Sydes, Michelle - QUT ePrints

Sydes, Michelle - QUT ePrints

This is the author’s version of a work that was submitted/accepted for publication in the following source: Sydes, Michelle L, Wickes, Rebecca L, & Hi...

NAN Sizes 0 Downloads 9 Views

Recommend Documents

PDF (142kB) - QUT ePrints
in excess of 20% expected in 2007 in each global region (Kane, 2004; .... the top ten selling video games were based on

PDF 177kB - QUT ePrints
7 Singer & Friedlander Ltd v John D Wood & Co [1977] 2 EGLR 84. 8 Zubaida v Hargreaves [1995] 1 EGLR 127. 9 Singer & Fri

PDF (71kB) - QUT ePrints
(with 26.2 millions of skier visits) in North America and generated annual revenue over USD $2 billion. Four conglomerat

PDF 88kB - QUT ePrints
2 Renee Gastaldon, Dividing Fences and Dangerous or Intrusive Trees: The Draft ... The issue of spite hedges or spite fe

PDF (252kB) - QUT ePrints
Apr 2, 2015 - Ted Sarandos, Netflix's chief content officer, told triplej's Hack program that they are: open to producin

McEwan, Mitchell - QUT ePrints
Jul 21, 2012 - falling into different categories in their typology for both a racing and a golf game, failed to fully su

PDF 256kB - QUT ePrints
George Weston Foods Limited v Goodman Fielder Ltd. 24Bread packaging declared in large typeface: “Now Twice the Fibre*

PDF 504kB - QUT ePrints
Hunter, 2005) and, on the French Minitel network, where user pseudo creation subverts the official terms and conditions

PDF (222kB) - QUT ePrints
'Teenagers are scientifically proven to be stupid,' he'd say. 'Try to fight ... 'It was a dare,' Richard Crinkle said in

wahawaewao - QUT ePrints
Auckland University of Technology's Motion Capture studio, and the expansive, rugged, physical landscape of Mt Ida, Kyeb

This is the author’s version of a work that was submitted/accepted for publication in the following source: Sydes, Michelle L, Wickes, Rebecca L, & Higginson, Angela (2014) The spatial concentration of bias: An examination of the community factors that influence residents’ perceptions of bias crime. Australian and New Zealand Journal of Criminology, 47 (3), pp. 409-428. This file was downloaded from:

c Copyright 2014 The Author(s)

Notice: Changes introduced as a result of publishing processes such as copy-editing and formatting may not be reflected in this document. For a definitive version of this work, please refer to the published source:

The spatial concentration of bias: An examination of the community factors that lead to resident reported bias incidents

Michelle Sydes School of Social Science, University of Queensland Rebecca Wickes School of Social Science, University of Queensland Angela Higginson Institute for Social Science Research, University of Queensland

Running Head: The spatial concentration of bias

Word Count: 7810


Abstract: Emerging scholarship indicates that bias crimes are concentrated in particular types of places. Currently only a small number of studies consider the ecological factors that influence official reports of bias crime. Results from these studies indicate that the community processes and structures associated with the occurrence of non-bias crime may operate differently for bias crime. We use administrative and survey data from approximately 4,000 residents living across 148 communities in Brisbane, Queensland to examine the ecological drivers of bias crime. Using multi-level logistic regression, we examine the community and household factors associated with residents’ perceptions of bias crime. Here we focus not only on the structural demographics of the community, but the degree to which community cohesion influences whether or not residents perceive bias crime as a problem in their community. We find that poverty and ethnic diversity are positively associated with residents’ perceptions of bias crime. Further residents living in communities with higher levels of community cohesion are less likely to perceive bias crime as a problem in their community. The level of community cohesion fully mediates the impact of ethnic diversity and partially mediates the effect of poverty on residents’ perceptions of bias crime.

Key Words: Bias Crime, Social Cohesion, Ethnicity, Community


Introduction Bias crime1 refers to unlawful, violent, destructive or threatening behaviour in which the perpetrator is motivated by prejudice towards the victim’s race, ethnicity, religion, sexual orientation, gender identity or physical/mental impairment (Craig 1999, Green, McFalls and Smith, 2001; Victoria Equal Opportunity and Human Rights Commission (VEOHRC), 2010). The impact of bias crime is devastating for the victim (Barnes and Ephross, 1994; Boeckmann and Turpin-Petrosino, 2002; Herek, Cogan and Gillis, 1999; Garcia & McDevitt, 1999), however, the effects of bias crime are felt throughout the targeted community and society more broadly as they undermine diversity, harmony and equality (Boyd, Berk and Hamner, 1996; Finn, 1988; Ignaski, 2001; VEOHRC, 2010). Incidents of ethnically motivated bias crime, the focus of this study, captivate global news headlines. With over 1000 active hate groups in the United States alone, Neo Nazi groups, white nationalist groups, the Ku Klux Klan, neo-Confederate groups and black separatist groups continue to thrive united by a shared hatred towards an entire group of people (Southern Poverty Law Center, 2012). In England and Wales, 35,816 ethnically motivated crimes were reported to police between 2011 and 2012 (Home Office, 2012). Some evidence from Australia also suggests that ethnically motivated bias crime is a problem many minorities face (Human Rights and Equal Opportunity Commission’s (HREOC), 2004; Johnson, 2005; Mason, 2009). Studies in the United States indicate that bias crimes are concentrated in particular types of communities (Grattet, 2009; Green, Strolovitch and Wong, 1998; Lyons, 2007; Lyons, 1

Scholars argue that ‘bias crime’ more accurately describes incidents motivated by prejudice as it emphasizes the offender’s bias, rather than hatred, towards a victim (Iganski 2008; Lawrence 1999; Perry 2005/2006). It is a term coined in the United States in the 1980’s in response to an emergence of violence directed specifically at homosexual, African American and Jewish populations (Green et al 2001). In the literature bias crime is often used interchangeably with hate crime (Perry 2005/2006).


2008). Yet the community characteristics and processes associated with bias crimes are distinct from non-bias crime (Grattet, 2009; Green et al., 1998; Lyons, 2007; Lyons, 2008), providing some evidence that particular community processes may differentially predict bias crime for different targeted populations. This has led some scholars to argue that specific ecological theories are needed to explain bias crime (Disha, Cavendish and King, 2011; Grattet, 2009; Green et al., 1998; LeVine and Campbell, 1972; Lyons, 2007; Lyons, 2008). To date the majority of ecological studies of bias crime take place in the United States, a country with a unique racial/ethnic history. Whether or not theories derived from the United States’ context generalise to an Australian setting has not been tested. Further, almost all ecological studies of bias crime rely solely on administrative data. There are two major problems with the use of official statistics to understand bias crime: victims are unlikely to report their victimisation to the police, and police statistics are often incorrect or incomplete (Iganski, 2002; Levin, 1999; Perry, 2001; VEOHRC, 2010). This latter point is particularly problematic in Australia where only Victoria and New South Wales regularly collect information on bias crime incidents (New South Wales Police Force, 2011; Victoria Police, 2013). As a nation built on the immigrant experience and one of the most ethnically diverse populations in the world (ABS, 2010), Australia provides a unique context in which to examine bias crime. In this paper, we progress the bias crime literature in two ways. First, we combine survey data from the Australian Community Capacity Study (ACCS) – a multi-city longitudinal study of community processes, disorder and victimisation – with census data to test the efficacy of U.S. derived ecological theories of bias crime in an Australian context. Using multi-level logistic regression, we examine the community and household factors that are associated with perceptions of bias crime across 148 state suburbs in Brisbane, Australia. Here we focus not only on the structural demographics of the community, but the degree to 4

which changes in these structural demographics, in particular ethnicity, influence bias crime. Second, considering the serious limitations associated with police incident data on bias crime, and in line with recent advances in England, Scotland and Victoria, Australia (Police Scotland 2013; HM Government, 2012; VEOHRC, 2010), we rely on residents’ accounts of bias crime in their community as opposed to police data. The ACCS survey asks over 4,000 residents to indicate if individuals in their community are attacked or harassed because of their skin colour, ethnicity or religion. This approach provides a reliable assessment of the extent of community problems from those who have the greatest experience of living in the community (Skogan, 2012) and allows us to more fully explore the community characteristics that predict bias motivated attacks in the community context. Additionally, a survey approach that centres on residents’ perceptions of crime and disorder is particularly useful when the goal is to identify incidents that are unlikely to be called into the police (Skogan, 2012). Our approach is also fruitful for bias crime prevention efforts as it may provide a more reliable identification of the types of locations where bias crimes are occurring. In what follows we provide a review of the neighbourhood effects literature on bias crime and unpack generalised theories of crime such as social disorganisation theory and its more contemporary reformulations, as well as specific theories of bias crime that treat the etiology of bias crime as distinct from non-bias crime. We then discuss the ACCS in greater depth and present our main findings. Literature Review Several inter-related theories are used to explain the concentration of bias crime in particular types of urban communities in the criminological literature. These include generalised ecological theories of crime such as social disorganisation and collective efficacy


as well as specific ecological theories of bias crime like the defended neighbourhood and power differential theses. Although the central propositions of each theory differ, a common assumption of all approaches is that certain ‘kinds of places’ have an effect on bias crime. Four key community characteristics and processes associated with bias crime in the community are highlighted in these theories. These are the ethnic/racial composition of the area, levels of poverty, residential mobility, and community cohesion. Generalised Ecological Theories of Crime Social disorganisation theory and its systematic reformulations represent a generalised ecological theory of crime. The central theoretical proposition in these approaches is that particular structural features erode the community networks and the capacity (or indeed willingness) to engage in informal social control which then leads to an increase in crime (Bursik, 1988; Bursik and Grasmick, 1993, Sampson and Groves, 1989; Sampson and Raudenbush, 1997; Shaw and McKay, 1942). In this literature, three social structural characteristics are often associated with lower informal social control: racial heterogeneity, poverty and residential mobility. Social disorganisation theory argues that racial heterogeneity leads to higher levels of crime as cultural and language barriers impede the social networks necessary to prevent crime (Bursik, 1988; Elliot et al., 1996; Sampson and Groves, 1989; Smith and Jarjoura, 1988). Racially diverse communities struggle to find some degree of social solidarity in order to organise a collective response to community problems like bias crime (Kornhauser, 1978). Poverty is another key feature that is linked to crime (Kornhauser, 1978; Morenoff, Sampson and Raudenbush, 2001; Sampson and Groves, 1989; Smith and Jarjoura, 1988). Impoverished communities are unlikely to have the necessary resources to engage in informal social control or prevent crime through other means (Kornhauser, 1978; Sampson and Groves, 1989). Finally residential instability also influences crime as it “weakens voluntary organisations and thereby directly reduces both informal and 6

formal sources of social control” (Stark, 1987, p.900). Here it is argued that residential turnover prevents residents from developing friendships and ties to organisations (Bursik and Grasmick, 1993; Sampson and Groves, 1989; Smith and Jarjoura,1988; Stark, 1987). Collective efficacy is a recent extension of social disorganisation theory and shifts the focus from the impact of neighbourhood conditions on networks to a focus on social cohesion and perceived informal social control (Sampson et al., 1997). According to Sampson and colleagues (1997, p.717), collective efficacy can be defined as “social cohesion among neighbours combined with their willingness to intervene on behalf of the common good”. They suggest that collective efficacy is able to mediate the relationship between ecological conditions and crime. If a community is collectively efficacious, even if it is characterised by higher levels of racial/ethnic heterogeneity, poverty, or residential mobility, it will experience lower levels of crime compared to structurally similar communities with low levels of collective efficacy (Sampson et al., 1997). The association between structural characteristics, social processes like collective efficacy, and crime is demonstrated in the criminological literature (Browning, 2002; Browning, Dietz and Fienberg, 2004; Morenoff et al., 2001; Sampson et al., 2008; Sampson, Raudenbush and Earls, 1998; Sampson and Wikstrom, 2007). There is some evidence that this association may also hold true for bias crime. Lyons (2007) argues that generalised ecological theories of crime can explain bias crime, as both bias and non-bias crime reflect antisocial behaviour. To date, only two studies consider the efficacy of generalised ecological theories in explaining bias crime. In his study of 103 neighbourhoods in Sacramento, Grattet (2009) found that concentrated disadvantage and residential turnover significantly increased the prevalence of bias crime. In communities with high levels of concentrated disadvantage, bias crimes doubled; however, ethnic heterogeneity had no effect on bias crimes.


Lyons’ (2007) study provides a more comprehensive test of the utility of generalised ecological theories of crime in explaining bias crime. He examined the effect of economic status, residential mobility, racial composition, race-specific economic status and economic inequality on both anti-white and anti-black bias crimes in Chicago. Lyons (2007) found that anti-black crimes were positively associated with community affluence and negatively associated with indicators of disadvantage. In contrast, anti-white crimes were more likely in disadvantaged communities with high levels of residential instability. Further anti-black crime was found to be more likely to occur in relatively organised communities with high levels of informal social control whereas anti-white crime was found to be more likely to occur in communities with lower levels of informal social control. Thus while anti-white crime appears to share a similar causation with non-bias crime, anti-black crime does not. Specific Ecological Theories of Bias Crime Unlike the generalised ecological theories of crime, specific ecological theories of bias crime postulate that bias crimes are not the result of a breakdown of informal social control or social cohesion but instead occur when resources or identities are threatened or challenged. The power differential thesis, for example, states that bias crimes occur at a higher rate in areas with a larger proportion of white residents. In contrast to social disorganisation or collective efficacy theory, the power differential thesis claims that residential stability and racial homogeneity promote bias crimes. That is, the presence of minority groups in communities where residents desire racial homogeneity leads to bias crime. From this perspective, white community members perceive a low level of risk for retribution attacks or law enforcement involvement when minority numbers are few. Thus minority groups are at a greater risk of victimisation in areas where they make up a smaller proportion of the population (LeVine and Campbell, 1972).


Brimcombe and colleagues (2001) focused on the influence of stability in racial composition on bias crime in the London Borough of Newham. They found that in any ward where the white population was over 50 per cent, rates of victimisation increased for the minority groups. Further, minority group victimisation was highest in wards with the white population at 75 per cent or above (Brimcombe, Ralphs, Sampson and Tsui, 2001). More recently in the United States, Disha and colleagues (2011) examined the relationship between racial composition and bias crimes directed towards Arabs/Muslims. They found that Arabs and Muslims were at higher risk of victimisation after 9/11 in counties where their proportions were extremely small. This was particularly the case when the white residents comprised a very large proportion of the population. Disha and colleagues (2011) argued that in such areas, the small minority group is highly visible with little protection and is thus vulnerable for bias attacks. Another specific ecological theory of bias crime is the defended neighbourhood thesis. Originally developed by Suttles (1972), the defended neighbourhood thesis proposes that bias crime represents an extreme means for defending community values from the threat posed by minority groups. The defended neighbourhood thesis identifies two important factors that facilitate bias crime: changes in racial composition and community attachment. According to the defended neighbourhood thesis, higher rates of bias crime will occur in traditionally homogenous communities experiencing demographic change (Green et al., 1998). Demographic change operates as a catalyst for affirmative action amongst white residents who perceive the increasing presence of ethnic minorities to threaten their status, shared values and way of life (Green et al. 1998). As a result, some residents attack minority group members in a bid to preserve racial homogeneity. Using census data and police reports, Green and colleagues (1998) tracked official bias crimes and changes in racial composition across 51 New York City community districts. 9

Their findings suggest that bias crime is more likely when minorities increase in population size in areas that have long been predominantly white, providing evidence for the defended neighbourhood approach. Similarly, Grattet (2009) found predominantly white neighbourhoods which experienced a large growth in the non-white population had a much greater expected number of bias crimes compared to more diverse populations with low nonwhite migration. However both Green and colleagues (1998) and Grattet (2009) failed to empirically examine the second component of the defended neighbourhood thesis – community attachment. According to Lyons (2008), demographic changes are most threatening to homogenous areas with higher levels of attachment such as tight knit communities where residents know each other and are able to identify outsiders easily. Conversely, white homogenous communities without strong attachment are less likely to be compelled to commit a bias motivated crime. Without attachment or a collective community identity, residents may not perceive demographic change to be a threat to community values. Lyons (2008) found that communities with higher levels of attachment, measured by the density of networks and the respondent’s time at their current address, have more anti-black crime than those with lower levels of attachment. In support of the defended neighbourhood thesis, Lyons (2008) argues that bias crimes are more likely to occur in strongly attached communities undergoing demographic change. The Limitations of Current Bias Crime Research An ecological approach to the study of bias crime is in its infancy, with only a handful of studies that consider the neighbourhood’s socio-demographic characteristics and the neighbourhood processes, like community cohesion, that lead to bias crime. While some research suggests that the etiology of bias crime is similar to that of non-bias crime, other 10

studies indicate that racial heterogeneity, residential stability and community attachment differentially influence the occurrence of bias crimes. As Heitgard and Bursik (1987, p.786) argue, bias crimes “may not represent internal social disorganisation but organised responses to perceived external threats.” Ecological studies of bias crime are largely concentrated on the United States (Disha et al., 2011: Grattet, 2009; Green et al., 1998; Grattet, 2009; Lyons, 2007; Lyons, 2008). Yet the United States has a particular racial history and thus the findings here may not generalise to other countries like Australia. Structurally, there are also important differences between the United States and Australia. Australia is one of the most ethnically diverse populations in the world (ABS, 2008). In a population of 22 million, Australians speak more than 400 languages, identify with over 270 ancestries and practice a wide range of cultural and religious traditions (ABS, 2008). According to the 2006 census, immigrants made up almost a quarter of the Australian population with 16 per cent speaking a language other than English at home (ABS 2008). Unlike the United States, Australia does not have homogenous ethnic groupings or ‘ethnic ghettos’ and there is little evidence of immigration segregation due to race-based poverty in the Australia context (Jupp, McRobbie and York, 1990). Another limitation of the bias crime literature is the exclusive reliance on official crime statistics to measure bias crimes across communities (Brimcombe et al., 2001; Disha et al., 2011; Grattet, 2009; Green et al., 1998; Lyons, 2007; Lyons, 2008). The inherent limitations of official statistics prevent researchers from developing an accurate picture of the scope of bias crime in the community. While the challenges associated with official statistics are not limited to bias crime, both Australian and international researchers have found minority bias victimisations are less likely to be reported than general non-bias crimes (Ignaski, 2001; HREOC, 2004; Levin, 1999; Perry, 2001; Zaykowski, 2010). According to HREOC (2004), many victims fail to report bias crime due to fear of retributive attacks, a 11

distrust of authority, language and cultural barriers, a general unawareness of the appropriate reporting channels and an overall lack of confidence in the criminal justice system. The problem with official reporting of bias crime is so severe, that leading scholars in the field believe that official reports are completely unsatisfactory for any analytic modelling (Perry, 2009). Utilising official bias crime statistics for an Australian study is particularly problematic due to differing legislation across states in regards to what constitutes bias crime and a lack of detail included in police incidents reports. While Victoria and New South Wales police collect bias crime data, for an incident to be officially recorded as bias crime relies on law enforcement agencies to identify and record the bias motivation. Further, Queensland police are not required to indicate whether a crime is motivated by bias and thus official statistics provide very little insight into the nature and extent of bias crime especially across Brisbane communities. The Current Research Project This research paper examines bias crime in the Australian context and utilising resident perceptions of bias crime. In particular, we are interested in testing the utility of generalised ecological theories of crime such as social disorganisation and collective efficacy in explaining the spatial variation of resident perceptions of bias crime. There are three main research questions which underpin this study. 1. How do community characteristics such as ethnic composition, poverty and residential mobility influence perceptions of bias crime? 2. Are perceptions of bias crime more likely to occur in communities that have become more ethnically diverse over time? 3. How does social cohesion and trust and informal social control impact perceptions of bias crime?


Methods Australia Community Capacity Study (ACCS) The ACCS is a leading longitudinal study of communities that includes four waves of data collection in Brisbane (see Mazerolle et al., 2007; Wickes et al., 2010; Mazerolle et al., 2011), one wave of data collection in Melbourne (see Mazerolle et al., 2011), four in-depth case studies of Brisbane neighbourhoods (see Wickes, 2010; Sargeant et al., 2013) and an ethnic community sample of residents from Indian, Vietnamese and Arabic speaking backgrounds in both Brisbane and Melbourne (see Mazerolle et al., 2011). The ACCS is funded by several Australian Research Council Grants (R0700002, DP1093960, DP1094589, DP0771785). Commencing in 2005, the overarching aim of the ACCS program of research is to investigate how different community processes such as collective efficacy may explain the spatial variation in crime over time. The current study employs data collected in 2010 representing the third wave of the ACCS in the Brisbane Statistical Division (BSD) located in Queensland. The Brisbane ACCS sample comprises 148 randomly drawn state suburbs with a residential population ranging from 245 to 20,999 (total state suburbs in the BSD = 429 with a residential population ranging from 15 to 21,001). Many of these neighborhoods comprise residents from Indigenous and immigrant backgrounds. In Brisbane the ACCS survey has been collected across four waves in 2005, 2008, 2010 and 2012. The number of residents needed to achieve ecometric reliability was assessed using power analyses from Optimal Design Software for multi-level samples. The overall consent and completion rate for the Brisbane sample was 68.52 percent which is calculated as the number of interviews completed proportional to the number of in-scope contacts (for further information see Mazerolle et al., 2012). This paper uses data collected in Wave 3 in 2010. The participant sample for Brisbane employed in Wave 3 comprises 4404 respondents,


randomly selected (using random digit dialing). The final sample used for analyses includes 148 suburbs and 3895 respondents. Data for Wave 3 of the ACCS survey was collected from 25 August to 15 December 2010 by the Institute for Social Science Research at the University of Queensland. Trained interviewers used computer-assisted telephone interviewing to administer the survey which lasted approximately 24 minutes. The in-scope survey population comprised all people aged 18 years or over who were usually resident in private dwellings with land-line telephones in the selected neighborhoods. Variable Information and Analytic Approach Two sources of data were used in our analyses: the Wave 3 ACCS survey data and 2001 and 2006 Australian Bureau of Statistics (ABS) census data aggregated to the suburb level. The variables included in this study are theoretically and empirically central to ecological theories of bias crime (Brimcombe et al., 2001; Disha et al., 2011; Green et al., 1998; Grattet, 2009; Lyons, 2007; Lyons, 2008) as we have outlined in our literature review. Dependent Variable Resident perceptions of bias crime: Due to the significant limitations associated with official statistics, we measure perceptions of bias crime from residents living across the 148 suburbs. In the ACCS survey, residents were asked to indicate the extent to which people being attacked or harassed because of their skin colour, ethnic origin or religion is a problem in their community. Respondents replied with ‘1’ = no problem, ‘2’ = some problem or ‘3’ = big problem. In our analyses we collapsed these into two categories ‘0’ = bias crime not a problem (n=3321) and ‘1’ bias crime a problem (n=574). As we argued earlier, this approach to measuring bias crime across communities aligns with third party reporting of bias crime incidents (VEOHRC, 2010; 2012, Police Scotland, 2013; HM Government, 2012).


Individual Level Control Variables: A number of individual level control variables are included in the models, these include; age, female (male:0, female:1), Indigenous (Non ATSI:0, ATSI:1), annual household income (less than $20,000, $20,000 to $39,999, $40,000 to $59,999, $60,000 to $79,999, $80,000 to $99,999, $100,000 to $119,999, $120,000 to $149,999, $150,000 or more), employed (not employed 0: employed:1), renting (own:0, rent:1), married (not married:0, married:1), children (no children:0, has children:1), speaks language other than English (LOTE) at home (English Only:0, LOTE:1) and time at current address (less than 6 months, 6 months to less than 12 months, 12 months to less than 2 years, 2 years to less than 5 years, 5 years to less than 10 years, 10 years to less than 20 years, 20 years or more). To retain in the analysis the 19.1 percent of respondents who did not give their income, we included a dummy variable for missing income (not missing:0, missing:1). <> Suburb Level Independent Variables According to generalised ecological theories of crime, certain community conditions and processes influence bias crime. These include: residential instability, poverty, ethnic composition and collective efficacy. To investigate the effect of these community characteristics on resident perceptions of bias crime, indicators that reflect these community processes and structural features of suburbs are included in the models. Residential Turnover: Two measures of residential mobility at the suburb level were drawn from ABS 2006 census data: the percentage at a different address than five years ago and the percentage of renters.


Poverty: Two measures of poverty were drawn from the 2006 ABS census data. These were the percentage of the population unemployed and looking for work, and the weekly median household income measured in $00s. Ethnic Composition: To control for ethnic composition at a community level, the percentage Indigenous2 and the Blau Diversity Index (Language)3 also derived from 2006 census data are included in the models. To capture the amount of variation of language within each suburb, we use the Blau Diversity index: 1 - Σ𝑝𝑖2 where p is the proportion of the total group who are members of a given category i. A completely homogenous community would receive a score of 0 and the score for an entirely heterogeneous community would approach 1. The Blau Diversity Index (Language) was multiplied by 100 to help interpretation. Change in Ethnic Composition: To examine the effect of changes in ethnic composition at the community level, changes in the percentage Indigenous and the Blau Diversity Index (Language) derived from 2001 and 2006 census data are included in the model. Collective Efficacy Scale: The CE scale comprises nine ACCS survey items and was constructed to measure informal social control and social cohesion and trust across communities. The items used in the ACCS CE scale are listed in Appendix 1. These items were measured on a five point likert scale with 0= very unlikely and 5= very likely. The CE scale used in the analyses is reliable with a Cronbach’s alpha of α=.750. The mean of the nine


Due to Australia’s history of poor relations with the Indigenous community, it is important to control for Indigenous status at a community level. 3 Language diversity rather than country of birth is the chosen indicator of diversity due to the high rate of migration to Australia from English speaking nations. In the Australian context, language can create a sense of otherness and cause an individual to be perceived as an outsider (Anderson, 1991; Calhoun; 1992; Leigh, 2006; Wickes et al., 2013).


items was calculated for each respondent to create an individual level CE scale. This variable was then aggregated to the community level by calculating the mean of the social cohesion scale for all respondents in each suburb. Participants who failed to provide a response to two or more items were not included in the calculation of the suburb level CE scale. The suburb level CE scale was mean-centred, so that 0 represents the average suburb level collective efficacy. << Table Two Here>> Analytic Strategy To account for the nested nature of the data and the variation between suburbs, a multilevel model was used to investigate whether respondents in suburbs with certain structural conditions or processes perceived differences in bias crime in their community. We estimate the prediction equation for a random intercept multilevel logistic regression model with a level 1 variable 𝑥1𝑖𝑗 and a level 2 variable 𝑥2𝑗 . The model is expressed mathematically as: 𝜋𝑖𝑗

𝑙𝑜𝑔𝑖𝑡 (𝜋𝑖𝑗 ) = 𝑙𝑜𝑔 ( ) = 𝛽0 + 𝛽1 𝑥1𝑖𝑗 + 𝛽2 𝑥2𝑗 + 𝑢𝑗 1−𝜋 𝑖𝑗

The coefficient 𝛽1 is the effect on the log-odds of a 1-unit increase in 𝑥1 for individual 𝑖 in suburb 𝑗. The coefficient 𝛽2 is the contextual effect of the suburb level variable 𝑥2 for suburb 𝑗. The random effect for group is 𝑢𝑗 ~ N (0, 𝜎 2 0). Thus this multilevel logistic regression accounts for variation in perceptions of bias crime at the individual/household and community level. To aid in the interpretation of the relationships, we report coefficients using the odds ratio.


The model is built in three stages. The individual level control variables are entered into Model 1. Model 2 incorporates the community characteristics associated with bias crime including ethnic diversity/ethnic homogeneity, changes in ethnic composition, residential instability and poverty.4 Collective efficacy is introduced in Model 3. Prior to the inclusion of explanatory variables, we ran a variance components model (or null model) with no explanatory variables to calculate the variation in bias crime that is attributable to living in a particular suburb. The variance partition coefficient (VPC) represents the percentage of variance in the outcome variable that is explained by the suburb. The VPC is calculated as the ratio of the level 2 variance to the sum of the level 1 and level 2 variances. The level 1 variance for a standard logistic distribution is 3.29. The VPC is also calculated for each model to determine the amount of suburb-level variation that has been explained by the addition of explanatory variables. Results The results of the multilevel models are displayed in Table 3. The initial VPC calculated from the null model indicates that 30.14 percent of the variation in perceptions of bias crime is due to living in a particular suburb. <
> As illustrated in Table 3, Model 1 shows control variables such as age, gender and renting are significantly associated with resident perceptions of bias crime. When the model only includes individual level variables, older respondents are less likely to perceive bias crime to be a problem in their community (OR= 0.98, p<0.001), female respondents are more likely to perceive bias crime to be a problem in their community compared to males (OR= 4

To assess whether the effect of a change in ethnic composition differs at high and low levels of ethnic diversity, interaction terms were created by multiplying the Ethnic Composition variables by the corresponding Change in Ethnic Composition variables. The interaction terms were not significant and thus not included in the final model.


1.23, p<0.05), and renters are more likely than home owners to perceive bias crime to be a problem (OR=1.45, p<0.05). However the inclusion of these household variables explains very little of the suburb level variation in perceptions of bias crime. With the addition of structural suburb level variables, in Model 2, the VPC is reduced to 9.62 percent. Thus individual and community characteristics explain around 69.27 percent of the 30.14 percent of the variation in the outcome due to suburbs. In Model 2, two structural indicators are significantly associated with resident perceptions of bias crime: poverty (as measured by median weekly household income) and ethnic diversity (as measured by the Blau diversity index and percentage of the population Indigenous). Changes in ethnic composition are not significantly associated with resident perceptions of bias crime. The inclusion of the community variables in Model 2 reduced the magnitude of the coefficient and the significance of the association between resident perceptions of bias crime and renting at the individual/household level. Model 2 shows that respondents from communities with greater language diversity are more likely to perceive bias crime to be a problem in their community (OR=1.01, p<0.05). These findings indicate a significant relationship between resident perceptions of bias crime and language diversity in the community. The relationship between the presence of Indigenous people in the community and resident perceptions of bias crime is also significant (OR=1.18 p<0.05). The greater the indigenous presence in area, the more likely residents will perceive bias crime to be a problem. Suburb level poverty is also seen to be a key predictor of bias crime with residents from communities with a higher weekly median household income less likely to perceive bias crime to be a problem in their community (OR=0.85, p<0.01). Yet while social disorganisation theory suggests residential turnover increases crime, residential instability is not linked to greater resident perceptions of bias


crime in this analysis. There appears to be no relationship between resident perceptions of bias crime and changes in ethnic diversity. The suburb level collective efficacy scale is introduced in Model 3. The inclusion of this variable reduces the VPC to 7.06 percent. Model 3 shows a significant negative relationship between collective efficacy and resident perceptions of bias crime (OR=0.08, p<0.001). Predicted probabilities are calculated to aid in understanding the relationship between collective efficacy and residents perceptions of bias crime in their community. The typical person living in a typical suburb with an average level of collective efficacy has the predicted likelihood of perceiving bias crime to be a problem of 10.95 percent (Odds= 0.12). However, by decreasing the level of collective efficacy to -0.4 under the mean, the same person now has the predicted likelihood of perceiving bias crimes to be a problem of 25.31 percent (Odds=0.34). Thus this decrease in collective efficacy has increased the likelihood of perceiving bias crime to be a problem from 10.95 percent to 25.31 percent. Additionally, the introduction of collective efficacy into the model fully mediates the effect of Indigenous presence on resident perceptions of bias crime (OR=1.11, n.s.), mediates the significance of language diversity on resident perceptions of bias crime (OR=1.01, n.s.) and partially mediates the effect of median weekly household income on resident perceptions of bias crime (OR=0.88, p<0.05). These findings suggest that respondents living in ethnically diverse communities perceive bias crime to be a problem in their area, whereas respondents from areas with higher median household incomes do not perceive bias crime as a problem in their community. However once collective efficacy is introduced into the model, the effect of ethnic diversity is fully mediated and the effect of poverty is partially mediated on perceptions of bias crime. Living in a cohesive community where residents work together to solve local problems is an important predictor of lower perceptions of bias crime in Australian communities. 20

Discussion Ecological theories emphasise the importance of specific community characteristics and processes that create an environment conducive to bias crime. Drawing on the core tenets of ecological theories of crime, this paper examined the effect of residential mobility, poverty, stability and change in ethnic composition and community cohesion on resident perceptions of bias crime in their community. Our first question asked whether or not greater ethnic diversity, poverty and residential mobility influence resident’s perceptions of bias crime. The findings indicate that ethnic diversity and poverty are positively associated with resident perceptions of bias crime. Similar to previous research in the United States and United Kingdom (Brimcombe et al., 2001; Disha et al., 2001; Grattet, 2008; Green et al., 1998; Lyons, 2008), the ethnic composition of a community is a key indicator of perceived bias crime in the Australian context. In contrast to studies that find ethnic homogeneity is associated with more bias crime (Brimcombe et al., 2001; Disha et al., 2001; Grattet, 2008; Green et al., 1998; Lyons, 2008), in Australia it is ethnic diversity that predicts perceptions of bias crime. This suggests that language and cultural differences may prevent residents from reaching the degree of social solidarity necessary to organise a collective response to bias crime. We also find that respondents from disadvantaged communities perceived more bias crime. Our second research question asked if perceptions of bias crime are more likely to occur in communities that have become more ethnically diverse over time. Despite considerable support for this theory in the United States, the results reveal no support for this relationship in the Australian context (Grattet, 2009; Green et al., 1999; Lyons, 2008). Increases in ethnic heterogeneity between 2001 and 2006 did not increase the probability that residents would perceive bias crime as a problem in their community.


Finally in line with the defended neighbourhood thesis and collective efficacy theory we asked how community cohesion (as measured by collective efficacy) influences resident perceptions of bias crime. Our findings reveal that residents living in collectively efficacious communities are less likely to perceive bias crime as a problem. Further collective efficacy mediates the effect of ethnic diversity and poverty on resident perceptions of bias crime. This suggests that community cohesion and residents’ willingness to work together to resolve local problems may help to develop an environment that is more conducive to ethnic harmony. Our findings contrast with those found in the United States (Lyons 2007; Lyons, 2008). In Australia, living in cohesive communities where residents are perceived as able to work together to solve local problems is a protective factor against bias crime. In this context, generalised ecological theories of crime like social disorganisation theory and collective efficacy theory can help to explain the spatial concentration of bias crime. From our results, it would appear as though bias crime share a similar etiology with non-bias crime. One explanation for the contrasting findings presented here could be that there are different dependent variables employed across the various studies. In this paper, we utilise resident perceptions of bias crime in their community whereas previous research into the ecology of bias crime has almost exclusively utilised official bias crime statistics. Yet official crime statistics present misleading results as bias crime is underreported in police incident reports. Resident perceptions of bias crime overcome these measurement challenges as they do not rely upon the victim to report an incident to authorities or require an official to document the crime as a bias crime. Rather, resident perceptions rely on the accounts of residents who live in the community (Skogan, 2012). The various social, cultural and political differences between Australia and the United States may also account for the differences in research findings. As previously discussed, homogenous ethnic groupings are uncommon in Australian communities whereas in the 22

United States ethnic ghettoes are quite prevalent (Jupp et al., 1990), and African Americans (the most commonly targeted group for bias crime) make up around 13.1 percent of the population in the United States (FBI UCR, 2010; Lyons 2007). Alternatively in Australia, Indigenous Australians experience the highest levels of racism however only constitute 2.5 percent of the population (ABS, 2010; Dunn 2003). These distinct differences in the racial composition of urban communities in Australia in contrast to the United States could potentially lead to differential results concerning the effect of ethnic/racial diversity or homogeneity on bias crime. These findings offer some significant considerations for policymakers in tackling bias crime. We argue that local governments can reduce the occurrence of bias crime by improving levels of collective efficacy and addressing the structural conditions which affect residents. Collective efficacy can be developed in communities through programs designed to emphasise respect for local diversity, trust amongst residents and foster a sense of social responsibility. Yet there is no one size fits all approach to building community cohesion and we suggest local government bodies interact with residents to identify local issues and develop targeted programs which meet the unique needs of the community. Additionally, as collective efficacy does not fully mediate the relationship between poverty and perceptions of bias crime, economic strategies and public spending are important. Further, Sampson (1998) warns poverty can actually dampen a community’s collective efficacy. Thus to create collectively efficacious communities, alleviating the negative effects of poverty is also necessary (Wetherell, LaFleche and Berkely, 2007). Future research could further develop and expand upon our approach to understanding bias crime in the community by including follow up survey questions which investigate the socio-demographic characteristics of the victim and aggressor, where the crime took place, whether the aggressor acted alone or in a group, and how the victim was attacked. This would 23

help determine which minority groups are specifically targeted, highlight ethnic conflicts between various groups in communities and provide information on bias crime hotspots. Further, while the current study focused specifically on crime motivated by ethnic/racial bias, we suggest that using residents’ perceptions of bias crime may also be helpful in understanding the spatial variation of bias crime directed towards other minority groups such as the LGBT or disabled community. As it stands, little is known about the community contexts in which these types of bias attacks take place. Conclusion Bias crime has a particularly deleterious impact on society as it not only affects the individual but generates widespread fear amongst minority group members. While previous research has determined bias crime is a problem in Australia, no Australian study has explored whether or not these crimes are spatially concentrated and what community characteristics might help to explain their occurrence. By adopting an integrated theoretical approach, this research offers a valuable contribution to the current research literature by examining the ecology of bias crime in Australia. Our findings suggest community cohesion is an important protective factor in preventing bias crime across communities. Therefore by implementing community based strategies designed to strengthen community cohesion and by addressing the structural conditions of a community, local governments can effectively reduce the occurrence of ethnically or racially motivated bias crime.


References Anderson, B. (1991). Imagined communities: Reflections on the origin and spread of nationalism. New York: Verso. Australian Bureau of Statistics (2008). Cultural Diversity. Available at:[email protected]/Lookup/2071.0main+features902012-2013 (accessed 10 October 2012). Australian Bureau of Statistics (2010). Migration, Australia 2010-2011. Available at:[email protected]/0/84074889D69E738CCA257A5A00120A69?open document (accessed 20 October 2013) Barnes, A. and Ephross, P. (1994). The impact of hate violence on victims: Emotional and behavorial responses to attacks. Social Work 39(3): 247-251. Blau, P.M. (1977). Inequality and Heterogeneity: A Primitive Theory of Social Structure. New York: The Free Press. Boeckmann, R. and Turpin-Petrosino, C. (2002). Understanding the harm of hate crime. Journal of Social Issues 58(2): 207-225. Boyd, E., Berk, R. and Hamner, K. (1996). Motivated by hatred or prejudice: Categorization of hate-motivated crimes in two police divisions. Law & Society Review 30(4): 819-850. Brimcombe, A.J., Ralphs, M.P., Sampson, A. and Tsui, H.Y. (2001). Analysis of the role of neighbourhood ethnic composition in the geographical distribution of racially motivated incidents. British Journal of Criminology 41(2): 293-308. Browning, C.R. (2002). The span of collective efficacy: Extending social disorganization theory to partner violence. Journal of Marriage and Family 64(4): 833-850. Browning, C.R., Dietz, R.D. and Feinberg, S.L. (2004). The paradox of social organization: Networks, collective efficacy, and violent crime in urban neighborhoods. Social Forces 83(2): 503-534. Bursik, R.J. (1988). Social disorganization and theories of crime and delinquency: Problems and prospects. Criminology 26: 519–551. Bursik, R.J. and Grasmick, H.G. (1993). Neighborhoods and crime: The dimensions of effective community control. New York: Lexington. Calhoun, C. (1992). Nationalism and Ethnicity. Annual Review of Sociology: 211-239. Craig, K.M. (1999). Retaliation, fear or rage an investigation of African American and White reactions to racist hate crimes, Journal of Interpersonal Violence 14(2): 138-151. Disha, I., Cavendish, J. and King, R. (2011). Historical events and spaces of hate: crimes against Arabs and Muslims in post 9/11 America. Social Problems 58(1): 21-46. Dunn, K. (2003). Racism in Australia: findings of a survey on racist attitudes and experiences of racism, paper presented at the 2003 conference “the Challenges of Immigration and Integration in the European Union and Australia” University of Sydney 2003, National Europe Centre Paper No.77. 25

Elliott, D., Wilson, W.J., Huizinga, D., Sampson, R., Elliott, A., and Rankin, B. (1996). The effects of neighborhood disadvantage on adolescent development. Journal of Research in Crime and Delinquency 33:389–426. FBI Uniform Crime Reports (2010). Hate crimes. Available at: (accessed 12 October 2012). Finn, P. (1988). Bias crime: Difficult to define, Difficult to prosecute. Criminal Justice 3(3): 19-25. Garcia, L. and McDevitt, J. (1999). The psychological and behavioural effects of bias and nonbias motivated assault. Washington: U.S Department of Justice, National Institute of Justice. Grattet, R. (2009). The urban ecology of bias crime: a study of disorganized and defended neighbourhoods. Social Problems 56(1): 132-150. Green, D.P., McFalls, L. and Smith, J. (2001). Hate crime: An emergent research agenda. Annual Review of Sociology 27(1): 479-504. Green, D., Strolovitch, D. and Wong, J. (1998). Defended neighbourhoods, integration and racially motivated crime. American Journal of Sociology 104(2): 372- 403. Heitgard, .J.L. and Bursik, R.J. (1987). Extracommunity dynamics and the ecology of delinquency. American Journal of Sociology 92: 775-787. Herek, G., Gillis, R.J. and Cogan, J.C. (1999). Psychological sequelae of hate-crime victimization among lesbian, gay and bisexual adults. Journal of Consulting and Clinical Psychology 67(6): 945-951. HM Government (2012). Challenge it, report it, stop it: The Government’s plan to tackle hate crime. Available at: attachment_data/file/97849/action-plan.pdf (accessed 20 October 2013) Home Office (2012). Hate crimes, England and Wales 2011 to 2012. Available at: (accessed 5 May 2013) Human Rights and Equal Opportunity Commission (2004). IsmaU – Listen: national consultations on eliminating prejudice against Arab and Muslim Australians. Sydney: Human Rights and Equal Opportunity Commission. Ignaski, P. (2001) Hate crimes hurt more. American Behavioral Scientist 45(4): 627-638. Iganski, P. (2002). The hate debate: Should hate be punished as a crime? London: Profile Books. Ignaski, P. (2008). ‘Hate crime’ and the city. Great Britain: The Policy Press. Johnson, H. (2005). Experiences of crime in two selected migrant communities. Trends and Issues in crime and criminal justice, 302. Canberra: Australian Institute of Criminology. Jupp, J. McRobbie, A. and York, B. (1990). Metropolitan ghettoes and ethnic concentrations. Wollongong: Centre for Multicultural Studies. 26

Kornhauser, R. (1978). Social sources of delinquency. Chicago: University of Chicago Press. Lawrence, F.M. (1999). Punishing hate: Bias crimes under American law. Cambridge: Harvard University Press. Leigh, A. (2006). Trust Inequality and Ethnic Heterogeneity. The Economic Record, 82(258), 268-280. Levin, B. (1999). Hate crimes: Worse by definition. Journal of Contemporary Criminal Justice 15(1), 6 -21. LeVine, R.A. and Campbell, D.T. (1972). Ethnocentrism. New York: John Wiley. Lyons, C. (2007). Community disorganisation and racially motivated crime. American Journal Sociology 112 (3): 815-863. Lyons, C. (2008). Defending turf: Racial demographics and hate crime against blacks and whites. Social Forces 87(1): 357-358. Mason, G. (2009). Hate crime laws in Australia: are they achieving their goals? Criminal Law Journal 33(6): 326-340. Mazerolle, L., Wickes, R., Cherney, A., Murphy, K., Sargeant, E. and Zahnow, R. (2012). Community variations in crime: A spatial and ecometric analysis wave 3. ARC Centre of Excellence in Policing and Security: Brisbane. Mazerolle, L., Wickes, R., Rombouts, S., McBroom, J., Shyy, T-K., Riseley, K. Homel, R.J., Stewart, A., Spencer, N., Charker, J. & Sampson, R.J. (2007). Community Variations in Crime: A Spatial and Ecometric Analysis, Technical Report No.1. ARC Centre of Excellence in Policing and Security: Brisbane. Morenoff, J.D., Sampson, R.J. and Raudenbush, S.W. (2001). Neighbourhood inequality, collective efficacy and the spatial dynamics of urban violence, Criminology 39(3), 517-558. New South Wales Police Force (2011). NSW Police Force priorities for working in a culturally, linguistically and religiously diverse society. Available at: http://www.police.nsw. (accessed 20 October 2013). Perry, B. (2001). In the name of hate: Understanding hate crimes. New York: Routledge. Perry, B. (2005/2006). A crime by any other name: The semantics of ‘hate’. Journal of Hate Studies, 4: 121-137. Perry, B. (2009). Hate Crimes. Green Publishing Group. Police Scotland. (2013). Hate Crime and Third Party Reporting. Available at: (accessed 20 October 2013) Sampson, R. and Groves, B. (1989). Community structure and crime: Testing social disorganization theory. American Journal of Sociology 94: 774-802.


Sampson, R., Raudenbush, S. and Earls, F. (1997) Neighbourhoods and violent crime: A Multilevel Study of Collective Efficacy. Science 277: 918-24. Sampson, R. and Wikström, P-OH. (2008). The Social Order of Violence in Chicago and Stockholm Neighborhoods. A comparative inquiry. In: Shapiro I, Kalyvas S and Masoud T (eds) Order, conflict and violence. New York and Cambridge: Cambridge university press. Sargeant, E. Wickes, R. and Mazerolle, L. (2013). Policing community problems: Exploring the role of formal and social control in shaping collective efficacy. Australian & New Zealand Journal of Criminology 46(1): 70-87. Shaw, C.R. and McKay, H.D. (1942). Juvenile delinquency and urban areas: A study of rates of delinquents in relation to differential characteristics of local communities in American cities. Chicago: University of Chicago Press. Skogan, W.G. (2012). Disorder and Crime. In: Welsh, BC and Farrington DP Oxford Handbook of Crime Prevention. Oxford Univerity Press. Smith, D. and Jarjoura, R. (1988). Social structure and criminal victimization. Journal of Research in Crime and Delinquency 25:27-52. Southern Poverty Law Centre (2012). Active Hate Groups in the United States. Available at: (accessed 1 June 2013). Stark, R. (1987). Deviant places: A theory of the ecology of crime. Criminology 25:893-909. Suttles, G.D.(1972). The social construction of communities. Chicago: University of Chicago Press. Victorian Human Rights and Equal Opportunities Commission (2010). Submission to the review of identity-motivated hate crime. Available at: 589:submission-to-the-review-of-identity-motivated-hate-crime&Itemid=496 (accessed 12 April 2012) Victoria Police (2013). Prejudice motivated crime. Available at: (accessed 20 October 2013) Wetherell, M., LaFleche, M. and Berkely, R. (2007) Identity, Ethnic Diversity and Community Cohesion. Califorina: Sage. Wickes, R. (2010). Generating Action and Responding to Local Issues: Collective Efficacy in Context, Australian New Zealand Journal of Criminology, 48(3). Wickes, R. Homel, R. McBroom, J. Sargeant, E. and Zahnow, R. (2011) Community Variations in Crime: A Spatial and Ecometric Analysis Wave 2. ARC Centre of Excellence in Policing and Security: Brisbane. Wickes, R. Zahnow, R., White, G. and Mazerolle, L. (2013). Ethnic Diversity and its Impact on Community Social Cohesion and Neighbourly Exchange, Journal of Urban Affairs DOI: 10.1111/juaf.12015. 28

Zaykowski H (2010) Racial disparities in hate crime reporting. Violence 25(3): 378-94.

Appendix 1: ACCS Items –Collective Efficacy Scale.

Social Cohesion and Trust Items

Informal Social Control Items

1. People around here are willing to help their neighbours.

5. If a group of community children were skipping school and hanging around on a street corner, how likely is it that people in your community would do something about it? 6. If some children were spray painting graffiti on a local building, how likely is it that people in your community would do something about it? 7. If there was a fight in front of your house and someone was being beaten or threatened, how likely is it that people in your community would break it up? 8. If a child was showing disrespect to an adult, how likely is it that people in your community would scold that child? 9. Suppose that because of budget cuts the fire station closest to your home was going to be closed down. How likely is it that community residents would organise to try and do something to keep the fire station open?

2. This is a close-knit neighbourhood. 3. People in this neighbourhood can be trusted. 4. People in this neighbourhood do not share the same values.


Table One: Descriptive Statistics for Residents’ Perceptions of Bias Crimes and ACCS control variables (N=3895) Variables



Resident Perceptions of Bias Crime

No problem Problem

85.27% 14.73%


Male Female

39.61% 60.39%



Mean: 51.91

Time at Address

Less than 6 months 6 months to less than 12 months 12 months to less than 2 years 2 years to less than 5 years 5 years to less than 10 years 10 years to less than 20 years 20 years or more

0.74% 1.33% 2.93% 15.22% 25.10% 29.80% 24.87%


Own Rent

88.71% 11.29%

Marital Status

Not married Married

32.91% 67.09%


No children Has children

62.62% 37.38%


Employed Not employed

57.52% 42.48%

Annual Household Income

Missing Less than $20,000 $20,000 to $39,999 $40,000 to $59,999 $60,000 to $79,999 $80,000 to $99,999 $100,000 to $119,000 $120,000 to $149,999 $150,000 or more

19.10% 6.16% 12.53% 11.63% 11.47% 11.65% 9.21% 7.60% 10.65%

English only or LOTE

English Only LOTE

93.89% 6.11%

Indigenous Status


99.10% 0.90%


Table Two: Descriptive Statistics for Community Level Variables (N=148) Variables





Blau Diversity Index (Language)










% At Different Address Five Years Ago 42.31%




% Renting





Median Household Income (Weekly)

$1224.60 331.55



% Unemployed

















2006 ABS Census

2001 and 2006 Census % Absolute Change in Blau Diversity Index % Absolute Change in ATSI ACCS Data Collective efficacy scale (centred)


Table 3: Multilevel Models (N=148, n=3895) Model 1 Odds Ratio S.E

Model 2 Odds Ratio S.E

Individual level Age 0.98 *** 0.00 0.98 Female 1.23 * 0.13 1.23 Married 1.01 0.17 1.05 Has children 0.80 0.10 0.82 Time at Address 0.99 0.04 0.99 Renting 1.45 ** 0.22 1.30 Employed 0.98 0.11 0.94 Annual Income .98 0.03 1.00 Missing Income 0.80 0.15 0.89 LOTE 1.15 0.22 1.09 ATSI 1.80 0.79 1.74 Suburb Level % Different Address Five Years 1.01 Ago % Renting 1.00 % Unemployed 1.17 Median Household Income 0.85 (Weekly) % Indigenous 1.18 Blau Diversity Index 1.01 (Language) % Change Blau Diversity Index 1.04 % Change Indigenous 0.90 Collective Efficacy Scale Constant 0.28 * 0.12 0.33 Chi2 44.05 *** 182.55 LR test of improvement 226.90 *** 43.22 VPC 28.63% 9.62% Significance: # p<0.1, * p< 0.05, ** p<0.01, *** p<0.001

*** *

Model 3 Odds Ratio S.E

0.00 0.13 0.12 0.10 0.04 0.20 0.11 0.03 0.16 0.21 0.73

0.98 *** 1.25 * 1.05 0.82 0.98 1.32 0.96 1.01 0.92 1.11 1.84

0.00 0.13 0.12 0.10 0.04 0.20 0.11 0.03 0.17 0.21 0.77





0.01 0.15 0.00

0.99 0.99 0.88 *

0.01 0.12 0.04

* *

0.09 0.01

1.11 1.01

0.01 0.01

0.02 0.13

*** ***

1.02 0.89 0.08 *** 0.37 0.50 214.04 *** 23.92 *** 7.06%

0.02 0.13 0.04 0.40