routine activity theory.

First, individual sociodemographic contexts—like having children, marriage, and age cohort—may be powerful factors in shaping perceptions. Indeed, a few studies show that there are a consistent set of individual characteristics that are related to disorder perceptions. First, older individuals often perceive less disorder (Hipp, 2010; Sampson, 2009; Sampson & Raudenbush, 2004). Older individuals do not use their neighborhood as widely as younger individuals (Hindelang, Gottfredson, & Garofalo, 1978), perhaps staying within a sphere they are comfortable in, thus impacting their exposure to their neighborhood and its problems. Race is also consistently related to differences across disorder perceptions, most notably, non-Whites typically perceive less disorder than Whites. Sampson and Raudenbush (2004) suggest that Blacks see far less disorder than Whites due to exposure to high levels of disorder (through generations of segregation and stratification), which creates a higher bar for disorder to be considered a problem. Thus, Blacks, and perhaps other individuals similarly exposed to poor neighborhood conditions over extended periods of time, have a higher threshold for considering disorder a problem.

Disorder perceptions may also be affected by neighborhood tenure. Long-term residents’ disorder perceptions may be framed by their in-depth understanding of their community; unlike newer residents, long-term residents have years of information and experience in their community that frames their disorder perceptions. Long-term residents may consider disorder more of a problem given their existing personal and economic investment in a community (Hipp, 2010). To some extent, this applies to homeowners, regardless of length of time in the neighborhood; homeowners may be more attuned to their environment given their need to assure the profitability of their investment (Hipp, 2010). Skogan (1992) notes that the presence of neighborhood disorder is a depressor for the neighborhood’s housing market. However, long-term residents may be sensitized to neighborhood problems or even feel hopeless about change depending on the persistence of such problems (Xu, 2005). New residents, by contrast, may see their community in an overly positive light (Hipp, 2010).

Fear of crime also impacts disorder perceptions. Individuals’ fear of crime forces them to become more aware of their surroundings, thus shaping their disorder perceptions (Hipp, 2010; Jackson, 2004; LaGrange, Ferraro, & Supancic, 1992). There are differences across individuals in their fear of crime levels and perception of risk (Kanan & Pruitt, 2002; Roundtree & Land, 1996; Taylor & Hale, 1986). In general, women and older adults are most likely to fear crime (Joseph, 1997; Taylor & Hale, 1986; Warr, 1984), while non-Whites are less likely to fear crime (Joseph, 1997; Kanan & Pruitt, 2002; Roundtree & Land, 1996). However, non-Whites are significantly more likely than Whites to view being in their neighborhood alone at night as dangerous (Kanan & Pruitt, 2002). All these individuals, non-Whites, females, and older adults, may feel more at risk of victimization and therefore mind their surroundings to a greater degree and have higher disorder perceptions (Hipp, 2010). The altruistic fear of crime suggest that parents and guardians are more aware of their surroundings (Hipp, 2010; Warr & Ellison, 2000) and thus are more concerned about disorder and crime because they are responsible for protecting those in their care (Hipp, 2010). There is contradicting evidence for this explanation though. While Hipp (2010) finds that individuals with children perceive more crime and disorder, other research shows no link between children and perceptions of disorder (Ross & Mirowsky, 2001).

In addition to fear of crime and individual characteristics, an individual’s attachment to his or her neighborhood may also impact disorder perceptions. Behavioral neighborhood attachment can arise through two means: through neighboring and formal neighborhood activities (Woldoff, 2002). Behavioral attachment through neighboring comes in the form of having dinner with neighbors, knowing the names and faces of neighbors, or watching a neighbor’s property while they are away (Woldoff, 2002). Resident participation in neighborhood life has long been seen as the key to a successful community (Putnam, 2000; Swaroop & Morenoff, 2005), and when individuals are attached to the neighborhood, they are more likely to express satisfaction with their community (Woldoff, 2002). This satisfaction likely taints their perceptions of disorder—namely, that residents who are attached to and satisfied with their neighborhood are less likely to see disorder as a problem. There is limited evidence of this: Taub et al. (1984) show that residents interpret the physical environment of their neighborhood depending on the trajectory they believed the neighborhood to be on. Taub et al. find that residents who feel their neighborhood is on an upswing interpret cues like boarded storefronts to mean positive change is on the way. Formal neighborhood activities involve attending neighborhood association or watch meetings, block groups, or any formal neighborhood meeting involved with the police. Individuals engaged in these formal neighborhood groups are more aware of neighborhood problems simply through the role the neighborhood groups play. These individuals may perceive higher levels of disorder given their understanding of neighborhood problems.

Finally, the routine activities theory explains variation in disorder perceptions by suggesting that individuals will use the public spaces differently based on their daily activities, and this will increase or decrease their chance for victimization (Cohen & Felson, 1979). Routine activities will also afford individuals knowledge regarding neighborhood conditions (Hindelang et al., 1978; Hipp, 2010), thus shaping their disorder perceptions. For example, individuals out at night may see more public drunkenness, whereas those out at the neighborhood when school lets out will likely see more loitering. Furthermore, individuals in disadvantaged neighborhoods have less access to private space than individuals living in middle-class neighborhoods (Baldassare, 1982; Venkatesh, 2006). Poverty forces individuals into the public areas of neighborhoods, allowing for more knowledge about the neighborhood (Baldassare, 1982; Venkatesh, 2006). However, there are mixed results for the routine activities explanation, particularly in regard to age: Older adults perceive less disorder (Sampson, 2009; Sampson & Raudenbush, 2004) though they also restrict their activities in their neighborhood (Hindelang et al., 1978).

Current Study

In an effort to better understand what is responsible for variation in disorder perceptions among individuals, this study investigates the role of a variety of individual characteristics, routine activities, and neighborhood attachment in generating bias in perception. While in this study I develop a core set of individual characteristics that produce differences in perception, below, I also stipulate two primary hypotheses that guide my analyses. First, neighborhood attachment, both formally and informally, may impact disorder perceptions. Involvement in formal neighborhood activities may make residents more conscious of neighborhood problems, whereas less formal neighborhood involvement may enable residents to see their neighborhoods in an overly positive light. Thus, I suggest that

Hypothesis 1a: Formal neighborhood attachment will increase individuals’ disorder perceptions net of individual characteristics and neighborhood controls.

Hypothesis 1b: Informal neighborhood attachment will decrease individuals’ disorder perceptions net of individual characteristics and neighborhood controls.

Second, the structure of people’s daily routines may push them into certain public spaces of a neighborhood more frequently. This exposure affords people more information about the disorder that is present in their neighborhood, potentially altering perceptions. Thus, I suggest,

Hypothesis 2: Individuals whose routine activities allow them to engage the public spaces of a neighborhood more frequently will be exposed to disorder more frequently; thus, they will perceive more disorder net of individual characteristics and neighborhood controls.

In the following sections, I detailed how the current work tests the above hypotheses.

Method

Data

This study employs Miethe’s (1991) data set, “Testing Theories of Criminality and Victimization in Seattle, 1960-1990,” which is available from ICPSR. Collected in Seattle, the data set explores criminal opportunity theories of victimization, in addition to perceptions of disorder. The survey contains telephone interviews, collected in 1990, of Seattle residents nested within 100 neighborhoods or census tracts. Postcleaning, the final sample consists of 4,721 respondents within 100 neighborhoods. Respondents who are homeowners or college educated are overrepresented in the sample (Miethe & Meier, 1994). The collection of cases used in this article’s analyses reflects Miethe and Meier’s (1994) concerns: 84% of the sample is White, 14% of the sample is under 29 years, 71% of the sample has taken some college or more, and 67% of the sample is homeowners. Table 1 contains summary statistics for all variables.

Table 1.

Summary Statistics for Outcome and Predictors.

	M or %	SE
Outcome
Disorder Perception Scale	−0.003	0.01
Individual predictors—By substantive test
Female	50%	—
White	84%	~—
Black	6%	—
Other	8%	~—
Age 17-19	1%	—
Age 20-29	13%	—
Age 30-39	25%	—
Age 40-49	19%	—
Age 50-59	11%	—
Age 60 and over	31%	—
Married	54%	—
Single	23%	—
Divorced	11%	—
Widowed	10%	—
Household size	242%	—
Children under 6 in HHLD	13%	—
No high school diploma	6%	—
High school graduate	23%	—
Some college or more	71%	—
Income-centered	30861.3	334.3
Full time employment	49%	—
Home owner	67%	—
Length of time in NHD, 0-2 years	24%	—
Length of time in NHD, 3-6 years	25%	—
Length of time in NHD, 7+years	49%	—
Neighborhood victimization	60%	—
Neighborhood attachment variables
Formal neighborhood activities	53%	—
Recognize strangers on block	47%	—
Know neighbors’ first name	20%	—
Watch neighbors’ property	71%	—
Dinner with neighbors	51%	—
Helped neighbor solve problem	73%	—
Lifestyle/routine activities variables
Use public transportation at least 5 times/month	22%	—
Number of evenings out/week	2.44	0.03
Take an evening walk at least once a week	28%	—
Visited a bar/nightclub last week	21%	—
In places where teens hang out	19%	—

Note: For dummy variables, the mean is reported as a percentage. HHLD = Household; NHD = Neighborhood.

Even with the age of the data, there remain two substantial benefits of using Miethe’s data set. First, the data were collected outside of Chicago. Some of the most influential works on disorder are based on data collected in Chicago. While every major metropolitan area, including Seattle, may produce distinctive results, it is important to test disorder theories in other metropolitan contexts to get a sense of how disorder theory applies more generally to all urban areas. Moreover, unlike Chicago, Seattle has not seen the long-term, drastic concentration of poverty so often associated with high levels of disorder. Due to this, Seattle provides an excellent venue for testing biases in individuals’ perceptions of disorder. Should the results be similar to those found in other cities, it indicates that the mechanisms behind biases in disorder perceptions are similar across cities. Second, one of the strongest explanations behind variation in disorder perceptions arises out of the routine activities perspectives. Miethe’s data set offers several variables that serve as measures for this perspectives.

Dependent Variable

The dependent variable for the following analyses is respondents’ perceptions of neighborhood disorder. To construct an individual-level disorder measure, a disorder perception scale was created using four disorder cues. The disorder cues, coded Yes or No, include “Are there teenagers within three blocks surrounding your home?” “Is there litter or trash within three blocks surrounding your home?” “Are there abandoned or rundown buildings within three blocks surrounding your home?” and “Is there vandalism within three blocks surrounding your home?” The measures of disorder in this data set are different from other surveys; rather than asking the respondent “how much of a problem” disorder is, respondents report the occurrence of disorder in their immediate three blocks. Although this is an attempt at more objective estimates of disorder, however, biases in perception of disorder remain¹. The disorder perception scale has a Cronbach’s alpha of .61 and is standardized (appendix shows the item-test and item-rest scores for each item in the scale). This form of the disorder perception scale may be problematic, shown by the modest alpha. Multiple ways of constructing the dependent variable were undertaken, including Poisson models and a factor score. All three methods yield highly similar results and these results are also analogous to those found by Hipp (2010) and Sampson and Raudenbush (2004). Given that the results here have been replicated by others, the linear form was assumed.²

Independent Variables

Individual predictors come in five categories: demographic characteristics, life-course characteristics, aspects of neighborhood life, neighborhood attachment, and lifestyle/routine activities.³ The demographic characteristics include race, sex, and age. Gender is a dummy coded variable where 1 represents females. Race is measured through two dummy variables representing Blacks and other races. Age has been collapsed into six categories: 17 to 19, 20 to 29, 30 to 39, 40 to 49, 50 to 59, and 60 and above, with the 20 to 29 variable being the reference. Life-course variables measure those mutable aspects of individuals and include marital status, education, household size, children under 6 in the household, homeownership, full-time employment, and income. Similar to age, marital status is broken down into four dummy variables: married, single, divorced, and widowed. Education is divided into three dummy variables representing no high school diploma, high school graduate or equivalent, and some or more college. Household size is coded from 0 to 4, with 0 representing single-person households and 4 representing households with five or more members. Although this is not a direct measure of whether the respondent has children, respondents are asked whether there are children less than six years old present in their household. This variable is a dummy variable where 1 represents the presence of children under six years old in the household. Like children under 6, homeownership and full-time employment are dummy variables. Income is a continuous variable. There was missing data on the income variable; nearly 14% of respondents did not report their income. Income was therefore imputed and included in the model as income divided by 1,000.⁴ There are two general neighborhood variables: length of time in neighborhood and neighborhood victimization. Length of time in the neighborhood, originally, was a categorical variable, measured from 0 to 9, with 0 representing less than 1 year of residency in the community and 9 being 9 or more years of residency. Due to the skewed distribution for this variable, it has been split into three dummies of relatively equal categories: (a) 0 to 2 years in the neighborhood, (b) 3 to 6 years in the neighborhood, and (c) 7 years or more in the neighborhood. Neighborhood victimization is a dummy variable created from a series of victimization questions. Respondents were asked if they had been a victim of the following crimes in their neighborhood, specifically within three blocks of their home: a home break-in (both attempted and actual), having property stolen from their yard, being physically assaulted by a stranger, mugging, vandalism, or a car theft. If a respondent answered “yes” to any one of these questions, they were given a score of 1 for neighborhood victimization.

The neighborhood attachment variables include measures of both formal and informal attachment to one’s neighborhood. The variable “formal neighborhood activities” is a dummy variable denoting whether the respondents currently participate in a neighborhood block association or in a block activity with the Seattle Police Department. The following are informal neighborhood attachment variables. “Recognize strangers on block” is a dummy variable denoting that the respondents report that they can differentiate between residents and strangers on their block. “Know neighbors first name,” a dummy variable, shows that the respondents report knowing most or all of their neighbor’s names on their block. “Watch neighbors property” is a dummy variable signifying that the respondent has watched a neighbor’s property when not at home. “Dinner with neighbors” is a dummy variable denoting that the respondent has had dinner or lunch with a neighbor. Finally, the dummy variable “Helped a Neighbor Solve a Problem” denotes when a respondent has helped a neighbor with a problem.

Five variables consist of the routine activities variables test. The first variable “whether or not the respondent frequently uses public transportation” is a dummy variable showing the respondent uses public transportation five or more times in a month. “Number of evenings out” is a measure of the number of days in a week that the respondent goes out in the evening; the variable is centered. The variable “takes an evening walk at least once a week” is a dummy variable that shows that the respondent does walk at night one or more times in a week. “Visited a bar/nightclub last week,” also a dummy variable, indicates that the respondent has been to a bar or nightclub that serves alcohol during the previous week. Finally, “In places where teens hang out” is a dummy variable that denotes that, during the last week, the respondent was in a public place where groups of teenagers or young adults were hanging out on the street.

Modeling

To analyze the effects of individual characteristics on disorder perceptions, this study uses a fixed-effects technique—an individual-level model in which the effect of the neighborhood is controlled for through a series of dummy variables representing each neighborhood unit. The benefit of fixed effects is that they eliminate all between-unit variation (Halaby, 2004); thus, any differences in disorder perceptions among respondents are not due to their neighborhood environment but rather are a function of the individual.⁵ A benefit of a within-neighborhood effects model is the partial control of selection bias. In this case, selection bias would enter the model if disorder perceptions are influenced by omitted characteristic of the neighborhood or individual that is related to who is in or remains in the neighborhood. Because fixed-effects models control for the observed and unobserved characteristics of the neighborhood, so long as the omitted characteristics are at the neighborhood level, selection bias is controlled for (Harding, 2003). I estimate robust standard errors adjusted for clustering. Finally, to be sure multicollinearity was not present, variance inflation factors, or VIF, statistics were run on all the models. The average VIF for all predictors is approximately 2, with the highest VIF associated with length of residence, 7-9 years at 5.34 in the lifestyle/routine activities model, well below 10 and thus not a concern.

Results

In Table 2, I begin with discussing the baseline model that establishes the set of individual characteristics that are typically associated with variation in disorder perceptions. While the baseline model is displayed in both Tables 2 and 3 as a comparison point, I will only discuss it here. In addition, I discuss the baseline results in sections: demographic variables, life-course variables, and neighborhood variables, though they are part of the same model. Similar to the results of Sampson and Raudenbush (2004) and Hipp (2010), Blacks, non-Whites, and people above 40 perceive less disorder than Whites or 20- to 29-year-olds. Important to note is the relationship between age and disorder perceptions: Higher age categories see larger, negative effects on disorder perceptions. The effect of age runs contradictory to the fear of crime explanation of bias in disorder perceptions: Individuals who have physical reasons to fear crime, like older adults and women, should perceive more disorder. However, because older individuals are less likely to use their neighborhood in ways that younger individuals do (Hindelang et al., 1978), they may have less information about disorder and may thusly perceive less. Females perceive more disorder than males. There are also several significant life-course variables. Single and widowed individuals perceive less disorder than married individuals, though for single individuals, the effect is only moderately significant (p < .10). Interestingly, level of education also plays a role in disorder perceptions: Individuals with some or more college perceive significantly more disorder than those with only a high school education. People with larger households perceive less disorder; this may have something to do with guardianship as a larger household has a greater ability to look out for its members than smaller households. The altruistic fear explanation of variation in disorder perceptions suggests that guardians are more concerned about disorder and crime because they are responsible for protecting those in their care (Hipp, 2010; Warr & Ellison, 2000). This is supported here: Individuals with children under the age of 6 in their home perceive significantly more disorder than those with older or no children in their home. As for the neighborhood variables, only neighborhood victimization is a significant predictor of disorder perceptions. This is expected: People who have been victimized in their community are likely to see more problems with it. It is also important to discuss the nonsignificant variables. Neighborhood tenure is theorized to have an impact on disorder perceptions. Long-term residents are thought to perceive more disorder given their knowledge about the neighborhood problems, while newer residents have a more positive picture of their neighborhood given their inexperience with their neighborhood. Here, long-term residents (3 years and above) are no different than short-term residents (the reference category) in regards to disorder perceptions. Homeowners are expected to perceive more disorder than renters simply because they have a long-term investment to protect (Hipp, 2010). Here we see that homeowners are no different from renters in their disorder perceptions. Here, two explanations of bias in disorder perceptions, homeownership or economic investment, and neighborhood tenure are unsupported.

Table 2.

Baseline and Neighborhood Attachment Fixed Effect Regression Models Predicting Disorder Perceptions.

	Baseline model		Neighborhood attachment model
	B	SE	B	SE
Female	.0515***	(0.0169)	.0543***	(0.0169)
Black	−.237***	(0.0543)	−.222***	(0.0547)
Other	−.153***	(0.0395)	−.135***	(0.0392)
Age 17-19	−.0998	(0.125)	−.120	(0.128)
Age 30-39	−.0532	(0.0335)	−.0683**	(0.0337)
Age 40-49	−.108**	(0.0418)	−.133***	(0.0425)
Age 50-59	−.173***	(0.0448)	−.192***	(0.0458)
Age 60 and over	−.295***	(0.0463)	−.301***	(0.0457)
Single	−.0476*	(0.0278)	−.0279	(0.0275)
Divorced	−.0543	(0.0341)	−.0363	(0.0344)
Widowed	−.113***	(0.0302)	−.0961***	(0.0292)
Household size	−.0366***	(0.00928)	−.0350***	(0.00899)
Children under 6 in HHLD	.0801**	(0.0350)	.0698**	(0.0342)
No high school diploma	−.0236	(0.0373)	−.0102	(0.0367)
Some college	.0840***	(0.0213)	.0717***	(0.0211)
Income, centered	−.0009	(0.0005)	−.0009	(0.0005)
Full time employment	−.00469	(0.0236)	−.00117	(0.0230)
Home owner	−.00983	(0.0195)	−.0263	(0.0191)
Length of time in NHD, 3-6 years	.00255	(0.0239)	−.0126	(0.0243)
Length of time in NHD, 7+years	−.0340	(0.0267)	−.0628**	(0.0270)
Neighborhood victimization	.165***	(0.0194)	.154***	(0.0193)
Formal neighborhood activities			.0705***	(0.0240)
Recognize strangers on block			−.0341*	(0.0199)
Know neighbors’ first name			−.0261	(0.0241)
Watch neighbors’ property			.0668***	(0.0231)
Dinner with neighbors			.0196	(0.0218)
Helped neighbor solve Problem			.0711***	(0.0257)
Constant	.116**	(0.0553)	.0382	(0.0557)
Observations	4721		4721
R ²	0.083		0.094
Number of tracts	100		100

Notes: Robust standard errors in parentheses.

p < .10. **p < .05. ***p < .01.

Table 3.

Baseline and Routine Activities/Lifestyle Fixed Effect Regression Models Predicting Disorder Perceptions.

	Baseline model		Routine activities model
	B	SE	B	SE
Female	.0515***	(0.0169)	.0592***	(0.0169)
Black	−.237***	(0.0543)	−.230***	(0.0526)
Other	−.153***	(0.0395)	−.143***	(0.0382)
Age 17-19	−.0998	(0.125)	−.0704	(0.119)
Age 30-39	−.0532	(0.0335)	−.0293	(0.0349)
Age 40-49	−.108**	(0.0418)	−.0814*	(0.0423)
Age 50-59	−.173***	(0.0448)	−.132***	(0.0454)
Age 60 and over	−.295***	(0.0463)	−.230***	(0.0475)
Single	−.0476*	(0.0278)	−.0643**	(0.0277)
Divorced	−.0543	(0.0341)	−.0708**	(0.0348)
Widowed	−.113***	(0.0302)	−.107***	(0.0292)
Household size	−.0366***	(0.00928)	−.0363***	(0.00914)
Children under 6 in HHLD	.0801**	(0.0350)	.0896**	(0.0347)
No high school diploma	−.0236	(0.0373)	−.0261	(0.0364)
Some college	.0840***	(0.0213)	.0705***	(0.0207)
Income, centered	−.0009	(0.0005)	−.0009	(0.0006)
Full time employment	−.00469	(0.0236)	−.00431	(0.0235)
Home owner	−.00983	(0.0195)	−.00600	(0.0196)
Length of time in NHD, 3-6 years	.00255	(0.0239)	−.000407	(0.0241)
Length of time in NHD, 7+years	−.0340	(0.0267)	−.0289	(0.0265)
Neighborhood victimization	.165***	(0.0194)	.158***	(0.0191)
Use public transportation at least 5 times/month			.0227	(0.0244)
Number of evenings out/week			.0198***	(0.0056)
Take an evening walk at least once a week			−.0140	(0.0194)
Visited a bar/nightclub last week			−.008	(0.0254)
In places where teens hang out			.167***	(0.0241)
Constant	.116**	(0.0553)	.0561	(0.0544)
Observations	4721		4721
R ²	0.083		0.099
Number of tracts	100		100

Note: Robust standard errors in parentheses.

p < .10. **p < .05. ***p < .01.

In addition to the baseline model, Table 2 shows the results of the neighborhood attachment model. Neighborhood attachment does offer insight into variation in perceptions. When individuals are involved in formal neighborhood activities, they perceive higher levels of disorder. Thus, it seems that individuals actively working to combat neighborhood problems in formalized settings like Chicago Alternative Policing Strategy (CAPS) meetings or block groups are more attuned to disorder. This may be due to the simple nature of the activity; neighborhood groups are often formed to combat neighborhood problems and bring such problems to the forefront for their members. Next, an individual’s ability to recognize strangers on their block predicts lower levels of disorder perceptions, though this effect is small and only moderately significant. Given that people on the street can often be seen in a disorderly light, the ability to recognize strangers from residents may allow for residents to perceive disorder and people in a less threatening manner. In addition, watching a neighbor’s property and helping neighbors solve problems both have a positive and significant effect on disorder perceptions. Like participation in formal neighborhood activities, when respondents report engaging in these two activities, paying attention to disorder becomes rather important. Thus, neighborhood attachment activities that can be seen as protective activities have a strong, positive relationship with disorder perceptions.

Table 3 displays the results of the routine activities model contrasted to the baseline model. The number of evenings out has a positive impact on disorder perceptions—as individuals spend more time out of the house at night, they perceive greater amounts of disorder. Regularly being in an area where teens hang out, however, does have a dramatic and positive effect on disorder perceptions. Here the size of the slope is roughly equivalent to the effect of neighborhood victimization. Regularly being in an area where teens hang out should increase disorder perceptions because, depending on the context, teens hanging out can be seen as disorderly. Exposure to teens hanging out may also expose one to other disorder cues, like littering or drug sales. Thus, engaging in disorderly activity does not necessarily reduce one’s perception of disorder, simply being exposed to disorder increases one’s disorder perceptions of it. Only about 19% of the respondents report regularly hanging out around with teens. This is the evidence that a mildly deviant lifestyle impacts increases disorder perceptions.

Discussion

Perceptions of disorder play a strong role in how individuals view a neighborhood. Previously, individuals’ perceptions of disorder were thought to be primarily a function of the neighborhood environment with individuals’ disorder perceptions reflecting the “on the ground” reality of disorder in the neighborhood (Ross & Mirowsky, 1999).The most recent additions to disorder theory show that individuals’ disorder perceptions can be socially mediated by stigma and stereotypes at the neighborhood level (Sampson, 2009; Sampson & Raudenbush, 2004). Even with these strides, the role of the individual has again been excluded from much discourse on disorder perceptions. This article inquires into which personal and social contexts for individuals, particularly neighborhood attachment and routine activities, contribute to their disorder perceptions net of the neighborhood environment. Using fixed-effects models, this study shows that neighborhood attachment and routine activities help us understand in greater detail why individuals differ in their disorder perceptions.

The primary results of this study are as follows: Both neighborhood attachment and routine activities provide insight into variation in disorder perceptions. When testing the neighborhood attachment hypothesis, the strongest impact came from the attachment variables that involved protective activities, like formal neighborhood activities or watching a neighbor’s property. These activities make residents “guardians” allowing them to purposefully pay greater attention to disorder. In regard to the routine activities theory, regularly being in an area where teens hang out has a sizable effect on disorder perceptions. A similar effect is found in other studies; Latkin et al. (2009) find that individuals who regularly hang out with drug users spend more time on the street and thus have higher disorder perceptions. Thus, the routine activities explanation for disorder perceptions holds if one is around deviant and disorderly activities. The effect of routine activities on biases in disorder perceptions is not exclusively related to high levels of neighborhood exposure via routine activities, but also about exposure to disorderly areas and people. Interestingly, neither the neighborhood attachment nor the routine activities variables reduced the effects of the main individual characteristics, such as non-Whites or older ages, associated with disorder perceptions. Systematic bias in disorder perceptions originates from these core characteristics: The strongest predictors of disorder perceptions remain age, race, and neighborhood victimization. Combining this core set with routine activities and neighborhood attachment provides for a clearer understanding of differences in disorder perception, though it is clear that those perceptual differences still exist and that research is far from understanding their root cause.

The limitations of this study primarily surround data issues. First, because these data originate from Seattle, there are few Blacks in the sample. Ideally, this study would have been able to test the association between race and disorder perceptions, but the small number of Blacks did not afford much variation. In addition, the disorder measure can be seen as problematic. Due to how disorder was assessed, the disorder scale can be also constructed as a count variable. Comparative tests were run on linear and Poisson models of disorder perceptions; no significant differences were found. However these limitations are outweighed by several major benefits of the data set. First, disorder is often assessed in Chicago (see, for example, Sampson & Raudenbush, 1999). Studying Seattle offers an interesting glimpse into disorder perceptions. This data set also has a slew of routine activities and neighborhood attachment variables that are often unavailable in data sets aimed at capturing neighborhood phenomena.

This study shows that many hypotheses regarding systematic bias in disorder perceptions fall short of being powerful explanations (homeownership, neighborhood tenure) while others emerge as important contexts to consider (routine activities and neighborhood attachment). Yet, differences still remain and we are left searching for other potential explanations for these differences. One suggestion might be that some individuals are more tolerant of disorder. Indeed Sampson and Raudenbush (2004) suggest that Black may have a higher threshold for disorder simply because they have been exposed to generations of disadvantaged neighborhoods that are typically high in disorder. In addition, there is some evidence to suggest that even though neighborhood problems are something residents want changed, they consider them a low priority in their lives and thus become tolerant of them (Wells, Schafer, Varano, & Bynum, 2006). Unfortunately, individuals’ inactivity regarding crime and delinquency is often misunderstood as tolerance for it. Rather, Sampson and Bartusch (1998) show that “it thus appears that there is an ecological structuring to normative orientations—‘cognitive landscapes’ where crime and deviance are more or less expected and institutions of criminal justice are mistrusted” (p. 800). Later research finds that legal cynicism explains the durability of homicide in Chicago neighborhoods during a drastic crime decline in the 1990s (Kirk & Papachristos, 2011). Another explanation may surround neighborhood historical contexts. Disorder perceptions may be affected by one’s history with their neighborhood. For example, while not explicitly dealing with disorder, in Villa Victoria, Small (2004) finds that two generations of residents have different “neighborhood narratives”; the contexts the residents use to interpret their neighborhood are disparate across generations. Without understanding individuals’ personal historical context, it is impossible to comprehend how individuals view their communities (Small, 2004), naturally, this includes disorder.

As we search for the golden rule that enables us to better understand why individuals perceive disorder so differently, perhaps it is more prudent to ask, “Do we have a good understanding of what disorder means, at least from the individual perspective?” It is possible that the bias we see in perceptions is reflective of individuals interpreting disorder in ways that current research, including this study, does not tap. Kubrin (2008) suggests that “definitions of disorder used by researchers and officials studying and practicing broken windows policing are not necessarily consistent with residents’ perceptions in their own community” (p. 206). Indeed, Gau and Pratt (2008) found that respondents could not differentiate between crime and disorder. If we examine the long history of urban ethnography, we find several examples that show disorder being interpreted outside of its traditional definition. Sudhir Venkatesh (2006) details how the underground economy, such as drug selling or prostitution, is an economically functional aspect of a Black community in south Chicago and seen in a different, less criminogenic light by residents. By approaching disorder perceptions from the perspective of offenders, St. Jean (2007) shows that offenders, particularly those selling drugs, understand the nuances surrounding the production of disorder in their neighborhoods and do not directly link its presence to their criminal intentions. St. Jean (2007), similar to Venkatesh (2006), also finds that individuals living in neighborhoods where an illicit economy is prominent do not often see disorder itself as a problem, but rather as social situations tied to financial gain. In this case, the issue for residents is not, say, the act of prostitution or drug dealing, it is where within their neighborhood these crimes occur (St. Jean, 2007).

In conclusion, without substantial evidence supporting explanations of bias, the main consequence of variation in disorder perceptions calls into question our current understanding of how individuals interpret disorder. The importance of individuals’ perceptions to disorder demands that all individuals perceive disorder in the same fashion and that disorder conveys the same neighborhood quality and fear of crime meanings to those individuals. Given the problems of conceptualization and the evidence that there is variation in perceptions, the meaning of disorder is bound to shift. The retraction and addition of several explanations of bias disorder perceptions potentially opens a new research arena where disorder is not defined by the researcher but instead by the resident.

Footnotes

Appendix

Scale Statistics for the Disorder Scale

	Item–test correlation	Item–rest correlation	α (if variable is not included)
Teens	0.6805	0.3901	0.5311
Litter	0.6979	0.4133	0.5120
Abandoned buildings	0.6349	0.3219	0.5811
Vandalism	0.6971	0.4136	0.5118
Overall α			0.6052

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was partially supported by the Housing and Urban Development Dissertation Research.

Notes

References

Baldassare

(1982). The effects of neighborhood density and social control on resident satisfaction. Sociological Quarterly, 23, 95-105.

Cohen

L. E.

Felson

(1979). Social change and crime rate trends: A routine activities approach. American Sociological Review, 44, 588-608.

Gau

J. M.

Pratt

T. C.

(2008). Broken windows or window dressing citizens’ (in) ability to tell the difference between disorder and crime. Criminology & Public Policy, 7, 163-194.

Halaby

C. N.

(2004). Panel models in sociological research: Theory into practice. Annual Review of Sociology, 30, 507-544.

Harcourt

B. E.

(2001). Illusion of order: The false promise of broken windows policing. Cambridge, MA: Harvard University Press.

Harding

D. J.

(2003). Counterfactual models of neighborhood effects: The effects of neighborhood poverty on dropping out and teenage pregnancy. American Journal of Sociology, 109, 676-719.

Hindelang

M. J.

Gottfredson

M. R.

Garofalo

(1978). Victims of personal crime: An empirical foundation for a theory of personal victimization. Cambridge, MA: Ballinger.

Hipp

J. R.

(2010). Resident perceptions of crime and disorder: How much is “bias”, and how much is social environment differences?*. Criminology, 48, 475-508.

Jackson

(2004). Experience and expression: Social and cultural significance in the fear of crime. British Journal of Criminology, 44, 946-966.

10.

Joseph

(1997). Fear of crime among Black elderly. Journal of Black Studies, 27, 698-717.

11.

Kanan

J. W.

Pruitt

M. V.

(2002). Modeling fear of crime and perceived victimization risk: The (in) significance of neighborhood integration. Sociological Inquiry, 72, 527-548.

12.

Kirk

D. S.

Papachristos

A. V.

(2011). Cultural mechanisms and the persistence of neighborhood violence. American Journal of Sociology, 116, 1190-1233.

13.

Kubrin

C. E.

(2008). Making order of disorder: A call for conceptual clarity. Criminology & Public Policy, 7, 203-214.

14.

LaGrange

R. L.

Ferraro

K. F.

Supancic

(1992). Perceived risk and fear of crime: The role of social and physical incivilities. Journal of Research in Crime & Delinquency, 29, 311-334.

15.

Latkin

C. A.

German

Hua

Curry

A. D.

(2009). Individual-level influences on perceptions of neighborhood disorder: A multilevel analysis. Journal of Community Psychology, 37, 122-133.

16.

Miethe

T. D.

Testing Theories of Criminality and Victimization in Seattle, 1960-1990 [Computer file].2nd ICPSR version. Blacksburg, VA: Virginia Polytechnic Institute [producer], 1991; Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 1998. DOI:10.3886/ICPSR09741.v2

17.

Miethe

Meier

R. F.

(1994). Crime and its social context: Towards an integrated theory of criminal offender, victims, and situations. Albany: SUNY Press.

18.

Putnam

R. D.

(2000). Bowling alone: The collapse and revival of American community. New York, NY: Simon & Schuster.

19.

Raudenbush

S. W.

Bryk

A. S.

(2002). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: SAGE.

20.

Ross

C. E.

Mirowsky

(1999). Disorder and decay: The concept and measurement of perceived neighborhood disorder. Urban Affairs Review, 34, 412-432.

21.

Ross

C. E.

Mirowsky

(2001). Neighborhood disadvantage, disorder, and health. Journal of Health and Social Behavior, 42, 258-276.

22.

Roundtree

P. W.

Land

K. C.

(1996). Perceived risk versus fear of crime: Empirical evidence of conceptually distinct reactions in survey data. Social Forces, 74, 1353-1376.

23.

Sampson

R. J.

(2009). Disparity and diversity in the contemporary city: Social (dis)order revisited. British Journal of Sociology, 60, 1-31.

24.

Sampson

R. J.

Bartusch

D. J.

(1998). Legal cynicism and (subcultural?) tolerance of deviance: The neighborhood context of racial differences. Law & Society Review, 32, 777-804.

25.

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, 603-651.

26.

Sampson

R. J.

Raudenbush

S. W.

(2004). Seeing disorder: Neighborhood stigma and the social construction of “broken windows.” Social Psychology Quarterly, 67, 319-342.

27.

Skogan

W. G.

(1992). Disorder and decline: Crime and the spiral of decay in American neighborhoods. Berkeley: University of California Press.

28.

Small

M. L.

(2004). Villa Victoria: The transformation of social capital in a Boston Barrio. Chicago, IL: University of Chicago.

29.

St. Jean

P. K. B.

(2007). Pockets of crime: Broken windows, collective efficacy, and the criminal point of view. Chicago, IL: University of Chicago Press.

30.

Swaroop

Morenoff

J. D.

(2005). Building community: The neighborhood context of social organization. Social Forces, 84, 1665-1696.

31.

Taub

R. P.

Taylor

D. G.

Dunham

J. D.

(1984). A theory of reactions to crime and urban neighborhood change paths of neighborhood change (pp. 167-181). Chicago, IL: University of Chicago Press.

32.

Taylor

R. B.

Hale

(1986). Testing alternative models of fear of crime. Journal of Criminal Law and Criminology, 77, 151-189.

33.

Venkatesh

(2006). Off the books: The underground economy of the urban poor. Boston, MA: Harvard University Press.

34.

Warr

(1984). Fear of victimization: Why are the elderly and women more afraid? Social Science Quarterly, 65, 681-702.

35.

Warr

Ellison

C. G.

(2000). Rethinking social reactions to crime: Personal and altruistic fear in family households. American Journal of Sociology, 106, 551-578.

36.

Wells

Schafer

Varano

Bynum

(2006). Neighborhood residents’ production of order: The effects of collective efficacy on responses to neighborhood problems. Crime & Delinquency, 52, 523-550.

37.

Wilson

J. Q.

Kelling

G. L.

(1982). Broken windows: The police and neighborhood safety. Atlantic Monthly, 127, 29-38.

38.

Woldoff

R. A.

(2002). The effects of local stressors on neighborhood attachment. Social Forces, 81, 877-116.

39.

(2005). Discovering the impact of community policing: The broken windows thesis, collective efficacy, and citizens’ judgment. Journal of Research in Crime and Delinquency, 42, 147-186.