Introduction

The international research on fear of crime has proliferated impressively but is associated with measurement problems and conceptual and methodological ambiguity. In much scholarly work on fear of crime, the methodological convention for quantifying fear of crime is asking study participants how safe they feel in their residential area alone in the evening or when it is dark. This formulation has been criticized for not mentioning crime or fear and for defining the subject of the research as something happening when someone is alone, outdoors and when it is dark. When asked about perceived fear outdoors, research generally shows that women tend to be more insecure than men and that younger and older people are more insecure than middle-aged people (Hale, 1996; Heber, 2007). Fear of crime research has been the subject of extensive and legitimate criticism for methodological, theoretical and conceptual reasons (Nair, Ditton and Phillips, 1993; Hale, 1996; Farrall et al., 1997, 2012; Gilchrist et al., 1998; Short and Ditton, 1998; Ditton, Bannister, et al., 1999; Ditton, Farrall, et al., 1999; Farrall and Ditton, 1999; Lee, 1999, 2001, 2013; Ditton, 2000; Ditton et al., 2000, 2004; Farrall, 2004; Farrall and Gadd, 2004; Chadee and Ditton, 2005; Chadee, Austen and Ditton, 2007; Gray et al., 2008; Farrall, Godfrey and Cox, 2009; Farrall, Jackson and Gray, 2009a, 2009b; Jennings et al., 2015; Lee et al., 2020). There have been discussions among researchers on the validity of the survey as a tool for measuring emotions. The concept of fear of crime can be understood as a collection of cognitive processes, subjective and ill-suited to be measured with operationalizations such as “are you afraid to walk around in your neighbourhood after dark?” There has been concern that quantitative studies significantly overestimate fear of crime and are of questionable validity (Hale, 1996; Farrall et al., 1997; Farrall and Gadd, 2004; Farrall, Jackson and Gray, 2009b; Jackson, 2009). In studies with qualitative design, fear of crime tends to be described not as a statistical quantifiable phenomenon with serious implications in people’s lives, but rather transient, temporary, occasional and spatially definite (Koskela, 1999a, 1999b; Koskela and Pain, 2000; Heber, 2007). This criticism can perhaps be summarized by Hale's (1996) statement: "Fear of crime is a product of how it has been researched, rather than how it is".

Curiously, this lack of scientific validity has not hindered extensive institutionalization across different national contexts. This paper attempts to shed some light over this paradox, through asking what the role of fear of crime research plays in the conception and framing of the crime problem. By unpacking the theoretical assumptions inherent in fear of crime methodology and examining their function and through discussing the empirical example of Sweden in relation to accounts on fear of crime in the USA and UK, this paper aims to fulfil a twofold aim. First, it aims to provide an empirical account of how fear of crime research was implemented and institutionalized in Sweden, and how the occurrence of governmental fear of crime surveys has developed over time. Secondly, it presents a discussion on the role that fear of crime discourse plays in the politics of punitivity. This paper concludes that fear of crime research rhetorically enables continued political exploitation of crime, constructs crime as a problem of order maintenance and legitimizes further expansions of state control.

Methods

The matter of government engagement in fear of crime research is addressed using government documents and a survey. The documents were generally obtained through the websites of government agencies and through email correspondence with the concerned agencies, and include reports, questionnaires, technical reports and decision-making protocols. In addition to the survey,Footnote 1 data were obtained via email and phone conversations with municipal representatives and representatives of Swedish governmental agencies.

Emergence and Expansion of the Swedish Fear of Crime Research Discourse

The speed and intensity of the expansion of fear of crime research and how it managed to penetrate many segments of Swedish governance must be considered remarkable. The research discourse began humbly in Sweden with the addition of a general fear of crime indicator in the survey of Swedish living conditions in 1978. Fear of crime research in Sweden took a major step forward through the first specialized fear of crime and victimization survey, The Stockholm Project. The study collected data in 1989 and published results in 1990 (Wikström, 1990) and was theoretically and methodologically inspired by Skogan and Maxfield (1981). A period of specialized fear of crime and victimization studies followed through the 1990s, implemented by either the Sweden’s National Crime Council, or the Research group of the Police, in collaboration with Swedish criminologists. An important motivation behind these was the altered phrasing of the mission of the Swedish police, changed to "decrease crime and increase safety (trygghet)" in 1992. Still, in 1995 fear of crime research was rare in Sweden. The average number of fear of crime measurements per municipality was less than 0.1 during the period of 1995–1999, and 94% of municipalities did not conduct any form of fear of crime measurement, according to the survey of Swedish municipalities.

Figure 1 depicts the development of the total number of survey participations in municipalities over time. Ranging from 1995 to 2018, there is an exponential increase, from almost zero survey measurements to almost five hundred surveys in the Swedish municipalities during the last period. Figure 2 depicts the average number of fear of crime measurements per municipality, also increasing exponentially during the period 1995–2018. Figure 3 depicts the percentage of municipalities without measurements of fear of crime. This graph shows a similar tendency compared to previous figures; very few municipalities did fear of crime measurements during the earlier periods, but only 16% said they do not measure fear of crime during the last period of 2015–2018.

Fig. 1
figure 1

Total number of fear of crime measurements per period in Swedish municipalities

Full size image
Fig. 2
figure 2

Average number of fear of crime measurements per period in Swedish municipalities

Full size image
Fig. 3
figure 3

Percentage of municipalities without fear of crime measurements, per period

Full size image

While the accumulation of fear of crime surveys began during the 1990s in the municipalities, and in 1978 in terms of national fear of crime surveys, the period of most rapid development clearly took place during the 2000s. Figure 4 illustrates the establishment of national fear of crime surveys over time; The Survey of Living Conditions in 1978, Local Youth Politics Survey in 2003, The National Public Health survey in 2004, The Citizen Survey in 2005, The Swedish Crime Survey in 2006 and The Swedish Contingencies Agency Survey in 2007 (Sahlin Lilja, 2021). Most of the surveys do not have fear of crime as a main theme, but all contain fear of crime indicators. From one national-level government survey containing a fear of crime indicator (ULF) in 2001, the number of surveys increased by one per year until 2007, when six surveys collected data on fear of crime. The rate has been more or less constant since then, with minor fluctuations caused by semi-annual surveys. This period also saw the introduction of the Swedish Crime Survey (NTU), arguably the most influential of them all in terms of putting "otrygghet" Footnote 2 on the agenda.

Fig. 4
figure 4

Number of national governmental fear of crime surveys per year

Full size image

On the local level, interest and engagement in the fear of crime issue also accelerated during the 2000s. During the earlier period of 2000–2004, local fear of crime measurements were still somewhat uncommon, with an average number of fear of crime measurements per municipality at 0.4 and a total number of 74 across all municipalities. Only 26% of municipalities measured fear of crime during this period. By 2005–2009, there is an expansion, as the average increased to 0.9, 168 measurements are reported, and the percentage of municipalities participating increased to 49%. This may be seen as a kind of tipping point, where measuring fear of crime at the local level is becoming the norm.

Additionally, 89 percent of municipalities do some form of nonquantitative measurement of fear of crime. These include " fear of crime walks" which is a method developed by the Swedish Crime Council (BRÅ, 2010c). The intended purpose is for a group consisting of municipality workers and people living in the area to examine the local milieu from a fear of crime perspective. The guide suggests engaging “municipality administration, politicians, landlords, renter’s associations, police, local entrepreneurs, local associations and other actors” (BRÅ, 2010c: p. 9), and “citizen dialogues", in order to improve relations between citizens and local government, to gain legitimacy, provide more efficient services and improve knowledge (SKL, 2013). Aside from the surveys and inquiries into fear of crime among citizens, another indication of the saliency of fear of crime in local government is the professions it has generated, such as “brottsförebyggare” (crime preventers) and “trygghetssamordnare” (safety coordinators). Almost all municipalities except for the very smallest seem to have staff working solely with fear of crime issues, which must be considered a recent development.

During the 2010s, there were five national-level government surveys measuring fear of crime, and regional and local fear of crime projects at all levels of governance, not to mention private enterprises. The average number of fear of crime measurements in the municipalities surveyed was 1.55 by 2010–2014 and 2.4 for the last examined period, 2015–2018. The total number of reported fear of crime measurements on the local level is up to 301 in 2010–2014 and, for the last period of 2015–2018, the count is 469. The percentage of participating municipalities increased as well, from 70% in 2010–2014 up to 84% in 2015–2018. At the end of the examined period of 1978–2018, the range and wealth of examples of the salience of “otrygghet” in contemporary Swedish society is impressive. Measurements of fear of crime have increased, and there are now many participating actors from both governmental and private sectors of Swedish society. The research discourse has grown in scope and become integrated in other forms of political and social life, as an increasingly potent symbol (Hermansson, 2018, 2019; Sahlin Lilja, 2018, 2021).

Results and Response Rates of Swedish National Fear of Crime Surveys

Five of the government agency surveys contain fear of crime indicators that are presented as a percentage of fearful participants. The results from these surveys—the Survey of Living Conditions, the Swedish Crime Survey, the Local Youth Politics Survey, and the National Public Health Survey—are shown in Fig. 5. The legend displays the different surveys and indicators that resulted in that percentage; how many were considered otrygga, according to each survey and indicator. First, change over time for each specific survey is rather modest compared to the discrepancy between surveys. Only the Survey of Living Conditions offers data from before the new millennium since most of the surveys began to collect data sometime in the early 2000s. The curve during the 2000s is slightly U-shaped but rather flat in most of the surveys, depicting an initial small decrease and an increase during the last 2–4 years. Most depict modest variance over time, but the Swedish Contingencies Agency Survey is an exception; between 2007 and 2010 people became a lot more fearful according to this study.

Fig. 5
figure 5

Results and indicators from Swedish Governmental Fear of Crime Surveys. sources: (Folkhälsomyndigheten and SCB, 2004, 2005, 2006, 2007, 2008, 2009, 2010a, 2010b, 2011, 2012, 2013, 2014, 2015, 2016, 2018a, 2018b; Brottsförebyggande Rådet, 2007; BRÅ, 2007, 2008a, 2008b, 2009a, 2009b, 2010a, 2010b, 2022a2011a, 2011b, 2012a, 2012b, 2013a, 2013b, 2014a, 2014b, 2015a, 2015b, 2016a, 2016b, 2016c, 2017, 2018a, 2018b, 2019a, 2019b, 2019c, 2020a, 2020b, 2020c, 2021a, 2021b, 2022a, 2022b; MUCF, 2015a, 2015b, 2016a, 2016b, 2016c, 2016d, 2016e, 2017a, 2017b; Folkhälsomyndigheten, 2018a, 2018b, 2019, 2023; SCB, 2019; Undersökningarna av levnadsförhållanden (ULF), 2021)

Full size image

Secondly, the discrepancy between the different surveys and indicators is often large. This means that for any given year when multiple surveys have been implemented, several answers to the question of how “otrygga” people are can be considered scientifically correct. What is being used and reported as a general fear of crime indicator in different surveys results in different answers to that question. Take the year 2017 as an example: 15 percent of respondents in the Survey of Living Conditions, 19 percent in the Swedish Crime Survey, and only 6 percent of youths aged 13–19 years participating in the Local Youth Politics Survey are otrygga. The National Public Health survey was not implemented in 2017, but according to the 2018 version, 28 percent are otrygga, while according to the Swedish Contingencies Agency Survey 63 percent were fearful in 2010 (Fig. 6).

Fig. 6
figure 6

Nonresponse rates in Swedish fear of crime surveys. Sources: (BRÅ, 2007, 2008b, 2009b, 2010b, 2011b, 2012b, 2013b, 2014b, 2015b, 2016b, 2017, 2018b, 2019c, 2019d; BRÅ & SCB, 2017; Folkhälsomyndigheten & SCB, 2004, 2005, 2006, 2007, 2008, 2009, 2010a, 201b, 2011, 2012, 2013, 2014, 2015, 2016, 2018b, 2018c; MSB & SCB, 2007a, 2007b, 2010a, 2010b, 2014a, 2014b; SCB, 2010, 2018a, 2018b, 2018c, 2019a, 2019c)

Full size image

Response rates are falling over time. For surveys that were launched in the early 2000s, the response rates have decreased from around 60 percent to around 40 percent for the Citizen Survey, the Public Health Survey, and the Swedish Contingencies Agency Survey. For these surveys, the decline is steady, and a 50/50 equilibrium was reached around 2012–2014. For the Swedish Crime Survey, the development is more dramatic, with decreasing response rates from an impressive 77 percent in 2006 to 37 percent in 2018. The Survey of Living Conditions had a remarkable response rate of 87 percent in the early 1980s. Since then, it has been decreasing and even though it is still slightly higher than in the other surveys, it has fallen to a much less remarkable 50 percent in recent years. The decline was slow during the 1980s and 1990s, and more rapid during the 2000s. As the response rate has gone down, the number of selected participants has gone up. This is most dramatically evident in the Swedish Crime Survey, where the changes implemented in the NTS-lokal included ten times more selected participants, from 20,000 to 200,000, but also a much lower response rate. It should be noted that when the survey increased the number of participants and much fewer responded, the survey recorded a much higher level of fear of crime.

Discussion

A peculiar feature of the fear of crime research discourse and its rapid expansion in Sweden is that it has coincided with a period of stable or decreasing crime levels. The period of the first tentative specialized fear of crime surveys, the mid 1990s, saw a turning point away from the tendency of increasing crime rates that had been ongoing since the post-war period. The period of most rapid expansion of fear of crime research was the 2000s for governmental surveys and the 2010s for municipal fear of crime measurements. Crime levels in Sweden, as well as in the USA and UK, have been stable or decreasing for most the 21th century so far (von Hofer, 2011; Kivivuori, 2014; Tonry, 2014; Nilsson et al., 2017; Lehti et al., 2019; Aebi, 2021). Though crime is not increasing, analysis of the development of Swedish penal policy shows since the year of 1976, the vast majority of policy changes have increased the scope of the judicial sphere in terms of criminalization, sentencing and discretionary powers of the police (Tham, 2018a, 2018b). The conclusion can be drawn that Sweden has experienced a turn towards punitivity (Tham, 2018a).

Two main explanations for why a punitive turn has occurred in the west have been offered. The first, arguing that the punitive turn was a natural cause of rising crime rates has largely been debunked, by crime trajectory analysis showing that it occurred largely during the periods of stable or decreasing crime rates (Chambliss, 1994; Tonry, 2014). The second explanation asks what political and ideological role a particular conception of crime plays in the public debate, and thus focuses on the discourse of crime rather than actual crime rates (Estrada 2004). Demker and Duus-Otterström (2009), aiming to fulfil Tonrys request for fuller accounts of the political histories of countries, added an excellent analysis of the role of political actors in the developments in criminal policy in Sweden, through asking if both the Swedish public and Swedish crime policy has become more punitive, what occurred first? Through looking at political parties and their crime politics, they conclude that much of the shift in Swedish criminal policy can be explained by the increasingly victim-centred discourse on crime and criminality, and has been largely initiated by Moderaterna, for whom the rise of the crime victim represented a window of opportunity (Demker and Duus-Otterström, 2009). Demker and Duus-Otterström (2009) propose a causal chain where individualization leads to victimization of criminal policy, which in turn causes punitivity and fear of crime.

But accounts of fear of crime research’s origins have stressed how fear of crime itself originated in a political context, initiated by political actors, in order to further a specific understanding of crime (Lee, 2013). The fear of crime discourse represented a window of opportunity in itself. Through importing a discourse on how to produce fear of crime knowledge derived from an American context, the political actors interested in furthering the punitive turn in Sweden received an especially potent instrument. The next section will discuss how the specific theoretical and methodological assumptions of fear of crime research construct a conception of the crime problem that characterizes punitivity. Estrada (2004) uses Garland’s (2001) framework to analyse trends in crime and punitivity in Sweden and describes five defining features of late modern punitivity: a loss of ground by the treatment ideology, the rediscovery of the prison, the focus is on the victim, increased political exploitation of the crime problem and individualization. While the first two mainly has to do with societal reactions to crime, fear of crime research and discourse play a crucial part for the last three aspects: the increased victim centredness, political exploitation of crime and an individualized understanding of the causes of crime.

Fear of Crime Research Constructs a Conception of Crime as a Matter of Order Maintenance

Fear of crime research rests on theoretical assumptions largely derived from “Signs of Disorder” theory. The earliest Swedish fear of crime studies cite this theory (Wikström, 1990, 1991), and police officers involved with the implementation of early Swedish fear of crime research have described it as the only available theory in the contemporary discourse (Sahlin Lilja, 2021). Its assumptions shape what knowledge can be produced, which is visible for example in the specific examples of problems that are possible to choose for respondents of the first Swedish specialized fear of crime survey: littering, drunken people, fighting and violence outside, young people fighting or making trouble, women or children being harassed, homes for addicts, disruptive children and disruptive neighbours (Wikström, 1990: pp. 210–232). The authors of these early fear of crime studies write that a “lack of reaction” from society will lead to a downwards spiral (Wikström, Torstensson, & Dolmèn, 1997). The fear of crime studies organized by the police also inherited its methodology from the same American origins and are also explicit about its theoretical foundations:

“The theory can in simple terms be explained as an attempt to analyze levels of crime and disorder, and its consequences in accordance with the so-called ‘zero tolerance’ theory. The fear of crime survey is based on this hypothesis.”Footnote 3

The fear of crime research discourse frames the problem of crime as coming from outside through defining what kinds of crimes are fear-inducing and where they happen. The convention of operationalizing fear of crime by asking individuals how they feel alone in their neighbourhood after dark defines fear-inducing crime as crime committed by strangers, when the respondent is alone and away from home. We can note here that it is well known that women are most often victimized in their own homes, or the home of a man she knows (Jansson, 2007; Lehti et al., 2019). The assumption that women are safe in the home is an ancient patriarchal construction.

The dominance of this conventional operationalization defines fear-inducing crime as street crimes; an old problem of criminology, discussed for example by Hagan (2010). Wilson (1975) argues that his influential book, Thinking about Crime, will not deal with any victim-less crime nor any white-collar crime, and motivates this by saying that the police spend little to no time on these crimes, and neither will he. Wilson focuses on “predatory crime”: street crime committed by strangers. This theoretical inheritance that insists on focusing on street crime is evident in Swedish fear of crime surveys in several ways. The home is generally assumed to be a safe place, according to fear of crime survey operationalizations in Sweden. In addition to asking if the respondents feel safe outside of the home during night-time, other questions specifically ask how afraid they are of the crimes that constitute classic “street crimes”, namely robbery, assault and vandalism. The police survey specifically asks the respondent how afraid she is of being assaulted outside the home. If fear is thought to exist in the home, it is assumed to come from outside, for example, as seen in how the “Citizen survey” asks about respondents’ fear of being burglarized.Footnote 4 In addition, the “Police Survey” contains questions on what problems exist in your area and the possible answers are limited to the following groups: gangs of youth, drunk people and drug addicts. In addition to the Ennis operationalization, fear of crime is measured through asking about worry about burglary in the home or the garage, vandalization or theft of a vehicle and worry about being assaulted outside in the neighbourhood. The surveys often assume that the participant owns and worries about property, such as cars, bikes and motorcycles. There is a boundary drawn between what is owned and private—and must be protected—and the threatening public sphere.

I point here to a careful synergy between how right-wing criminological theory influences the instruments of measurement and the conception of the crime problem that the knowledge produced by these instruments construct. It is an understanding of crime and fear of crime as a problem of order maintenance and control, caused by using surveys that operationalize theory that rests on this assumption. It omits structural explanations and causes of crime by design. Indeed, Kelling, an author of Broken Windows theory, has called such an understanding of crime “disastrous” as it risks “de-policing” the issue of crime (Bratton and Kelling, 2015). It furthers an individualized understandings of crime, characteristic of punitivity, where causal discussion has come to focus less on social mechanisms and more on individual level factors, and crime policy focuses on immediate control and security measures (Estrada, 2004).

Fear of Crime Rhetorically Enables Continued Political Exploitation of Crime–No Matter the Crime Rates

A seemingly paradoxical aspect of the fear of crime discourse in Sweden is that it was largely implemented during a period of stable or decreasing crime trends. Tonry (2014) remarked that it is strange that only a few criminologists have paid attention to the fact that crime trends are falling throughout the western world. Hall writes that statistics—whether crime rates or opinion polls—have an ideological function: they appear to ground free floating and controversial impressions in the hard, incontrovertible soil of numbers (Hall et al., 2013/1978: p. 13). But what if the statistics do not show what they “should”? Andersson (2002) shows how when crime statistics diverged from the expected increasing trend and showed a levelling-off during the 1970s, they were considered less reliable and valid by Swedish politicians. In Andersson’s example the politicians question the validity of crime statistics by arguing that criminology cannot accurately measure the “dark figure” of crime, and by arguing that any crime is too much.

Through the emergence of fear of crime research a third strategy is made viable: crime is a problem because it makes people afraid. As long as there is fear of crime, crime must be fought. It removes the issue of crime from the measurable. A continual production of statistics of a percentage of the population that is afraid of crime rhetorically enables politicians to exploit the crime problem effectively even during periods of stable or decreasing crime trends. Let us examine how a bill from the Moderates uses fear of crime measurements rather than crime statistics to argue for a conception of crime as an existential threat:

Motion 2021/22:4031 av Ulf Kristersson m.fl. (M) Swedish original

Motion 2021/22:4031 av Ulf Kristersson m.fl. (M) English translation

Brottslighetsutvecklingen i Sverige kan bara beskrivas som extrem. På relativt kort tid har vi gått från ett av världens tryggaste samhällen, präglat av konsensus och låg konfliktnivå, till ett högkonfliktsamhälle. Detta utgör enligt vår analys för närvarande det allra allvarligaste samhällsproblem som politiken måste hantera. Utvecklingen visar sig bland annat i ökande otrygghet i princip i alla delar av samhället. Enligt Brottsförebyggande rådets (Brå) nationella trygghetsundersökning 2020 är den upplevda otryggheten i befolkningen större än vid något tidigare mättillfälle. En tredjedel av befolkningen drar sig för att gå ut på kvällen i sitt eget bostadsområde

The crime trend in Sweden can only be described as extreme. In a relatively short time we have gone from one of the world's safest societies, characterized by consensus and low conflict level, to a high-conflict society. This, according to our analysis, currently constitutes the most serious social problem that politics has to deal with. This development is manifested, among other things, in increasing fear of crime in basically all of society. According to the Swedish National Crime Survey of the Swedish Council for Crime Prevention (Brå) In 2020, the perceived fear of crime in the population is greater than in any previous measurement. A third of the population are reluctant to go out in the evening in their own residential area

Crime reaches its potential as a social problem only in crisis. Legitimacy for expansion of tools of control at the state’s disposal cannot be won by the type of discourse on crime that used to be the Swedish midcentury norm, which was expert-focused, measured and technological (Nilsson et al., 2017; Tham & von Hofer, 2014). In the example above, crime is described as a crisis, as growing, as extreme and in rapid change. While Sweden has experienced an increased level of shootings connected to organized crime, most categories of crime have been stable or decreasing. Fear of crime surveys reliably produce suitably shocking statistics on the severity of the crime problem. Examining the results from Swedish fear of crime surveys we can note that the percentage of fearful people resulting from different surveys during the same period can differ to a significant degree. Response rates are falling. The surveys could all be subjected to the same criticism that has been levelled at the research discourse of fear of crime historically; its lacking in conceptualization, theory and validity of measurements (Hale, 1996; Ditton et al., 2000; Lee, 2013). Curiously, in Sweden, as in the USA (Hale, 1996), or the UK (Farrall et al., 1997; Ditton et al., 2000), this lack of scientific validity never seems to hinder fear of crime research to emerge, expand and become institutionalized to an impressive degree. Perhaps the lack of validity should be understood as an explanation for its great political appeal and popularity. The research discourse reliably generates scientific knowledge (or, at least knowledge produced loosely according to the principles and methods derived from social science) that enables continual political exploitation of the crime problem. A percentage of people will be found fearful of crime no matter what crime trajectory analysis shows.

The Paradoxes of Fear of Crime

There is something nonsensical about fear of crime research. A cursory glance at the charts depicting the growth of fear of crime measurements, and the development of fear of crime indicators gives cause to dismiss the explanation that the increased engagement in the problem of fear of crime actually has made us feel safer. Even though fear of crime as a problem is widely acknowledged and fought across various parts of Swedish society, fear of crime does not seem to be decreasing. This conclusion has been reached before, for example by Ditton et al. as they analysed fear of crime research and its implementation in Great Britain:

The briefest possible history of the ‘fear’ of crime would be this. There was no ‘fear’ of crime in Britain until it was discovered in 1982. Crime surveyors liked it because whereas only about five interviewees in 100 could recall a crime victimization from the recall period (usually the previous year), 100 out of 100 could give usable data about their ‘fear’ of crime. Politicians and policymakers liked it because it seemed more amenable to manipulation and reduction than crime itself. Rates of ‘fear’ of crime, at one point, seemed about to become more important than rates of crime itself. People set about energetically trying to reduce the ‘fear’ of crime. They failed. (Ditton et al., 2000: p. 3).

Lee considers the question itself—if fear of crime research was ever intended to make us safer—to be naive. Lee shows how the very origin of fear of crime research is to be found in the need for an ideological project for the American Republicans.

My emphasis here is that one of the main foci of the New Right neo-conservative law and order political campaigns was precisely the object of investigation first “discovered” empirically in the administrative crime surveys: the fear of crime. Were it not for this “discovery,” the New Right campaigns would have been unable to engage in the same forms of populism that we have witnessed. Claims of the existence of a fearing population could now be backed up by statistical “proof” (Lee, 1999: p. 238).

The Implementation of Fear of Crime as a Neo-liberal Reform Project

The 45 years that have passed since fear of crime research began in Sweden has also been a period of intense economic change and restructuring. Economic inequality has increased, both in terms of ownership, with capital becoming increasingly concentrated, and in terms of income inequality, which is increasing (Therborn, 2018, 2020). Sweden has an international reputation as an egalitarian country: as social democratic, welfarist and equal. Indeed, in 1980, Sweden was ranked the least unequal country in the world in terms of income (Alvaredo et al., 2018). From this high watermark, Sweden has experienced radically increasing economic disparity which, today, puts it on par with the USA in terms of economic inequality. In 2017, the richest 1 percent in Sweden owned 37 percent of all private wealth, and the richest 10 percent owned 75 percent, while in the USA, the richest 1 percent owned 35 percent and the richest 10 percent owned 76 percent (Waldenström, Bastani and Hansson, 2018; Therborn, 2020). Overall wage inequality fell from the late 1960s to 1975, then stabilized until 1980 before resuming growth (Korpi and Tåhlin, 2011), and most of the growing inequality is explained by capital gains (Roine and Waldenström, 2012). The period of the most intensive economic restructuring was the 1990s, coincidentally (?) when fear of crime was introduced as a social problem in a Swedish context.

There is something to be said for the way fear of crime seems to crop up as an issue, across different national contexts, during periods of intense neo-liberal reformation of the economy following perceived crises of the welfare state. Lee (2001) points out how fear of crime was used discursively in the US congress to argue for increased punitivity already in 1969, and how it influenced the rhetoric of Nixon. Issues of the welfare state, growing crime rates and black insurrection ended up building rhetorical support for Nixon’s neo-liberal project. According to Lee, the new survey techniques of social science, the quantification of fear of crime and the resulting political discourse formed a fundamental part of this political project (Lee, 2001). In the UK, the way crime was exploited as a political issue during the 1970s economic crisis has been chronicled by Hall et al. (2013). The following 1980s, described in Ditton et al. (2000) as a "golden age for fear of crime research", were a period of rapid political, economic and social change in the UK from which the British welfare state never recovered. Margaret Thatcher’s political legacy is one of social conservatism combined with economic neoliberalism. Farrall (2006) summarizes her government’s policies and their various effects: deregulation of housing, rising unemployment, a general scaling back of the social responsibilities of the state. Others have remarked that the main result of her policies was to have “closed down the escape hatches from poverty” (Andrews and Jacobs, 1990), and have pointed out the criminogenic nature of many of these reforms (Riddell, 1985; Farrall, 2006). Thatcher’s policies promoted continued use of prisons, invested in new ones and disapproved of social intervention as a crime fighting policy; Thatcher said about social workers, “they create a fog of excuses in which the muggers and burglars operate.” Researchers have pointed out that the rising crime levels of the period helped sell a narrative of the failures of rehabilitation and social work (Farrall, 2006).

Political exploitation of crime in order to discredit the welfare state and legitimate further policy change in a punitive direction seems to be an enduring trait of the neo-liberal political project. Also, in Sweden, Demker and Duus-Otterström (200) consistently find that Moderaterna, historically the largest right-wing party of Sweden, has been working to move discourse on crime in Sweden in a more punitive direction:

When looking at the origins of the tougher policy embraced today by the parties to the right at least, it seems as if they must be sought in the 1980s, especially in Moderaterna’ s writings on criminal policy. It was they who more than others seem to have introduced what we may call an ‘alarmist’ attitude to crime: it was they who portrayed crime as a rampant social problem without political solutions (Demker and Duus-Otterström, 2009: pp. 2988–289).

Demker and Duus-Otterström find something interesting that causes them to reevaluate victim centredness as a driving cause. They find Moderaterna expressed interest in becoming tougher on crime before they expressed concern about crime victims. Their initial interest was in sharper sentencing, first expressed in 1969—during the heyday of the rehabilitative ideology. Sharper sentencing, however freely spoken about in today’s rhetoric on crime in Sweden, was unpalatable and unmodern then. Lee remarks on how social–scientific ideals, statistics and the social–democratic will to be a knowledgeable society contributed to a conservative discourse on crime policy, and the emergence of fear of crime a subject of governmental inquiry. Sweden should perhaps be considered a country more entrenched in that social–democratic will to be a knowledgeable society than most. As concluding remarks, I would like to argue for an understanding of fear of crime, not as a scientific object to be discovered or a social problem to be solved, but as a very specific discursive instrument used to institutionalize a conservative understanding of crime as a problem of order maintenance, perhaps extra malicious in that it omits and makes invisible the structural causes of crime, and thus makes social progress increasingly difficult.