Abstract

Fear of crime occupies a substantial area of research and theorizing in criminology. Yet, it has not been examined within a longitudinal framework of political socialization. Using insights from generational modelling, we explore how political cohorts influence the fear of crime and perceptions of antisocial behaviour. This ‘age, period and cohort’ (APC) approach recognizes the distinct temporal processes of (1) individual ageing, (2) current contexts and (3) generational membership and is crucial to understanding the origins and shape of social change. We employ repeated cross-sectional data from the British Crime Survey in an APC analysis to explore how worry about crime and perceptions of antisocial behaviour were impacted by the sociopolitical environment in which respondents spent their ‘formative years’. Our results underline the theoretical significance of political socialization and the methodological consequence of longitudinal analyses when exploring public perceptions of crime. We find that political socialization can have a distinctive and enduring impression on public perceptions of crime from childhood into middle age.

Introduction

For over 40 years, the fear of crime has been a stable of North American, British and European criminological research. Hundreds of publications have sought to illuminate the social and emotional risks associated with worry about crime (Ferraro 1995; Hale 1996; Visser et al. 2013). Undoubtedly, this work has increased our understanding of the subject. However, despite the breadth and depth of these studies, they tend to treat respondents as if they grew up during periods of similar political and economic character with comparable formative experiences. Herein we test whether the political period in which a cohort ‘came of age’1 exerted a meaningful effect on the way in which that same cohort perceived crime and antisocial behaviour over time. Hence, we explore if temporal forces do more than provide a ‘context’ in which to situate public anxiety about crime, questioning if political socialization is fundamental to the development of those emotional responses.

Age, period and cohort (APC) effects are powerful variables in the analysis of long-term social change (Mannheim 1928; Ryder 1965; O’Brien 2015). Individual ageing, historical contexts and generational membership are simultaneously related to the passage of time but have separate effects. To be precise, APC methods identify the following time dimensions:

  1. Age effects condense those developmental changes that take place over the life course (individual aging).

  2. Period effects occur due to cultural and economic changes that are unique to precise historical periods and induce similar changes in individuals of all ages.

  3. Cohort effects are the substance of social change and represent the effects of formative experiences (Ryder 1965). Cohort effects integrate the impact of early life conditions and exposure to these historical and social factors can affect an individual throughout the life course (Mannheim 1928).

A key question therefore is how much change measured in a group is due to the aging process of the individual; how much is due to the historical period in which the data were collected and how much of it was about the generation in which the participant grew up. By disentangling these three dimensions of experience, we can identify trends and illuminate our analysis of behaviour over time. Although this model was first articulated by Mannheim (1928), successive APC modelling has consolidated the idea that generations socialized in different periods can diverge markedly in their values and behaviours (Tilley 2002; Grasso 2014; Neundorf and Niemi 2014; Tilley and Evans 2014; Grasso et al. 2017).

Because emotional responses to crime and perceptions of antisocial behaviour have been demonstrated to express a range of objective and subjective sociopolitical and cultural cues (Ferraro 1995; Harcourt 2001; Sampson and Raudenbush 2004; Farrall et al. 2009; Innes 2014; Zhao et al. 2015), we test if modelling political generations might add some novel insights to this field. If people merge crime and antisocial behaviour with a general apprehension about the direction a society is perceived to be moving towards and the stability of their neighbourhood, it is conceivable that the political context in which they grew up, and the conditions and debates they were exposed to, may have a lasting influence on their perceptions of crime. For example, did growing up in the United States during the era of civil unrest in the 1960s provoke concerns about crime and lawlessness that remained with that generation into the 1970s and beyond? In the United Kingdom, did growing up during Margaret Thatcher’s political ascendancy—a period in which crime rates soared2 and a ‘law and order’ agenda intensified (Hay 1996; Garland 2001; Loader 2008)—differentiate this generation’s fear of crime from previous and subsequent generations? Did a harsh political emphasis on antisocial behaviour and penal populism in the late 20th century (Burney 2005) lead those who grew up under New Labour to view concerns about neighbourhood incivility differently from previous generations? How do generations who came of age in other political periods compare?

We believe unravelling APC effects may have the potential to improve theorizing on the fear of crime. For example, research suggests that age is related to risk perception and fear of crime (Pain 1991; Tulloch 2000), but in the absence of a complete analysis including period and cohort effects, we cannot be sure that the observed differences are related to individual aging, rather than other temporal processes. Similarly, researchers have found that perceptions of neighbourhood incivility vary with age (Kruger et al. 2007). However, this may also vary across historical periods and cohorts. In this article, we explore the heterogeneity in life-course patterns, time trends and birth cohorts in assessments of crime fears and perceptions of antisocial behaviour. We ask, what are the net APC effects on perceptions of crime and antisocial behaviour? Do relationships among the variables of interest remain static as broader political and economic conditions change? To what extent do age patterns and time trends previously found in the research actually reflect birth cohort differences once the full APC models are utilized? This study aimed to address some of these questions by using time-series data from the British Crime Survey (BCS) spanning 30 years in England and Wales (Jennings et al. 2015). To test for the existence of generational or cohort effects, we integrate APC modelling using the method presented by Grasso (2014). The result is an empirical model that accounts for nuanced time effects but is also theoretically driven to acknowledge the meaning and significance of public perceptions about crime in a political temporal framework (Pierson 2004).

Next, we consider the relevant literatures on political generations, fear of crime and the particular role that antisocial behaviour has filled in the United Kingdom. We reflect on the methodological issues and challenges of generational modelling—including the matter of ‘identification’ in APC models.

Incorporating a political generational lens

The APC analyses herein are premised on the theory of ‘political generations’ (Tilley 2002). Political generations signify the importance of political events and experiences that occur during the impressionable ‘formative years’ and differentiate cohorts from each other (Mannheim 1928). The concept is borne out of Mannheim’s theory of generations, which encourages us to examine groups of people who share an age location in history. Notably, when we employ the term ‘political generation’ we are not referring merely to the political allegiance of those holding power at the time, but to the wider dimensions of that political era, including those which shape its social and economic policies and its civil characteristics.

Discussing the concept of generations, Mannheim argued that ‘youth experiencing the same concrete historical problems may be said to be part of the same actual generation’ (1928: 232). These ‘historical problems’ are understood to bare exclusive influences on the population throughout the life-course (Mannheim 1928; Ryder 1965). The subsequent cohort or generational effects have been expressed by Markus as the ‘enduring inter-cohort distinctions that are attributable to the common “imprinting” of cohort members…. generational effects are often presumed to be the result of cohort members having shared similar socialising experiences, especially during late adolescence and early adulthood’ (1983: 718). In a similar vein, life-course criminology has highlighted the links between macro-level social history and social structures, and the developing lives of individuals and communities (Sampson and Laub 1993; Ferraro and Shippee 2009).

Contemporary APC approaches also draw heavily on Ryder’s (1965) seminal article on the significance of the ‘cohort’ as a mechanism of social change. Ryder stressed that cohort changes and replacements were central to the transformation of society, describing a cohort as ‘an aggregate of individuals’ who had ‘a distinctive composition and character reflecting the circumstances of its unique origination and history’ (1965:845). A number of studies have supported this idea and confirmed that political contexts can produce generations with distinct value-sets and patterns of behaviour (Tilley 2002; Frost 2010; Bartels and Jackman 2014; Grasso 2014; Neundorf and Niemi 2014; Smets and Neundorf 2014; Tilley and Evans 2014; Grasso 2016).

The method employed here also allows for members of a given political generation to be divided by social and individual cleavages such as gender and education. Nevertheless, generational theorists believe a cohort will share certain values because their formative years were spent in the same temporal/spatial locations. Mannheim (1928) likened generations to social classes arising from distinct positions in the economic sphere. Cohort and generational theories thus imply non-linear shifts in public opinion. This framework assumes that a singular political generation might express attitudes and beliefs that are different from the generations coming of age before or after them (Grasso 2014; 2016). Accordingly, the prominent shift towards New Right politics in the United States, Britain, Australia and other countries (Stiglitz 2002; Steadman-Jones 2012; Farrall and Hay 2014), which developed during the Reagan (1981–88) and Thatcher and Major Governments (1979–97) may influence an individual’s perception and relationship to crime. This then is the first study (we are aware of) to conduct an APC analysis on perceptions of crime and antisocial behaviour. Our approach allows us to examine how attitudes operate over an extended historical period and consider the following questions: what are the net APC effects on fear of crime and concerns about antisocial behaviour? Do the findings for crime fears remain robust when one takes into account all three types of APC effects?

The magnification of fear of crime and perceptions of antisocial behaviour

The body of work on the fear of crime is vast and international and straddles a number of social science disciplines (Hale 1996). It became a highly topical and policy-relevant area of research during the 1960s in the United States and grew in influence elsewhere from 1970s to 1980s. Data revealed that fear of crime was a substantial social problem (Skogan and Maxfield 1981; Warr 1990; Hale 1996) and suggested that relatively large proportions of citizens in the United Kingdom, the United States, Australia and Europe reported being worried about crime (Hough 1995; Hale 1996; Kury and Theodore 1998).3 Crime fears have also been linked to various negative outcomes such as the erosion of peoples’ quality of life, the restrictions on movement, the adoption of costly precautions and the encouragement of ‘flight’ from deprived areas. Crime fears are said to absorb and reflect racial tensions and, ultimately, harm social trust, inter-group relations and the capacity of communities to exercise social control (Skogan and Maxfield 1981; Warr 1990; Ferraro 1995; Hale 1996; Sampson and Raudenbush 2004). Notably, the fear of crime continues to be both a barometer of the operation of the criminal justice system and a central topic in contemporary criminological studies and theorizing.

In a comparable fashion, debates around antisocial behaviour and disorder have reached across various academic fields, although, most conspicuously in relation to criminology (Goffman 1971; Wilson and Kelling 1982; Warr 1990; Harcourt 2001; Innes 2014).4 One of the first systematic investigations into how urban disorder influenced citizen’s perceptions of crime and the environment emanated from the Chicago School (Shaw and McKay 1942). This team reported that urban neighbourhoods with high levels of poverty often experienced breakdown in their social structures and institutions. Their work has been extensively developed and characteristics such as the local environment, population turnover and heterogeneity are also said to influence perceptions of crime and disorder through the mediating influence of social bonds and trust (Bursik and Grasmick 1993; Sampson et al. 1997; Rose and Clear 1998).

The literature discussed earlier illuminates how both crime and antisocial behaviour have a ‘long reach’ beyond the impact of an individual incident, to profound outcomes for one’s neighbourhood and wider society. The research also indicates that these concerns play a unique role in our social and political arrangements. For example, in their seminal work Policing the Crisis: Mugging, the State and Law,Hall et al. (1978) examined what they identified as a ‘moral panic’ over young black male robbers in the period 1972–73 in the Great Britain. A central part of their thesis was that robbery or more colloquially ‘mugging’ had become a vivid metaphor for a breakdown of order, of declining stability and living standards in post-War British society. Emotive discourses by politicians, policy makers and journalists were said to have helped shape the tone of debates on ‘mugging’ and simultaneously redirected public insecurities onto a marginalized and disaffected black youth population. For similar studies in the United States see Gilmore (2007) and Hagan (2010). Other scholars have noted that since the late 1970s, crime and disorder became a prominent feature of the political landscape (Garland 2001; Loader 2008). Hay (1996) explored specifically how the Thatcher governments recast the industrial relations and economic crisis of the late 1970s as a wider crisis of law and order.Garland (2001) remarked how the urban riots of the early 1980s in London, Liverpool, Birmingham and Leeds transformed crime into a major political issue, and linked it to questions of race, class and education. Images of excluded and disaffected young males of the inner-city became resonant as the perception grew of them ‘as a newly dangerous, alien class’ (Garland 2001: 154). Later in the 1990s, as crime rates peaked, policymakers became increasingly sensitive to matters relating to crime and criminal justice. In Britain, Home Secretary Michael Howard and the Labour Shadow Home Secretary Tony Blair (later, Labour Party Leader) both sought to accentuate the punitive nature of their criminal justice policies to reflect perceived popular concern. Even after crime started to fall in the mid-1990s in England and Wales, the New Labour government who came into power in 1997, insisted that ‘antisocial behaviour’ constituted one of the most important problems facing British society and made it a major policy priority (Burney 2005, Jennings et al. 2018). Lord Williams stated that ‘Antisocial behaviour is a menace on our streets; it is a threat to our communities’ (Hansard 1998). A raft of new criminal and civil interventions were heavily promoted by government, such as the Antisocial Behaviour Order (ASBO), dispersal powers and parenting orders (Crawford 2009).5 Notably, Beckett and Western (2000) have identified the inspiration New Labour took from ‘third-way’ domestic security strategies in America during the 1990s. Tony Blair sought to replicate Bill Clinton’s ‘tough’ criminal justice policies and reclaim debates around crime and antisocial behaviour (Driver and Martell 2002). What is clear is that in England and Wales—as in the United States and other Western countries—from the late 1980s crime moved centre stage as ‘a central organising principle of political authority and social relations’ (Loader 2008: 399).

The significance of crime was not only mounting in terms of its reputation and symbolism at the end of the 20th century. Crime rates were also accelerating consistently throughout the West. Official data on property crime in England and Wales reveal a steady rate of growth for all domestic burglaries from 1960 (150,622 recorded offences) up to the mid-1980s, after which there was a sudden spike that lasted up until 1993 (1,369,584 recorded crimes) and a decrease thereafter (Office of National Statistics 2012). Meanwhile self-report data (the BCS) found the residential burglary rate per household more than doubled between 1981 and 1993 (from 41 to 86 per 1,000 households), then decreased by one-third up to 1999 (Farrington and Jolliffe 2004). ‘All crime’ followed a similar pattern, and while measuring long-term crime trends is beset with technical pitfalls (UK Statistics Authority 2014), most commentators agree that crime increased dramatically around the 1980s, peaking between 1992 and 1995 and declined thereafter (Newburn 2007).

Notably, recognizing the importance of crime rates in understanding public sentiments about crime and antisocial behaviour, researchers in America and the United Kingdom have identified feedback processes in democratic politics—between crime rates, public opinion and public policy that account for the growth of penal populism (Enns 2016; Jennings et al. 2016). These scholars have demonstrated that the public unmistakably recognize and respond in a ‘thermostatic’ fashion to rising (and falling) levels of crime, and that in turn public support for being ‘tough on crime’ is translated into patterns of imprisonment (Jennings et al. 2016). Similarly, research has demonstrated that public sentiments about crime are responsive to the intersection of both objective and subjective factors (Baumer et al. 2003; Frost 2010).

Integrating APC Effects

It is perhaps unsurprising that APC work has not become more popular, because it demands longitudinal data spanning many years. Moreover, the fundamental disentangling of APC effects on social and attitudinal change means that researchers must devise strategies to deal with the inherent ‘identification problem’ (Neundorf and Niemi 2014). The ‘identification problem’ occurs because the three effects cannot be estimated simultaneously due to the linear dependency among the APC variables. If two of the values are known (i.e. a respondent’s age and year of interview), the third value in the relationship (the birth cohort) is automatically known. To deal with this methodological obstacle and allow for the parallel estimation of all three effects, we must apply theoretical and analytic approaches that allow us to ‘interrupt’ this association.

Thus, to test whether one’s political socialization has a unique role over and above other temporal effects, it is necessary to apply constraints to the APC statistical model. We do this via a three-step method following the technique employed in Grasso (2014) and tested further in Grasso (2016) and Grasso et al. (2017). First, we categorize the age groups6 and generational cohorts (see Table 1), while the measure for ‘period’ (year of interview) remains a continuous variable. We then run multivariate logistic regression analyses, which allows us to model social and individual characteristics alongside the political generations. Second, we apply generalized additive models (GAMs) to plot the identified, smoothed cohort effect and test for intergenerational differences. GAMs are useful as they mitigate the loss of information caused by categorization (Grasso 2014; Jiang and Carriere 2014) (for alternatives see Yang and Land 2013). Because our study concentrates on data from a single political area (England and Wales), we do not need to apply hierarchical generalized additive mixed models. Third, to test for cohort differences, we also apply post-estimation Wald tests.

Table 1

Political generations

Pre- consensus generationPost-war consensus generationWilson/Callaghan generationa,bThatcher/Major generationNew Labour generationc
Formative period1930–44 (14 years)1945–64 (18 years)1965–78 (13 years)1979–96 (18 years)1997–2010 (13 years)
Years of birth1910–241925–441945–581959–761977–90
Total N (%)32,934 (7%)118,261 (26%)104,176 (24%)140,569 (32%)48,878 (11%)
Pre- consensus generationPost-war consensus generationWilson/Callaghan generationa,bThatcher/Major generationNew Labour generationc
Formative period1930–44 (14 years)1945–64 (18 years)1965–78 (13 years)1979–96 (18 years)1997–2010 (13 years)
Years of birth1910–241925–441945–581959–761977–90
Total N (%)32,934 (7%)118,261 (26%)104,176 (24%)140,569 (32%)48,878 (11%)

aThis period includes the Conservative Heath Government of 1970–74.

bThis period begins in 1965 given the general election took place, untypically in the autumn of 1964.

cThis period includes Blair and Brown in government.

Table 1

Political generations

Pre- consensus generationPost-war consensus generationWilson/Callaghan generationa,bThatcher/Major generationNew Labour generationc
Formative period1930–44 (14 years)1945–64 (18 years)1965–78 (13 years)1979–96 (18 years)1997–2010 (13 years)
Years of birth1910–241925–441945–581959–761977–90
Total N (%)32,934 (7%)118,261 (26%)104,176 (24%)140,569 (32%)48,878 (11%)
Pre- consensus generationPost-war consensus generationWilson/Callaghan generationa,bThatcher/Major generationNew Labour generationc
Formative period1930–44 (14 years)1945–64 (18 years)1965–78 (13 years)1979–96 (18 years)1997–2010 (13 years)
Years of birth1910–241925–441945–581959–761977–90
Total N (%)32,934 (7%)118,261 (26%)104,176 (24%)140,569 (32%)48,878 (11%)

aThis period includes the Conservative Heath Government of 1970–74.

bThis period begins in 1965 given the general election took place, untypically in the autumn of 1964.

cThis period includes Blair and Brown in government.

Various smoothing techniques can be employed with the GAMs; in this instance, smoothing splines using a generalized cross-validation method are used to plot the non-parametric smoothed curve for the effect of year of birth. The value of the GAM approach is that it provides a visual comparison to the APC results. Moreover, Beck and Jackman have argued that because ‘GAMs strike a sensible balance between the flexibility of nonparametric techniques and the ease of interpretation and familiarity of linear regression’ (1998: 596), they should be utilized more widely in the social sciences. In sum, while the APC models categorize generations, the GAMs estimate the cohort effect as a smoothly changing function of year of birth. Thus, if the results from the two types of modelling match up to each other, we can have greater confidence that the effects are correctly estimated and not simply the result of potentially arbitrary constraints applied in models (Yang and Land 2013; Bell and Jones 2014).

Categorizing generations

For the purposes of this study, we have assigned individuals to a political generation based on the period in which they spent the majority of their formative years (age 15–25) and have conceived political generations based on key shifts in political direction from 1930 to 2010 (see Table 1). Specifically, we distinguish between the ‘pre’- and ‘post-consensus’ generation, given the end of World War Two marked a transformation in social and political arrangements in the United Kingdom with the wide extension and consolidation of the welfare state (Addison 1994; Paterson 2008). As Butler and Kavanagh (1997) note, the popular vote was quite evenly split between the two major parties (Labour and the Conservatives) in this era, emphasizing what is known as the ‘political consensus’. From the mid-1960s, this political consensus began to decay (Paterson 2008), hence our third generation ‘the Wilson/Callaghan generation’. At this time, alternative political parties began to gain support, such as the Liberals and nationalist parties in Scotland and Wales, as the Labour party lost credibility following a rise in Trade Union militancy (López 2014). There was also a palpable change in attitudes towards the welfare state and a growing dissatisfaction with Keynesian economics. Such was the loss of credibility in the socio-economic and political arrangements that 18 consecutive years of conservative rule followed. The Conservative government from 1979 to 1997 (our fourth generation, the Thatcher/Major generation) embarked on a project to ‘roll back’ the state and give citizens greater choice, as well as greater responsibility to organize their schooling, medical care and accommodation, while reducing benefits for the poor and vulnerable in society (Paterson 2008). The government also promoted a service-based industry in place of manufacturing, which prompted a steep rise in unemployment and income inequality (Hay 1996). Moreover, these Conservative administrations marked a key change in how crime and criminal justice were framed politically, with a greater emphasis on ‘law and order’ (Hay 1996; Farrall et al. 2009). Notably, in her final election broadcast on the eve of the 1979 election, Margaret Thatcher referred to the importance for citizens of ‘feeling safe in the streets’ (Riddell 1985: 193). Two decades of New Right dominance eventually led to the rise of ‘New Labour’ who won a convincing majority in 1997. The leader, Tony Blair, presided over a newly reformed party that endorsed market economics and was influenced by Antony Giddens’ (1998) ‘third way’, which sought to synthesize capitalism and socialism. Although New Labour was perceived, in some respects, as a continuation of Thatcherism, the governing party introduced the minimum wage, reduced inequality and child poverty, as well as devolving power to regional governments in Scotland and Wales (Paterson 2008). The period of New Labour is our final political generation.

We have formulated these generations in terms of both which political party was occupying the seat of administrative power and the broader political ideology and culture of the time. Of course, during the total period we are investigating (1930–2010), there were other important social and cultural shifts. The most pertinent of which, in relation to this study, is the nature of long-term crime rates (see Figure 1). Although sociologists and criminologists have long been wary of the inferences that can be drawn from officially recorded crime rates (UK Statistics Authority 2014), there is no doubt that crime rose dramatically over the lengthy period being examined herein (Newburn 2007). The context of rising crime will have impacted the public’s perceptions of crime and disorder (Farrall et al. 2009), and to that end, we have included a measure of officially recorded crime taken at the time of interview, in combination with experience of victimization within the last 12 months in the model (see Table 2). Nonetheless, we hypothesize that individuals’ attitudes will not only be shaped by the statistical rise in crime, but also by the (1) types of crime that ebbed and flowed and (2) the manner in which crime was framed politically (Ferraro 1995; Hay 1996; Garland 2001; Sampson and Raudenbush 2004; Loader and Sparks 2016).7 As such, the steep rise in property crime during the Thatcher era, coupled as it was with popular debates around ‘law and order’ (Farrall et al. 2009), might heighten the sensitivity of this generation to express more anxiety around burglary than generations who came before or after. Meanwhile, those who spent their formative years under New Labour might be influenced by the heavy political emphasis on antisocial behaviour that became a conspicuous criminal justice policy directive. To support our understanding, we explored the nature of UK parliamentary debates from 1910 to 2005 using Hansard. This data collate the Official Report of debates in Parliament and describe the number of occasions a keyword or phrase was used. We tracked the number of occasions Parliament discussed six key offences that are part of our analytic model (see Figure 2).8 Specifically, it suggests that from 1920 to the mid-1990s, there was little discussion of the notion of ‘antisocial behaviour’. This is in keeping with our expectations since it emerged during the latter period of Major’s (1990–97) premiership as New Labour sought to strengthen its reputation on criminal justice by emphasizing (and later legislating against) ‘antisocial behaviour’. However, parliamentary debates in relation to ‘vandalism’ began to grow from the 1950s before decreasing in the 1980s. With regards to ‘mugging’ and ‘burglary’, the data indicate a slightly different pattern—one that began to rise simultaneously in the 1960s and 1970s, before peaking in the 1980s for ‘mugging’, and in the 1990s for ‘burglary’. Again, this is to be expected, given the aforementioned ‘moral panic’ around mugging in the late 1970s and the sharp rise in domestic burglary in the 1990s (Morgan 2014). Finally, government deliberations on ‘car theft’ were less common than all other crime types; references to it did not increase until later in the time period, peaking in the 1990s before quickly decreasing.

Total police recorded crime rate per 1,000 head: England and Wales 1950–2012. Data source: Office of National Statistics 2012
Fig. 1

Total police recorded crime rate per 1,000 head: England and Wales 1950–2012. Data source: Office of National Statistics 2012

Number of debates in which keywords were discussed in Parliament (Hansard) per decade 1920–2005. The last time-period is limited to January 2000 to March 2005, when the data cease.
Fig. 2

Number of debates in which keywords were discussed in Parliament (Hansard) per decade 1920–2005. The last time-period is limited to January 2000 to March 2005, when the data cease.

Table 2

Independent variables

Independent variables includedFormat
Total number of (self-reported) victimizations in previous yearNumeric
Recorded crime rate per 1,000 head of the population for England and WalesaNumeric
Year of interviewNumeric
SexBinary
Age groups (16–34; 35–59; 60+)3 categories
Political generations (see Table 1)5 categories
Ethnicity (white; black; Indian/Pakistani; mixed/other)4 categories
Marital status (married or cohabiting; single; widow; divorced; separated)4 categories
Education status (higher education; A/AS levels/B Tech; O levels/CSE/GCSE; other/no qualifications)4 categories
Employment status (employed full-time; employed part-time; unemployed; retired/education/homemaker/other)4 categories
Income status (low 25%; mid 25–75%; top 25%)3 categories
Tenure (owner/mortgage; renters; other)3 categories
Inner-city residentBinary
Independent variables includedFormat
Total number of (self-reported) victimizations in previous yearNumeric
Recorded crime rate per 1,000 head of the population for England and WalesaNumeric
Year of interviewNumeric
SexBinary
Age groups (16–34; 35–59; 60+)3 categories
Political generations (see Table 1)5 categories
Ethnicity (white; black; Indian/Pakistani; mixed/other)4 categories
Marital status (married or cohabiting; single; widow; divorced; separated)4 categories
Education status (higher education; A/AS levels/B Tech; O levels/CSE/GCSE; other/no qualifications)4 categories
Employment status (employed full-time; employed part-time; unemployed; retired/education/homemaker/other)4 categories
Income status (low 25%; mid 25–75%; top 25%)3 categories
Tenure (owner/mortgage; renters; other)3 categories
Inner-city residentBinary

aThese data are adjusted for the 1998/99 rule change (Home Office 2012; 2014).

Table 2

Independent variables

Independent variables includedFormat
Total number of (self-reported) victimizations in previous yearNumeric
Recorded crime rate per 1,000 head of the population for England and WalesaNumeric
Year of interviewNumeric
SexBinary
Age groups (16–34; 35–59; 60+)3 categories
Political generations (see Table 1)5 categories
Ethnicity (white; black; Indian/Pakistani; mixed/other)4 categories
Marital status (married or cohabiting; single; widow; divorced; separated)4 categories
Education status (higher education; A/AS levels/B Tech; O levels/CSE/GCSE; other/no qualifications)4 categories
Employment status (employed full-time; employed part-time; unemployed; retired/education/homemaker/other)4 categories
Income status (low 25%; mid 25–75%; top 25%)3 categories
Tenure (owner/mortgage; renters; other)3 categories
Inner-city residentBinary
Independent variables includedFormat
Total number of (self-reported) victimizations in previous yearNumeric
Recorded crime rate per 1,000 head of the population for England and WalesaNumeric
Year of interviewNumeric
SexBinary
Age groups (16–34; 35–59; 60+)3 categories
Political generations (see Table 1)5 categories
Ethnicity (white; black; Indian/Pakistani; mixed/other)4 categories
Marital status (married or cohabiting; single; widow; divorced; separated)4 categories
Education status (higher education; A/AS levels/B Tech; O levels/CSE/GCSE; other/no qualifications)4 categories
Employment status (employed full-time; employed part-time; unemployed; retired/education/homemaker/other)4 categories
Income status (low 25%; mid 25–75%; top 25%)3 categories
Tenure (owner/mortgage; renters; other)3 categories
Inner-city residentBinary

aThese data are adjusted for the 1998/99 rule change (Home Office 2012; 2014).

Once a categorization of cohorts is devised, it is crucial to establish how individuals are assigned to a generation. Because we are interested in political socialization, year of birth was used to allocate individuals to the five different political generations based on the historical phase in which they have spent the majority of their formative years (Table 1) (Grasso 2014). Given that we are examining the formation of attitudes in relation to the dominant political landscape, we expect that political socialization will occur during the mid-teens to the mid-20s (Grasso 2014). The assumption here is also that each political period was sufficiently different as to warrant the distinction.9 This means we define ‘Thatcher’s Children’ as those born between 1959 and 1976 and coming of age in the protracted period of Conservative rule between 1979 and 1997. Likewise, we define the New Labour generation as those born during 1977 and 1990 whose formative years occurred during the administration led by Tony Blair and later Gordon Brown between 1997 and 2010.

Data

The data we use come from a merged dataset of the BCS for the period between 1982 and 2010 (Jennings et al. 2015). The BCS is a repeated cross-sectional survey in which respondents are only ever interviewed once, but in which attitudinal and other questions are repeated over time.10 The collation of this material resulted in a dataset of over 440,000 individual respondents. It thus provides rich individual-level data on attitudes about crime as well as necessary control variables (age, gender, victimization) over a sufficiently long time to conduct an APC analysis. To provide a measure of official crime rates at the time of interview, we have also included one variable from the annual number of crimes for England and Wales (Home Office 2012; 2014).

The political generations (Table 1) form one of 13 independent variables tested, alongside a range of individual and socio-demographic markers and measures of recent victimization in the respective recorded crime rate at the time of data collection (Table 2). Notably, BCS data do not provide a consistent geographic marker (to protect the anonymity of participants); however, we have employed an indicator for ‘inner-city’ that is based on a population weight derived from ward-level information.11Table 3 lists the dependent variables; these questions follow standard wording on fear of crime and perceptions of disorder in the United Kingdom appearing in national and local government and police surveys. The questions include a range of fears relating to property and violent crime as well as concerns about popular manifestations of antisocial behaviour (Burney 2005) that are available on a longitudinal basis in the BCS (some measures of ‘worry’ in the survey only appear for a limited period).12 The scaled measures of fear of crime and antisocial behaviour were recoded into dichotomous variables.13

Table 3

Dependent variables

Question set 1: Fear of Crime. ‘Could you tell me how worried you are about...?’Original response categoriesRecode for analysis
1) Being mugged/robbed (1982–2010)1) very worried 2) fairly worried 3) not very worried 4) not at all worried1/2 = 1 worried
2) Being burgled (1982–2010)3/4 = 0 not worried
3) Having your car stolen (1984–2010)
Question set 2: Anti-social behaviour. ‘How common a problem in this area are...?’Original response categoriesRecode for analysis
4) Vandalism or deliberate damage to property? (1982–2010)1) very big problem 2) fairly big problem 3) not a very big problem 4) not a problem1/2 = 1 a problem
5) Groups of teenagers hanging around? (1982–2010)3/4 = 0 not a problem
6) Noisy neighbours (1984–2010)
Question set 1: Fear of Crime. ‘Could you tell me how worried you are about...?’Original response categoriesRecode for analysis
1) Being mugged/robbed (1982–2010)1) very worried 2) fairly worried 3) not very worried 4) not at all worried1/2 = 1 worried
2) Being burgled (1982–2010)3/4 = 0 not worried
3) Having your car stolen (1984–2010)
Question set 2: Anti-social behaviour. ‘How common a problem in this area are...?’Original response categoriesRecode for analysis
4) Vandalism or deliberate damage to property? (1982–2010)1) very big problem 2) fairly big problem 3) not a very big problem 4) not a problem1/2 = 1 a problem
5) Groups of teenagers hanging around? (1982–2010)3/4 = 0 not a problem
6) Noisy neighbours (1984–2010)
Table 3

Dependent variables

Question set 1: Fear of Crime. ‘Could you tell me how worried you are about...?’Original response categoriesRecode for analysis
1) Being mugged/robbed (1982–2010)1) very worried 2) fairly worried 3) not very worried 4) not at all worried1/2 = 1 worried
2) Being burgled (1982–2010)3/4 = 0 not worried
3) Having your car stolen (1984–2010)
Question set 2: Anti-social behaviour. ‘How common a problem in this area are...?’Original response categoriesRecode for analysis
4) Vandalism or deliberate damage to property? (1982–2010)1) very big problem 2) fairly big problem 3) not a very big problem 4) not a problem1/2 = 1 a problem
5) Groups of teenagers hanging around? (1982–2010)3/4 = 0 not a problem
6) Noisy neighbours (1984–2010)
Question set 1: Fear of Crime. ‘Could you tell me how worried you are about...?’Original response categoriesRecode for analysis
1) Being mugged/robbed (1982–2010)1) very worried 2) fairly worried 3) not very worried 4) not at all worried1/2 = 1 worried
2) Being burgled (1982–2010)3/4 = 0 not worried
3) Having your car stolen (1984–2010)
Question set 2: Anti-social behaviour. ‘How common a problem in this area are...?’Original response categoriesRecode for analysis
4) Vandalism or deliberate damage to property? (1982–2010)1) very big problem 2) fairly big problem 3) not a very big problem 4) not a problem1/2 = 1 a problem
5) Groups of teenagers hanging around? (1982–2010)3/4 = 0 not a problem
6) Noisy neighbours (1984–2010)

Results

The first section of the analysis presents the APC models on six separate dependent variables. The controls were generally significant and behaved as expected. The first category (the pre-consensus generation) was the reference category in all cases (Table 2).

Worry about burglary, robbery/mugging and car theft

The results presented in Table 4 confirm findings previously established in the fear of crime literature (Hale 1996). Recent victimization (12 months preceding the interview) is positively associated with worry about burglary, mugging/robbery and car theft. Similarly, as we might have expected, official recorded crime rates are associated with increased worry about all three crimes examined, although the odds ratios are small and not as powerful as recent victimization. Holding other variables constant, women reported higher levels of worry than men, particularly in relation to worry about mugging/robbery. Notably, people from black or minority ethnic backgrounds (compared to white people) also reported higher levels of fear about these three offences. Prominently, respondents from the Asian community expressed over 2.6 times more anxiety about mugging/robbery than their white counterparts. People who were married or cohabiting (compared to single/divorced/separated people); those with the lowest standard of education (compared to those in higher education); those on the lowest 25 per cent incomes (compared to those in the middle 50 per cent and top 25 per cent of incomes) and those living in an inner-city all identified higher levels of worry about these three separate offences. Respondents who identified as unemployed were more likely to state that they had worried about burglary and mugging/robbery than those who were employed. Meanwhile, worry about car theft increased for those who were in full-time work. Noticeably the age of the respondent was significant for car theft; the youngest age category (16–34) reported higher levels of worry about this property-related offence compared to the older groups. Age was not significant for mugging/robbery. Meanwhile participants in rented accommodation expressed lower levels of worry about burglary (compared to people with mortgages/homeowners), but higher anxieties about mugging/robbery and car theft. These results chime with earlier studies (Hale 1996) and describe a wide spectrum of individual, social and economic drivers to the fear of crime.

Table 4

Worry about burglary, mugging/robbery and theft of a car, APC logistic regression model

Worry about burglaryWorry about mugging/robberyWorry about theft of a car
Odds ratios
Sex (female = 0, male = 1)0.810***0.466***0.841***
Age category 16–34 (Ref)
Age category 34–590.9930.9990.957*
Age category 60+0.938*1.0350.919**
White (Ref)
Black1.455***1.540***1.435***
Asian1.894***2.631***1.670***
Mixed/other1.317***1.619***1.202***
Married/cohabiting (Ref)
Single0.802***0.936***0.962*
Widow0.771***0.829***0.816***
Divorced/separated0.786***0.879***0.883***
Employed fulltime (Ref)
Employed part-time0.971*1.0170.898***
Unemployed1.106***1.192***1.002
Education/retired/homemakers/other0.9791.0150.838***
Higher level education (Ref)
A/AS levels/B-tech1.232***1.353***1.307***
O-levels/CSE/GCSE1.346***1.504***1.366***
Other/none1.486***1.798***1.475***
Income—bottom 25% (Ref)
Income—mid 25–75%0.884***0.897***0.919***
Income—top 25%0.773***0.735***0.744***
Total victimization in previous year1.184***1.119***1.209***
Total recorded crime rate per 1,000 population at year of interview1.005***1.007***1.012***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.745***0.650***0.713***
Year of interview0.963***0.976***0.965***
Mortgage/owners (Ref)
Rent0.970**1.154***1.134***
Other/none0.815***1.057*1.003
Pre-consensus generation (Ref)
Post-consensus generation1.431***1.252***1.397***
Wilson/Callaghan generation1.598***1.304***1.545***
Thatcher/Major generation1.583***1.141***1.412***
New Labour generation1.482***1.265***1.566***
Statistics
Constant70,52941,96169,626
n341,665340,575263,274a
Log likelihood453,202417,451342,355
Pseudo R20.0750.1170.090
Worry about burglaryWorry about mugging/robberyWorry about theft of a car
Odds ratios
Sex (female = 0, male = 1)0.810***0.466***0.841***
Age category 16–34 (Ref)
Age category 34–590.9930.9990.957*
Age category 60+0.938*1.0350.919**
White (Ref)
Black1.455***1.540***1.435***
Asian1.894***2.631***1.670***
Mixed/other1.317***1.619***1.202***
Married/cohabiting (Ref)
Single0.802***0.936***0.962*
Widow0.771***0.829***0.816***
Divorced/separated0.786***0.879***0.883***
Employed fulltime (Ref)
Employed part-time0.971*1.0170.898***
Unemployed1.106***1.192***1.002
Education/retired/homemakers/other0.9791.0150.838***
Higher level education (Ref)
A/AS levels/B-tech1.232***1.353***1.307***
O-levels/CSE/GCSE1.346***1.504***1.366***
Other/none1.486***1.798***1.475***
Income—bottom 25% (Ref)
Income—mid 25–75%0.884***0.897***0.919***
Income—top 25%0.773***0.735***0.744***
Total victimization in previous year1.184***1.119***1.209***
Total recorded crime rate per 1,000 population at year of interview1.005***1.007***1.012***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.745***0.650***0.713***
Year of interview0.963***0.976***0.965***
Mortgage/owners (Ref)
Rent0.970**1.154***1.134***
Other/none0.815***1.057*1.003
Pre-consensus generation (Ref)
Post-consensus generation1.431***1.252***1.397***
Wilson/Callaghan generation1.598***1.304***1.545***
Thatcher/Major generation1.583***1.141***1.412***
New Labour generation1.482***1.265***1.566***
Statistics
Constant70,52941,96169,626
n341,665340,575263,274a
Log likelihood453,202417,451342,355
Pseudo R20.0750.1170.090

aThe sample size is smaller here as the question is only fielded to people who have regular access to a car.

*p < 0.05, **p < 0.01, ***p < 0.001.

Table 4

Worry about burglary, mugging/robbery and theft of a car, APC logistic regression model

Worry about burglaryWorry about mugging/robberyWorry about theft of a car
Odds ratios
Sex (female = 0, male = 1)0.810***0.466***0.841***
Age category 16–34 (Ref)
Age category 34–590.9930.9990.957*
Age category 60+0.938*1.0350.919**
White (Ref)
Black1.455***1.540***1.435***
Asian1.894***2.631***1.670***
Mixed/other1.317***1.619***1.202***
Married/cohabiting (Ref)
Single0.802***0.936***0.962*
Widow0.771***0.829***0.816***
Divorced/separated0.786***0.879***0.883***
Employed fulltime (Ref)
Employed part-time0.971*1.0170.898***
Unemployed1.106***1.192***1.002
Education/retired/homemakers/other0.9791.0150.838***
Higher level education (Ref)
A/AS levels/B-tech1.232***1.353***1.307***
O-levels/CSE/GCSE1.346***1.504***1.366***
Other/none1.486***1.798***1.475***
Income—bottom 25% (Ref)
Income—mid 25–75%0.884***0.897***0.919***
Income—top 25%0.773***0.735***0.744***
Total victimization in previous year1.184***1.119***1.209***
Total recorded crime rate per 1,000 population at year of interview1.005***1.007***1.012***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.745***0.650***0.713***
Year of interview0.963***0.976***0.965***
Mortgage/owners (Ref)
Rent0.970**1.154***1.134***
Other/none0.815***1.057*1.003
Pre-consensus generation (Ref)
Post-consensus generation1.431***1.252***1.397***
Wilson/Callaghan generation1.598***1.304***1.545***
Thatcher/Major generation1.583***1.141***1.412***
New Labour generation1.482***1.265***1.566***
Statistics
Constant70,52941,96169,626
n341,665340,575263,274a
Log likelihood453,202417,451342,355
Pseudo R20.0750.1170.090
Worry about burglaryWorry about mugging/robberyWorry about theft of a car
Odds ratios
Sex (female = 0, male = 1)0.810***0.466***0.841***
Age category 16–34 (Ref)
Age category 34–590.9930.9990.957*
Age category 60+0.938*1.0350.919**
White (Ref)
Black1.455***1.540***1.435***
Asian1.894***2.631***1.670***
Mixed/other1.317***1.619***1.202***
Married/cohabiting (Ref)
Single0.802***0.936***0.962*
Widow0.771***0.829***0.816***
Divorced/separated0.786***0.879***0.883***
Employed fulltime (Ref)
Employed part-time0.971*1.0170.898***
Unemployed1.106***1.192***1.002
Education/retired/homemakers/other0.9791.0150.838***
Higher level education (Ref)
A/AS levels/B-tech1.232***1.353***1.307***
O-levels/CSE/GCSE1.346***1.504***1.366***
Other/none1.486***1.798***1.475***
Income—bottom 25% (Ref)
Income—mid 25–75%0.884***0.897***0.919***
Income—top 25%0.773***0.735***0.744***
Total victimization in previous year1.184***1.119***1.209***
Total recorded crime rate per 1,000 population at year of interview1.005***1.007***1.012***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.745***0.650***0.713***
Year of interview0.963***0.976***0.965***
Mortgage/owners (Ref)
Rent0.970**1.154***1.134***
Other/none0.815***1.057*1.003
Pre-consensus generation (Ref)
Post-consensus generation1.431***1.252***1.397***
Wilson/Callaghan generation1.598***1.304***1.545***
Thatcher/Major generation1.583***1.141***1.412***
New Labour generation1.482***1.265***1.566***
Statistics
Constant70,52941,96169,626
n341,665340,575263,274a
Log likelihood453,202417,451342,355
Pseudo R20.0750.1170.090

aThe sample size is smaller here as the question is only fielded to people who have regular access to a car.

*p < 0.05, **p < 0.01, ***p < 0.001.

In addition to the expected variables related to crime fears, the results herein also reveal a complex political-generational story. The additional period and cohort effects in our model add new statistically significant considerations. The year of interview was significant for worry about burglary, mugging/robbery and car theft, although the odds ratios were modest. There were also significant coefficients in the model related to the political generation respondents were assigned to. The pre-consensus generation reported the least amount of worry about burglary, mugging/robbery and car theft compared to all other political generations. This in itself is noteworthy because this generation was the eldest when surveyed, suggesting that the fear of crime–age nexus is only part of the story. Compared to the reference category, the Wilson/Callaghan and Thatcher/Major generations expressed the highest level of worry about domestic burglary—which is salient, given the fact the Thatcher generation ‘came of age’ during the dramatic rise in property crime during the 1980s and the attendant focus on ‘law and order’ by politicians. The Wilson/Callaghan generation expressed the highest levels of worry about mugging/robbery. Interestingly, this political generation spent their formative years during a time in which there was considerable public attention to and anxiety about a ‘crime wave’ of ‘mugging’. More specifically, this narrative was wrapped up with ethnic tensions and the considered threat young black men posed on the streets (Hall et al. 1978). Examining car theft, we found that the New Labour and Wilson/Callaghan generation reported the highest levels of worry about theft of a car, although this offence peaked in frequency in the mid-1990s.14

Perceptions of antisocial behaviour

As one might expect from the literature, the results presented in Table 5 regarding public perceptions of antisocial behaviour tell a similar story to the results earlier in relation to worry about crime. Females (compared to males) and people from Asian backgrounds (compared to white) were more likely to perceive ‘teenagers hanging around’, ‘noisy neighbours’ and ‘vandalism’ in the local area as a fairly or very big ‘common problem’. Those with the lowest education (compared to those with the highest qualifications), those with the lowest incomes (compared to the top 25 per cent), those who were unemployed (compared to people in full-time employment), those who were married (compared to all other categories), those living in rented accommodation (compared to those who owned their homes) and those who resided in an inner-city all reported higher levels of concern about ‘antisocial’ behaviours, namely noisy neighbours, visible groups of teenagers ‘hanging around’ and vandalism. As before, these results reaffirm much of the literature on antisocial behaviour and disorder discussed previously.

Table 5

Perceptions of problematic behaviour in local area: APC logistic regression model

Common local problem: ‘noisy neighbours’Common local problem: ‘teenagers hanging around’Common local problem: ‘vandalism and deliberate damage’
Odds ratios
Sex (female = 0, male = 1)0.874***0.926***0.972***
Age category 16–34 (Ref)
Age category 34–590.956*0.971*1.055***
Age category 60+0.787***0.834***0.957***
White (Ref)
Black1.0500.9880.937*
Asian1.237***1.318***1.065*
Mixed/other1.185***1.096***0.963
Married/cohabiting (Ref)
Single0.9770.921***1.001
Widow0.725***0.863***0.799***
Divorced/ separated0.9880.953***0.942***
Employed fulltime (Ref)
Employed part-time0.908***0.954***1.068***
Unemployed1.162***1.072***1.150***
Education/retired/homemaker/other0.906***0.870***0.999
Higher level education (Ref)
A/AS levels/B-tech1.055*1.214***1.181***
O-levels/CSE/GCSE1.057***1.235***1.240***
Other/none1.121***1.277***1.240***
Income—bottom 25% (Ref)
Income—mid 25–75%0.877***0.870***0.910***
Income—top 25%0.599***0.648***0.672***
Total victimization in previous year1.159***1.262***1.362***
Total recorded crime rate per 1,000 population at year of interview0.993***1.0001.016***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.662***0.697***0.580***
Year of interview0.9980.988***1.018***
Mortgage/owners (Ref)
Rent1.655***1.239***1.167***
Other/none1.0281.0451.019
Pre-consensus generation (Ref)
Post-consensus generation1.463***1.461***1.421***
Wilson/Callaghan generation1.770***1.811***1.645***
Thatcher/Major generation1.955***2.012***1.693***
New Labour generation2.223***2.296***1.985***
Statistics
Constant1.50622.912−38.686
N422,680422,249422,033
Log likelihood256,885480,903453,099
Pseudo R20.0740.0870.102
Common local problem: ‘noisy neighbours’Common local problem: ‘teenagers hanging around’Common local problem: ‘vandalism and deliberate damage’
Odds ratios
Sex (female = 0, male = 1)0.874***0.926***0.972***
Age category 16–34 (Ref)
Age category 34–590.956*0.971*1.055***
Age category 60+0.787***0.834***0.957***
White (Ref)
Black1.0500.9880.937*
Asian1.237***1.318***1.065*
Mixed/other1.185***1.096***0.963
Married/cohabiting (Ref)
Single0.9770.921***1.001
Widow0.725***0.863***0.799***
Divorced/ separated0.9880.953***0.942***
Employed fulltime (Ref)
Employed part-time0.908***0.954***1.068***
Unemployed1.162***1.072***1.150***
Education/retired/homemaker/other0.906***0.870***0.999
Higher level education (Ref)
A/AS levels/B-tech1.055*1.214***1.181***
O-levels/CSE/GCSE1.057***1.235***1.240***
Other/none1.121***1.277***1.240***
Income—bottom 25% (Ref)
Income—mid 25–75%0.877***0.870***0.910***
Income—top 25%0.599***0.648***0.672***
Total victimization in previous year1.159***1.262***1.362***
Total recorded crime rate per 1,000 population at year of interview0.993***1.0001.016***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.662***0.697***0.580***
Year of interview0.9980.988***1.018***
Mortgage/owners (Ref)
Rent1.655***1.239***1.167***
Other/none1.0281.0451.019
Pre-consensus generation (Ref)
Post-consensus generation1.463***1.461***1.421***
Wilson/Callaghan generation1.770***1.811***1.645***
Thatcher/Major generation1.955***2.012***1.693***
New Labour generation2.223***2.296***1.985***
Statistics
Constant1.50622.912−38.686
N422,680422,249422,033
Log likelihood256,885480,903453,099
Pseudo R20.0740.0870.102

*p < 0.05, **p < 0.01, ***p < 0.001.

Table 5

Perceptions of problematic behaviour in local area: APC logistic regression model

Common local problem: ‘noisy neighbours’Common local problem: ‘teenagers hanging around’Common local problem: ‘vandalism and deliberate damage’
Odds ratios
Sex (female = 0, male = 1)0.874***0.926***0.972***
Age category 16–34 (Ref)
Age category 34–590.956*0.971*1.055***
Age category 60+0.787***0.834***0.957***
White (Ref)
Black1.0500.9880.937*
Asian1.237***1.318***1.065*
Mixed/other1.185***1.096***0.963
Married/cohabiting (Ref)
Single0.9770.921***1.001
Widow0.725***0.863***0.799***
Divorced/ separated0.9880.953***0.942***
Employed fulltime (Ref)
Employed part-time0.908***0.954***1.068***
Unemployed1.162***1.072***1.150***
Education/retired/homemaker/other0.906***0.870***0.999
Higher level education (Ref)
A/AS levels/B-tech1.055*1.214***1.181***
O-levels/CSE/GCSE1.057***1.235***1.240***
Other/none1.121***1.277***1.240***
Income—bottom 25% (Ref)
Income—mid 25–75%0.877***0.870***0.910***
Income—top 25%0.599***0.648***0.672***
Total victimization in previous year1.159***1.262***1.362***
Total recorded crime rate per 1,000 population at year of interview0.993***1.0001.016***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.662***0.697***0.580***
Year of interview0.9980.988***1.018***
Mortgage/owners (Ref)
Rent1.655***1.239***1.167***
Other/none1.0281.0451.019
Pre-consensus generation (Ref)
Post-consensus generation1.463***1.461***1.421***
Wilson/Callaghan generation1.770***1.811***1.645***
Thatcher/Major generation1.955***2.012***1.693***
New Labour generation2.223***2.296***1.985***
Statistics
Constant1.50622.912−38.686
N422,680422,249422,033
Log likelihood256,885480,903453,099
Pseudo R20.0740.0870.102
Common local problem: ‘noisy neighbours’Common local problem: ‘teenagers hanging around’Common local problem: ‘vandalism and deliberate damage’
Odds ratios
Sex (female = 0, male = 1)0.874***0.926***0.972***
Age category 16–34 (Ref)
Age category 34–590.956*0.971*1.055***
Age category 60+0.787***0.834***0.957***
White (Ref)
Black1.0500.9880.937*
Asian1.237***1.318***1.065*
Mixed/other1.185***1.096***0.963
Married/cohabiting (Ref)
Single0.9770.921***1.001
Widow0.725***0.863***0.799***
Divorced/ separated0.9880.953***0.942***
Employed fulltime (Ref)
Employed part-time0.908***0.954***1.068***
Unemployed1.162***1.072***1.150***
Education/retired/homemaker/other0.906***0.870***0.999
Higher level education (Ref)
A/AS levels/B-tech1.055*1.214***1.181***
O-levels/CSE/GCSE1.057***1.235***1.240***
Other/none1.121***1.277***1.240***
Income—bottom 25% (Ref)
Income—mid 25–75%0.877***0.870***0.910***
Income—top 25%0.599***0.648***0.672***
Total victimization in previous year1.159***1.262***1.362***
Total recorded crime rate per 1,000 population at year of interview0.993***1.0001.016***
Inner-city (0 = inner-city resident,1 = non-inner-city)0.662***0.697***0.580***
Year of interview0.9980.988***1.018***
Mortgage/owners (Ref)
Rent1.655***1.239***1.167***
Other/none1.0281.0451.019
Pre-consensus generation (Ref)
Post-consensus generation1.463***1.461***1.421***
Wilson/Callaghan generation1.770***1.811***1.645***
Thatcher/Major generation1.955***2.012***1.693***
New Labour generation2.223***2.296***1.985***
Statistics
Constant1.50622.912−38.686
N422,680422,249422,033
Log likelihood256,885480,903453,099
Pseudo R20.0740.0870.102

*p < 0.05, **p < 0.01, ***p < 0.001.

Notably, respondents who had been a victim of crime during the 12 months preceding the interview were more likely to report anxiety about all three measures of neighbourhood ‘problems’. However, national crime rates followed a different pattern to personal experience of crime; not only were the odds ratios more modest, but crime rates were positively associated with ‘vandalism’, negatively associated with ‘noisy neighbours’ and not significant for ‘teenagers hanging around’. This highlights how immediate and proximate experience measures of crime (i.e. victimisation) are more sensitive to perceptions of local disorder than aggregate crime rates (Farrall et al. 2009).

Turning to the impact of APC effects, the youngest age group (age 16–34) reported significantly higher levels of concern about ASB than all of their older counterparts. Year of interview was significant for perception of teenagers and vandalism, although the coefficients on these variables were small. As demonstrated with the fear of crime earlier, however, membership of a political generation was the strongest predictor in these multivariate models. The New Labour generation reported the highest levels of concern about these three forms of antisocial behaviour, followed by the Thatcher/Major generation. Those who grew up in the era of New Labour reported that ‘teenagers hanging around’ the local area, ‘noisy neighbours’ and vandalism posed a fairly or very big problem in the neighbourhood, compared to the reference category. This suggests the public discourse around antisocial behaviour that emerged towards the end of Major’s time in office and which became fully developed during the New Labour administrations (Burney 2005) had biggest impact on the youngest political generation in this study. Not only did this generation consider visible groups of teenagers—of whom they were not much older—as ‘problematic’, they were also more likely to associate noisy neighbours and vandalism as a problem. In short, the political generation who came of age during an era where criminal justice was intensely focused on ‘antisocial behaviour’ later exhibited the lowest tolerance for low-level public order offences.

Given the large sample size and the constraints of the APC model, it is imperative to test the results robustly (Yang and Land 2013; Bell and Jones 2014). To do this, we employed a three-step method to verify the estimates, involving regression analysis, Wald tests and GAMs (Grasso 2014). The Wald tests allow us to formally test whether certain political generations express stronger emotional reactions to crime and antisocial behaviour than others did. Although the APC logistic regression models presented in Tables 4 and 5 allow us to assess whether differences between each cohort included in the model and the reference category are significant, Wald tests allow us to test for coefficient differences between the cohorts included in the model. The results for the Wald tests are presented in Table 6 and confirm the overwhelming influence of political generations when considering worry about crime and perceptions of antisocial behaviour.

Table 6

Wald tests for intergenerational differences from the APC models

Worry about burglaryWorry about mugging/ robberyWorry about car theftHow common a problem… are noisy neighbours?How common a problem... are teenagers hanging around?How common a problem... is vandalism?
Post-War Consensus Generation &*****************
Wilson/Callaghan Generation
Post-War Consensus Generation &******n.s*********
Thatcher/Major Generation
Post-War Consensus Generation &n.sn.s***********
New Labour Generation
Wilson/Callaghan Generation &n.s*************
Thatcher/ Major Generation
Wilson/Callaghan Generation &**n.sn.s*********
New Labour Generation
Thatcher/Major Generation &******************
New Labour Generation
Worry about burglaryWorry about mugging/ robberyWorry about car theftHow common a problem… are noisy neighbours?How common a problem... are teenagers hanging around?How common a problem... is vandalism?
Post-War Consensus Generation &*****************
Wilson/Callaghan Generation
Post-War Consensus Generation &******n.s*********
Thatcher/Major Generation
Post-War Consensus Generation &n.sn.s***********
New Labour Generation
Wilson/Callaghan Generation &n.s*************
Thatcher/ Major Generation
Wilson/Callaghan Generation &**n.sn.s*********
New Labour Generation
Thatcher/Major Generation &******************
New Labour Generation

A significant result implies cohort differences between each given pair in the rows for each of the dependent variables in the columns. For the direction of difference see model estimates.

*p < 0.05, **p < 0.01, ***p < 0.001, n.s. signifies non-statistically significant differences.

Table 6

Wald tests for intergenerational differences from the APC models

Worry about burglaryWorry about mugging/ robberyWorry about car theftHow common a problem… are noisy neighbours?How common a problem... are teenagers hanging around?How common a problem... is vandalism?
Post-War Consensus Generation &*****************
Wilson/Callaghan Generation
Post-War Consensus Generation &******n.s*********
Thatcher/Major Generation
Post-War Consensus Generation &n.sn.s***********
New Labour Generation
Wilson/Callaghan Generation &n.s*************
Thatcher/ Major Generation
Wilson/Callaghan Generation &**n.sn.s*********
New Labour Generation
Thatcher/Major Generation &******************
New Labour Generation
Worry about burglaryWorry about mugging/ robberyWorry about car theftHow common a problem… are noisy neighbours?How common a problem... are teenagers hanging around?How common a problem... is vandalism?
Post-War Consensus Generation &*****************
Wilson/Callaghan Generation
Post-War Consensus Generation &******n.s*********
Thatcher/Major Generation
Post-War Consensus Generation &n.sn.s***********
New Labour Generation
Wilson/Callaghan Generation &n.s*************
Thatcher/ Major Generation
Wilson/Callaghan Generation &**n.sn.s*********
New Labour Generation
Thatcher/Major Generation &******************
New Labour Generation

A significant result implies cohort differences between each given pair in the rows for each of the dependent variables in the columns. For the direction of difference see model estimates.

*p < 0.05, **p < 0.01, ***p < 0.001, n.s. signifies non-statistically significant differences.

Generalized additive modelling

To validate the results from the APC models, we also assessed the results from the GAMs. In particular, we examined the plots of the smoothed cohort effect from the full model with all controls included, as presented in Figures 3–8. Across all six indicators, the patterns are unambiguous; there is an upward swing in worry about burglary, worry about mugging/robbery, car theft and perceptions of antisocial behaviour once the Wilson/Callaghan generation is born after the 1940s. These patterns suggest that anxiety about crime and antisocial behaviour grew considerably and notably peaked for some generations around particular historical moments. Worry about burglary (Figure 3) noticeably peaks for those who spent their formative years living under the Wilson/Callaghan and Thatcher/Major administrations. Worry about mugging/robbery (Figure 4) fluctuates somewhat but is most apparent, again, for those born after 1940 (the Wilson/Callaghan generation). Similarly, worry about car theft fluctuates with inclines around those born after the 1940s (the Wilson/Callaghan generation) and during the 1980s (the New Labour Generation) (Figure 5). Meanwhile, perceptions of antisocial behaviour (Figures 6–8) showed considerable synergy across the three measures; they all demonstrate an increase in concern amongst people born in the 1960s (i.e. the Thatcher/Major generation) before peaking for those who were born around the early 1980s (the New Labour Generation). In all examples, the curves do not return to the levels of the pre- and early consensus generations. As such, we have a clear and upwards tendency towards greater worry about crime and antisocial behaviour over time, which provides considerable support for the theoretical literature discussed earlier.

Worry about burglary
Fig. 3

Worry about burglary

Worry about mugging/robbery
Fig. 4

Worry about mugging/robbery

Worry about car theft
Fig. 5

Worry about car theft

How common a problem are teenagers hanging around?
Fig. 6

How common a problem are teenagers hanging around?

How common a problem are noisy neighbours?
Fig. 7

How common a problem are noisy neighbours?

How common a problem is vandalism or deliberate damage to property?
Fig. 8

How common a problem is vandalism or deliberate damage to property?

Conclusion

The criminological literature acknowledges that fear of crime and perceptions of antisocial behaviour may be shaped by the political and social landscapes an individual resides in. However, despite the scale of this work, this is the first study to adopt an APC method to explore the relevance of political generations on worry about crime and disorder. Drawing on generational modelling and political socialization theories, we have sought to test if the political era in which a cohort spent their formative years made them more likely to report anxiety or concern about manifestations of crime and antisocial behaviour. In addition, we asked if such cohort effects could persist net of associated individual-level attributes and other demographic factors. Our results outline the theoretical and empirical significance of political cohorts and although there are challenges involved in APC investigations, we found that the method has considerable potential, particularly for those committed to interdisciplinary evaluation.

Our results indicate that the political generation in which a respondent spent their ‘formative years’ was a significant predictor of various manifestations of crime fears, namely worry about burglary, mugging/robbery, theft of a car and identification of antisocial behaviour as a local ‘problem’. Although our modelling confirmed the well-established socio-demographic antecedents of fear of crime and perceptions of antisocial behaviour (Hale 1996; Ferraro 1995; Farrall et al. 2009), it also revealed a longitudinal story of enduring anxieties about crime via political generational differences. Notably, the Wilson/Callaghan and Thatcher/Major generations expressed the greatest level of worry about domestic burglary. This is a notable finding, given that these generations ‘came of age’ as crime rose dramatically (particularly so in terms of burglary during the 1980s, Morgan 2014) and political attention towards ‘law and order’ increased in the United Kingdom and United States (Garland 2001; Farrall et al. 2009). The Wilson/Callaghan generation expressed the highest levels of worry about ‘mugging’ and robbery—an offence that featured in intense public debates throughout the years of their political socialization (Hall et al. 1978). Meanwhile, the New Labour generation reported the highest level of concern about ‘problems’ in the local area, such as vandalism, ‘teenagers hanging around’ and noisy neighbours, all of which were controversial motifs (Burney 2005; Crawford 2009) of ‘antisocial behaviour’ that the New Labour administration strongly emphasized. Our analysis appears to confirm the notion that political socialization can be a formative process with the ability to cast an enduring influence on public perceptions of crime and antisocial behaviour. Indeed, we found there were substantial differences between the political cohorts tested, with some exhibiting high levels of concern about problems in their environment caused by antisocial behaviour, whereas others expressed greater worry about property crime or violent crime. The example herein demonstrates the value of developing a close dialogue with political history (c.f. Gottschalk 2006; Aviram 2015; Enns 2016; Loader and Sparks 2016), and perhaps, as criminologists we could profit further from creating further opportunities to examine generational effects.

This exploration has reiterated the diverse components that are associated with beliefs about crime. Fear of crime and perceptions of antisocial behaviour can be responsive to both episodic tangible experiences (i.e. recent victimization) and dynamic personal circumstances (i.e. tenure or employment), as well as borne out of deep-rooted attitudes that have their origins in the political conditions that individuals were exposed to growing up. Entangled in our narratives of crime and antisocial behaviour are hidden stories that can tell us something important about our current circumstances and, in some cases, our political history.

Our work implies that political socialization can cast a long-term impression on public attitudes about crime. It can exert an influence as individuals’ transition from young adulthood into middle age and as new policy cultures become established. Those who grew up during the Callaghan/Wilson generation (formative period 1965–78)—a time when Hall et al. wrote the seminal Policing the Crisis: Mugging, the State and Law—were more likely than others to report anxiety about mugging into their middle age. Likewise, children who grew up with the Thatcher and Major governments—with its concomitant high crime rates and broad emphasis on law and order (Hay 1996)—sustained high levels of worry about property crime and antisocial behaviour up to and into the 2010s. This underlies the idea that concerns about crime can involve an enduring legacy. As Elias has noted, ‘the strength, kind and structures of the fears and anxieties that smoulder or flare in the individual never depend solely on his own “nature”’. Rather they are ‘always determined, finally by the history and the actual structure of his relations to other people’ (1982: 327). Building temporally sensitive theoretical frameworks has been stressed by Pierson (2004), who has argued that when politics is understood as a ‘moving picture’ (2004: 2) rather than a static scene, it has the ability to enrich our understanding of social dynamics and improve the theories and methods that we use to explain them. Indeed, criminologists have much to learn from the writing of historical institutionalists who have for some time pointed to the influence of ‘slow-moving causal processes’, which means ‘systematically situating particular moments (including the present) in a temporal sequence of events and processes stretching over extended periods’ (Pierson 2004: 2).

From our perspective, the value of generational analysis is not simply about expanding our appreciation of the ‘wider contexts’. Fear of crime and perceptions of antisocial behaviour are likely shaped via a broad and long sequence of events, and the pattern of those events may be critical in determining the meanings ascribed to those particular crimes by the public. Assuming that, as we argue, people travel through life with some of the political anxieties and beliefs that they were exposed to in their formative years, then, as that generation takes on the mantle of social, economic and political leadership (i.e. as they become parents, employers, voters and part of the moral leadership of a society), so their formative experiences shape both their own responses to pressing social issues (such as crime) and the practical and policy responses they pursue. In this way, the fear of crime is not simply an immediate response to a perceived threat, but also a lagged consequence of earlier policy debates.

Although this article has presented some novel findings, there are other approaches one could take here. Alternative generational schema might also produce interesting results, such as generations based on economic phases or fluctuations in crime. As part of our work, we tested a method for dividing the sample into ‘crime cohorts’, based on marked shifts in the long-term patterns of crime in England and Wales.15 Although this configuration was theoretically and empirically important to test and provided some useful data, statistically it was not as powerful as those models based on political generations in this instance. Certainly, understanding the public’s emotional relationship to crime and antisocial behaviour is an intricate and dynamic task; we cannot directly explain the casual mechanisms that account for the association between political generations and fear of crime and perception of neighbourhood problems. It could reflect differences in the nature of policy debates that took place during one’s formative years (Garland 2001), or the ways in which a particular crime is framed culturally (Hall et al. 1978), but it could also represent the interaction of both macro or micro-level processes (Enns 2016; Jennings et al. 2016). Notably, although the modelling herein has revealed some meaningful findings, a substantial amount of variation remains unexplained. Employing secondary longitudinal data limits one to the variables that were considered necessary at the inception of the original survey, and comparable future research might benefit from other variables that might afford a tighter model. Nevertheless, the results clearly emphasize the relevance of political influence and suggest that a comprehensive understanding of fear of crime and perceptions of antisocial behaviour requires information about both the characteristics of individuals, their community and the political environment in which they live.

In Memoriam

In Memory of Sir Roger Jowell (1942–2011) and Professor Jason Ditton (1949–2015), both of whom worked tirelessly to produce the data needed to study social issues such as the fear of crime, and whose ideas have shaped a vast body of work, including our own.

Acknowledgments

We would like to thank the reviewers for their helpful comments and the ESRC for their generous funding.

Funding

This project was funded by the ESRC (award No. ES/K006398/1).

Footnotes

1

The terms ‘came of age’ and ‘formative years’ were developed by Mannheim (1928) and refer to a transitional period in which individuals are deeply sensitive to social events and form key opinions. This typically occurs between the age of 15 and 25. Older members of society may experience these same events but they will interpret them through a different (earlier) set of formative experiences.

2

Recorded property crime in England and Wales reveals a steady rate of growth for all domestic burglaries from 1960 to the mid-1980s, after which there was a sudden spike that lasted up until 1993 and a decrease thereafter (Office of National Statistics 2012).

3

Fear of crime in England and Wales, as measured by the BCS, began to fall in the mid-1990s after a peak in 1994 after which it reduced consistently before stabilizing around 2005.

4

We recognize the unique political and cultural relevance of the term ‘antisocial behaviour’ as it applies in the United Kingdom. In the 1990s, the New Labour government asserted that antisocial behaviour should be a major policy priority (Burney 2005). Antisocial behaviour referred not only to low-level criminal offences, but also to civil matters such as ‘noisy neighbours’ and ‘intimidation’ by groups of young people. Consequently, the Antisocial Behaviour Act 2003 gave the police powers to disperse groups of two or more individuals where their presence or behaviour had resulted, or was likely to result, in a member of the public being alarmed or distressed (Crawford 2009).

5

Given the political and media emphasis on antisocial behaviour in the late 1990s (Burney 2005), it is perhaps unsurprising that public concern about neighbourhood incivilities did not show a reduction during the early part of the 2000s. BCS data from 1984 to 2012/13 demonstrate that concern about ‘noisy neighbours’, ‘vandalism’ and ‘teenagers hanging around’ fluctuated considerably in the 1980s before stabilizing in the 1990s and has remained fairly constant up to 2012/13.

6

We tested a range of different age groupings including a three-level and four-level categorization. Because there were no statistical differences in the results between these two categorizations, we opted for the more parsimonious three-level version.

7

See also the Conclusion for further discussion on an analysis of cohorts based on crime rates.

8

The controversial ‘ASBO’ was borne of the Crime and Disorder Act 1998. Before this, incidences of intimidation, criminal damage or vandalism were processed separately through criminal law (Crawford 2009).

9

The key period of socialization will largely depend on the mechanism implied in theory (Bartels and Jackman 2014). Given that we are interested in the influence of political socialization on attitudes towards crime and antisocial behaviour, we follow Grasso’s (2014) method.

10

First conducted in 1982, the BCS moved to an annual collection from 2001. As well as collecting information on victimization and fear of crime, it includes attitudinal data and demographic information. The survey sampling is structured to be representative of two groups, namely residential households in England and Wales, and adults (aged 16 years and over) living in those households. The survey does not collect information from those living in alternative accommodation. The BCS sample size has increased from 11,000 in 1982 to over 47,000 in 2005/06. There have also been further methodological changes, e.g. adjustments for non-response to improve the quality of the survey (Jansson 2007).

11

Wards are the key building blocks of UK administrative geography. Across England and Wales, wards have a mean population size of 6,600 (Office of National Statistics 2013).

12

One measure of antisocial behaviour refers to whether ‘teenagers hanging around’ was perceived as problematic. Scholars have noted that from the late 20th century young people in the United Kingdom were unduly associated with crime and antisocial activities (Burney 2005; Crawford 2009). A United Nations report on the Rights of the Child stated that the human rights of young people in the United Kingdom were threatened by a ‘general climate of intolerance and negative public attitudes’ towards young people exemplified by antisocial behaviour prevention that degraded the youth population (2008: 6).

13

Following previous studies (i.e. Farrall et al. 2009; Jackson and Gray 2010; Pantazis 2000), we classify respondents who said they were ‘not very worried’ about various crimes as unworried for the purposes of our dichotomous modelling strategy (see Table 3). Similarly, those who said that local environmental matters were ‘not a big problem’ were recoded as not concerned about the form of ASB in question. Such responses acknowledge minimal or insignificant experience of worry or concern. We therefore believe they can be meaningfully compared to instances where explicit worry about crime and/or problematic antisocial behaviour is reported.

14

Unauthorized theft of a motor vehicle peaked in 1993 with 592,660 (Office of National Statistics 2012).

15

In testing different cohort formulations, we devised one based on ‘crime cohorts’ based on the long-term rise and fall of all recorded crime in England and Wales from 1940 to 2012. The analysis repeated the same three-step model employed herein, regression analysis, post-estimation Wald tests and GAM. The cohorts were ‘consistently low’ (1940–55); ‘slow and subtle increase’ (1956–76); ‘sharp rise’ (1977–92); ‘fluctuating’ (1993–2004) and ‘decreasing’ (2005–12). For further information on the results, please contact the corresponding author.

References

Addison
,
P
. (
1994
),
The Road to 1945: British Politics and the Second World War
, 2nd edn.
Pimlico
.

Aviram
,
H
. (
2015
),
Cheap on Crime: Recession-Era Politics and the Transformation of American Punishment
.
University of California Press
.

Bartels
,
L.
and
Jackman
,
S
. (
2014
), ‘
A Generational Model of Political Learning
’,
Electoral Studies
,
33
:
7
18
.

Baumer
,
E.
,
Messner
,
S.
and
Rosenfeld
,
R
. (
2003
), ‘
Explaining Spatial Variation in Support for Capital Punishment: A Multilevel Analysis
’.
American Journal of Sociology
,
108
:
844
75
.

Beck
,
N.
and
Jackman
,
S
. (
1998
), ‘
Beyond Linearity by Default: Generalised Additive Models
’,
American Journal of Political Science
,
42
:
596
627
.

Beckett
,
K.
and
Western
B
. (
2000
), ‘
Crime Control, American Style: From Social Welfare to Social Control
’, in
P.
Green
and
A
.
Rutherford
,
eds,
Criminal Policy in Transition
.
Hart Publishing
.

Bell
,
A.
and
Jones
,
K
. (
2014
), ‘
Don’t Birth Cohorts Matter? A Commentary and Simulation Exercise on Reither, Hauser, and Yang’s (2009) Age-Period-Cohort Study
’,
Social Science and Medicine
,
101
:
176
80
.

Burney
,
E
. (
2005
),
Making People Behave: Antisocial Behaviour, Politics and Policy
.
Willan
.

Bursik
,
R.
and
Grasmick
,
H
. (
1993
),
Neighborhoods and Crime: The Dimensions of Effective Community Control
.
Lexington Books
.

Butler
,
D.
and
Kavanagh
,
D
. (
1997
),
The British General Election 1997
.
Macmillan
.

Crawford
,
A
. (
2009
), ‘
Criminalizing Sociability through Antisocial Behaviour Legislation: Dispersal Powers, Young People and the Police
’,
Youth Justice
,
9
:
5
26
.

Driver
,
S.
and
Martell
,
L
. (
2002
),
Blair’s Britain
.
Polity Press
.

Elias
,
N
. (
1982
),
The Civilizing Process, State Formation and Civilization, Vol. 2
,
Basil Blackwell
.

Enns
,
P
. (
2016
),
Incarceration Nation: How the United States Became the Most Punitive Democracy in the World
.
Cambridge University Press
.

Farrall
,
S.
,
Jackson
,
J.
and
Gray
,
E
. (
2009
),
Social Order and the Fear of Crime in Contemporary Times
.
Oxford University Press
.

Farrall
,
S.
and
Hay
,
C
. (
2014
),
The Legacy of Thatcherism: Assessing and Exploring Thatcherite Social and Economic Policies
.
Oxford
:
Oxford University Press
.

Farrington
,
D.
and
Jolliffe
,
D
. (
2004
), ‘
England and Wales
’ in
D
.
Farrington
,
P
.
Langan
,
M
.
Tonry
, eds.,
Cross-National Studies in Crime and Justice
.
Bureau of Statistics
.

Ferraro
,
K
. (
1995
),
Fear of Crime: Interpreting Victimisation Risk
.
SUNY Press
.

Ferraro
,
K.
and
Shippee
,
T
. (
2009
), ‘
Aging and Cumulative Inequality: How Does Inequality Get Under the Skin
?’,
The Gerontologist
,
49
:
333
43
.

Frost
,
N
. (
2010
), ‘
Beyond Public Opinion Polls: Punitive Public Sentiment and Criminal Justice Policy
’,
Sociology Compass
,
4
:
156
68
.

Garland
,
D
. (
2001
),
The Culture of Control
.
Oxford University Press
.

Giddens
,
A
. (
1998
),
The Third Way: The Renewal of Social Democracy
.
Polity Press
.

Gilmore
,
R
. (
2007
),
The Golden Gulag: Prisons, Surplus, Crisis, and Opposition in Globalizing California
.
University of California Press
.

Goffman
,
E
. (
1971
),
Relations in Public
.
Basic Books
.

Gottschalk
,
M
. (
2006
),
The Prison and the Gallows: The Politics of Mass Incarceration in America
.
Cambridge University Press
.

Grasso
,
M
. (
2014
), ‘
Age-Period-Cohort Analysis in a Comparative Context: Political Generations and Political Participation Repertoires
’,
First View, Electoral Studies
,
33
:
63
76
.

—— (

2016
),
Generations, Political Participation and Social Change in Western Europe
.
Routledge
.

Grasso
,
M.
,
Farrall
,
S.
,
Gray
,
E.
,
Hay
,
C.
and
Jennings
,
W
. (
2017
), ‘
Thatcher’s Children, Blair’s Babies, Political Socialisation and Trickle-Down Value-Change: An Age, Period and Cohort Analysis
’,
British Journal of Political Science
,
1
20
, doi:.

Hagan
,
J
. (
2010
),
Who Are the Criminals?
Princeton University Press
.

Hale
,
C
. (
1996
), ‘
Fear of Crime: A Review of the Literature
’,
International Review of Victimology
,
4
:
79
150
.

Hall
,
S.
,
Critcher
,
C.
,
Jefferson
,
T.
,
Clarke
,
J.
and
Roberts
,
B
. (
1978
),
Policing the Crisis: Mugging, the State and Law
.
Macmillan
.

Hansard
. (
1998
),
February 1998, column 603
, available online at https://www.publications.parliament.uk/pa/ld199798/ldhansrd/vo980203/text/80203–30.htm (Accessed
July 2018
).

Harcourt
,
B
. (
2001
),
Illusion of Order: The False Promise of Broken Windows Policing
.
Harvard University Press
.

Hay
,
C
. (
1996
),
Re-stating Social and Political Change
.
Open University Press
.

Home Office
.(
2012
),
Recorded Crime Statistics for England and Wales 1898 – 2001/02
.
Home Office
.

—— (

2014
),
Recorded Crime Statistics for England and Wales 2002/03 – 2013/14
.
Home Office
.

Hough
,
M
. (
1995
),
Anxiety About Crime: Findings from the 1994 British Crime Survey, Home Office Research Study No. 147
.
Home Office
.

Innes
,
M
. (
2014
),
Signal Crimes: Social Reactions to Crime, Disorder, and Control
.
Oxford University Press
.

Jansson
,
K
. (
2007
),
British Crime Survey – Measuring Crime for 25 years
.
Home Office
.

Jackson
,
J.
and
Gray
,
E
. (
2010
). ‘
Functional Fear and Public Insecurities about Crime
’,
British Journal of Criminology
,
50
:
1
21
.

Jennings
,
W.
,
Farrall
,
S.
,
Gray
,
E.
and
Hay
,
C
. (
2016
), ‘
Penal Populism and the Public Thermostat: Crime, Public Punitiveness and Public Policy
’,
Governance: An International Journal of Policy, Administration and Institutions
,
30
:
463
81
.

Jennings
,
W.
,
Gray
,
E.
,
Hay
,
C.
and
Farrall
,
S
. (
2015
), ‘
Collating Longitudinal Data on Crime, Victimisation and Social Attitudes in England and Wales: A New Resource for Exploring Long-term Trends in Crime
’,
British Journal of Criminology
,
55
:
1005
1015
.

Jennings
,
W.
,
Farrall
,
S.
,
Gray
,
E
, and
Hay
,
C
. (
2018
), ‘
Moral Panics and Punctuated Equilibrium in Public Policy: an analysis of the criminal justice policy agenda in Britain
’,
Policy Studies Journal
. [Epub ahead of Print].

Jiang
,
B.
and
Carriere
,
K
. (
2014
), ‘
Age-Period-Cohort Models Using Smoothing Splines: A Generalised Additive Model Approach
’,
Statistics in Medicine
,
33
:
595
606
.

Kruger
,
D.
,
Reischl
,
T.
and
Gee
,
G
. (
2007
), ‘
Neighborhood Social Conditions Mediate the Association Between Physical Deterioration and Mental Health
’,
American Journal of Community Psychology
,
40
:
261
71
.

Kury
,
H.
and
Theodore
,
F
. (
1998
), ‘
The Victim’s Experience and Fear of Crime
’,
International Review of Victimology
,
5
:
93
140
.

Loader
,
I
. (
2008
), ‘
The Anti-Politics of Crime
’,
Theoretical Criminology
,
12
:
399
410
.

Loader
,
I.
and
Sparks
,
R
. (
2016
), ‘
Ideologies and Crime: Political Ideas and the Dynamics of Crime Control
’,
Global Crime
,
17
:
314
.

López
,
T
. (
2014
),
The Winter of Discontent: Myth, Memory and History
.
Liverpool University Press
.

Maguire
,
M
. (
2002
), ‘
Crime Statistics: the ‘Data Explosion’ and its Implications
’, in
M.
Maguire
,
R
.
Morgan
and
R
.
Reiner
, eds,
The Oxford Handbook of Criminology
3rd edition.
Oxford University Press
.

Mannheim
,
K
. (
1928
), ‘
The Problem of Generations
’, in
K.
Mannheim
, ed.,
Essays on the Sociology of Knowledge
.
Routledge
.

Markus
,
G
. (
1983
), ‘
Dynamic Modelling of Cohort Change: The Case of Political Partisanship
’,
American Journal of Political Science
,
27
:
717
39
.

Morgan
,
N
. (
2014
),
The Heroin Epidemic of the 1980s and 1990s and Its Effect on Crime Trends—Then and Now
.
Home Office
.

Neundorf
,
A.
and
Niemi
,
R
. (
2014
), ‘
Beyond Political Socialisation: New Approaches to Age, Period, Cohort Analysis
’,
Electoral Studies
,
33
:
1
6
.

Newburn
,
T
. (
2007
), ‘
Tough on Crime: Penal Policy in England and Wales
’,
Crime and Justice
,
36
:
425
70
.

O’Brien
,
R
. (
2015
),
Age-Period-Cohort Models. Approaches and Analyses with Aggregate Data
.
CRC Press
.

Office of National Statistics
(
2012
),
A Summary of Recorded Crime Data from 1898 to 2001/02
, available online at https://www.gov.uk/government/statistics/historical-crime-data#history (accessed
February 2018
).

Office of National Statistics
(
2013
),
Ward level mid-year population estimates (experimental): Mid 2011 (Census-based)
. Statistical Bulletin. https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/bulletins/wardlevelmidyearpopulationestimatesexperimental/2013-05-30 (accessed
July 2018
).

Pain
,
R
. (
1991
), ‘
Space, Sexual Violence and Social Control: Integrating Geographical and Feminist Analyses of Women’s Fear of Crime
’,
Progress in Human Geography
,
15
:
415
431
.

Pantazis
,
C
. (
2000
), ‘
Fear of Crime, Vulnerability and Poverty: Evidence from the British Crime Survey
’,
The British Journal of Criminology
,
40
:
414
36
.

Paterson
,
L
. (
2008
), ‘
Political Attitudes, Social Participation and Social Mobility: A Longitudinal Analysis
’,
British Journal of Sociology
,
59
:
413
34
.

Pierson
,
P
. (
2004
),
Politics in Time
.
Princeton University Press
.

Riddell
,
P
. (
1985
),
The Thatcher Government
.
Basil Blackwell
.

Rose
,
D.
and
Clear
,
T
. (
1998
), ‘
Incarceration, Social Capital, and Crime: Implications for Social Disorganization Theory
’,
Criminology
,
36
:
441
79
.

Ryder
,
N
. (
1965
), ‘
The Cohort as a Concept in the Study of Social Change
’,
American Sociological Review
,
30
:
843
61
.

Sampson
,
R.
and
Laub
,
J
. (
1993
),
Crime in the Making: Pathways and Turning Points through Life
.
Harvard University Press
.

Sampson
,
R.
and
Raudenbush
,
S
. (
2004
), ‘
Seeing Disorder: Neighborhood Stigma and the Social Construction of “Broken Windows”
’,
Social Psychology Quarterly
,
67
:
319
42
.

Sampson
,
R.
,
Raudenbush
,
S.
and
Earls
,
F
. (
1997
), ‘
Neighbourhoods and Violent Crime: A Multilevel Study of Collective Efficacy
’,
Science
,
277
:
918
24
.

Shaw
,
C.
and
McKay
,
H
. (
1942
),
Juvenile Delinquency and Urban Areas
.
University of Chicago Press
.

Skogan
,
W.
and
Maxfield
,
M
. (
1981
),
Coping with Crime
.
Sage
.

Smets
,
K.
and
Neundorf
,
A
. (
2014
), ‘
The Hierarchies of Age-Period-Cohort Research: Political Context and the Development of Generational Turnout Patterns
’,
Electoral Studies
,
33
:
41
51
.

Steadman-Jones
,
D
. (
2012
),
Masters of the Universe: Hayek, Friedman, and the Birth of Neoliberal Politics
.
Princeton University Press
.

Stiglitz
,
J
. (
2002
),
Globalization and Its Discontents
.
W. W. Norton
.

Tilley
,
J
. (
2002
), ‘
Political Generations and Partisanship in the UK, 1964–1997
Journal of the Royal Statistical Society, Series A (Statistics in Society)
,
165
:
121
35
.

Tilley
,
J.
and
Evans
,
G
. (
2014
), ‘
Ageing and Generational Effects on Vote Choice
’,
Electoral Studies
,
33
:
19
27
.

Tulloch
,
M
. (
2000
), ‘
The Meaning of Age Differences in the Fear of Crime: Combining Quantitative and Qualitative Approaches
’,
British Journal of Criminology
,
40
:
451
67
.

UK Statistics Authority
.(
2014
),
Assessment of Compliance with the Code of Practice for Official Statistics
.
UK Statistics Authority
.

United Nations Committee on the Rights of the Child
.(
2008
),
Forty-Ninth Session: Consideration of Reports Submitted by States Parties Under Article 44 of the Convention – UK and Northern Ireland
.
United Nations
.

Visser
,
M.
,
Scholte
,
M.
and
Scheepers
,
P
. (
2013
), ‘
Fear of Crime and Feelings of Unsafety in European Countries: Macro and Micro Explanations in Cross-National Perspective
’,
Sociological Quarterly
,
54
:
279
301
.

Warr
,
M
. (
1990
), ‘
Dangerous Situations: Social Context and Fear of Victimisation
’,
Social Forces
,
68
:
891
907
.

Wilson
,
J.
and
Kelling
,
G
. (
1982
), ‘
Broken Windows
’,
Atlantic Monthly
, March:
29
38
.

Yang
,
Y.
and
Land
,
K
. (
2013
),
Age-Period-Cohort Analysis: New Models, Methods and Empirical Applications
.
Chapman and Hall/CRC

Zhao
,
J. S.
,
Lawton
,
B.
and
Longmire
,
D
. (
2015
), ‘
An Examination of the Micro-Level Crime-Fear Link
’,
Crime and Delinquency
,
61
:
19
44
.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.