The Role of Radical Economic Restructuring in Truancy from School and Engagement in Crime

Abstract

Of late, criminologists have become acutely aware of the relationship between school outcomes and engagement in crime as an adult. This phenomenon—which has come to be known as the ‘school-to-prison-pipeline’—has been studied in North America and the United Kingdom, and requires longitudinal data sets. Typically, these studies approach the phenomenon from an individualist perspective and examine truancy in terms of the truants’ attitudes, academic achievement or their home life. What remains unclear, however, is a consideration of (1) how macro-level social and economic processes may influence the incidence of truancy, and (2) how structural processes fluctuate over time, and in so doing produce variations in truancy rates or the causal processes associated with truancy. Using longitudinal data from two birth cohort studies, we empirically address these blind spots and test the role of social-structural processes in truancy, and how these may change over time.

Full text

*Stephen Far rall, Depart ment of Criminolog y, College of Business, Law and the S ocial S ciences, University of Derby,
Derby, England; School of Law, University of Shefeld, Shefeld, England; S.Farrall@derby.ac.uk; Emily Gray, Depar tment
of Criminolog y, College of Business, L aw and the S ocial S ciences, University of Derby, Derby, England; Philip Mike Jones,
Department of Criminolog y, College of Busines s, Law and t he Socia l Sciences, Universit y of Derby, Derby, England.
© The Author(s) 2019. Published by Oxford University Press on behalf of the Centre for Crime and Justice Studies (ISTD).
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.
THE ROLE OF RADICAL ECONOMIC RESTRUCTURING IN
TRUANCY FROM SCHOOL AND ENGAGEMENT INCRIME
S F*, E G and P MJ
Of late, criminologists have become acutely aware of the relationship between school outcomes and
engagement in crime as an adult. This phenomenon—which has come to be known as the ‘school-
to-prison-pipeline’—has been studied in North America and the United Kingdom, and requires
longitudinal data sets. Typically, these studies approach the phenomenon from an individualist
perspective and examine truancy in terms of the truants’ attitudes, academic achievement or their
home life. What remains unclear, however, is a consideration of (1) how macro-level social and eco-
nomic processes may inuence the incidence of truancy, and (2) how structural processes uctuate
over time, and in so doing produce variations in truancy rates or the causal processes associated
with truancy. Using longitudinal data from two birth cohort studies, we empirically address these
blind spots and test the role of social-structural processes in truancy, and how these may change
over time.
Keywords: truancy, school, offending, anomie, life-course studies, school-to-prison-
pipeline, National Child Development Study, British Cohort Study 1970
Introduction
In recent years, scholars, especially those in North America, have shown a renewed
interest in what is now referred to as the ‘school-to-prison-pipeline’ (Rocque etal. 2017).
This phenomenon describes the ways in which schools have become a conduit to the
youth and criminal justice systems, whereby those children who do poorly at school,
who truant, feel excluded or who are expelled, or who drop out completely, will often
end up enmeshed in the juvenile correctional system and later the adult prison system.
In fairness, few studies actually study both schooling and imprisonment; most con-
tent themselves, as we do, by exploring truancy and later engagement in crime. Many
studies rely on individual-level factors and processes to account for truancy. Herein
we explore the role played by economic restructuring in triggering alienation from
school, truancy and offending, thereby challenging the general accounts which tend to
pathologize those young people who truant. One of our aims, then, was to disrupt the
mainstream account of neo-liberal criminology which focuses on the individual and
suppresses any consideration of wider structural processes and the role of governments
in shaping which communities and their members are affected by crime. Building on
insights from strain theory, and employing two longitudinal studies of people born in
1958 and 1970, we build a structural equation model to model economic restructuring,
doi:10.1093/bjc/azz040 BRIT. J. CRIMINOL. (2020) 60, 118 –140
Advance Access publication 28 July 2019
118
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

alienation from school and offending. The remainder of our article is constructed as
follows: rst we outline what is known about truancy and later life outcomes (including
offending). We then critique this literature, arguing that it tends to focus on individual-
level factors. Having outlined our research design, we discuss our theoretical model.
Drawing on strain theory, we treat truanting as a result of anomic feelings towards
formal institutions (such as schools and the labour market). We then locate our study
within the wider social and economic changes of the 1980s in the United Kingdom.
Data from the two birth cohorts we rely on are then examined and contrasted, before
we develop a theoretically informed model of truancy that incorporates socio-economic
and political forces and anomic reactions towards formal institutions (such as schools
and the labour market). This model we test using structural equation modelling, using
longitudinal data from the National Child Development Study (NCDS) and 1970 Birth
Cohort Study (BCS70). We end by reecting upon what our ndings contribute to the
scholarship on truancy and the ‘school-to-prison-pipeline’.
What Do We Know About the Relationship Between Truancy and Later Offending?
Research into the relationship between truancy (when a child elects not to attend some
or all classes for a day or more) and later offending is not new in criminology. In the
middle of the twentieth century, studies in the United States by Shaw and McKay (1942),
Glueck and Glueck (1950) and Reiss (1951) all reported associations between truancy,
and delinquency or offending and recidivism at a later age. It is accepted that truancy
and offending are not directly related to one another in an immediate causal way, but
may lead indirectly to offending in adulthood. Garr y (1996), e.g. argues that truancy
is a ‘gate-way’ into later delinquency. Truants are more likely to use drugs, consume al-
cohol and become involved in violent activities (Rocque etal. 2017:596), and are likely
to engage in early sexual activities and gang membership (Dryfoos 199 0). However,
there is uncertainty about the causal ordering; truancy may lead to drug use and delin-
quency, whereas these may help to encourage and reinforce truancy. As Rocque etal.
(2017:593) note, it is still surprising that there have not been more studies of the rela-
tionship between truancy and offending in later life. Current thinking is that truancy
will lead to offending via a series of ‘stepping-stones’; events and processes which leave
the individuals involved more likely to commit offences as anadult.
Various characteristics have been found to be associated with truancy itself. At the
individual level, truants are usually found to be more likely to be male (Garr y 1996).
They are also more likely to dislike school (Attwood and Croll 2006) and to have
achieved fewer qualications (Farrington 1980; Vaughn et al. 2013). Those with low
non-verbal IQ, ‘daring’ attitudes and who were troublesome were also more likely to
truant (Far rington 1996). Familial processes have also been found to be associated with
truancy; those from lower income groups (Attwood and Croll 2006), whose parents
give them less attention (Fa rrington 1980), have conictual relationships, are disinter-
ested in education (Farrington 198 0), or who live in disadvantaged neighbourhoods
(Farr ington 1980) are more likely to truant than others. Moreover, having a sibling who
had behavioural problems or separation from a parent was also associated with truancy
(Farr ington 19 96). Some school factors have also been shown to be related to truancy,
with large school sizes, failure to motivate pupils and poor attendance policies having
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
119
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

greater rates of truancy (Strand and Lovrich 2014). Reid’s (2008:337–338) overview of
what is known about truancy suggests that styles of school leadership, poorly managed
school-to-school transitions, poor or absent pastoral support, poor, or non-existent
attendance-monitoring, and the children’s exclusion from school decision-making pro-
cedures all contributed to truancy. Educational support and early interventions can
reduce truancy (Strand and Lovrich 2014).
In the United Kingdom, one of the few longitudinal studies that investigated the
relationship between truancy from school and offending as an adult has been the
Cambridge Study in Delinquent Development (CSDD). Farrington (1980) found that
truancy (measured by self- and teacher-reported data) was associated with later nega-
tive life outcomes such as a low status job at 18, smoking at 18 and involvement in crime
at 18. At age 32, truancy at school was associated with numerous negative outcomes,
including offending (Farrington 1996). Rocque et al., using the CSDD, extended this
work up to age 50, exploring the consequences of truancy at ages 12–14, nding that
it was strongly related to criminal convictions at age 50; self-reported offending at 32;
problem-drinking at 18 and 32; poor accommodation at 48; and employment problems
at ages 32 and 48. Those who truanted at school were found to earn less and to have
more psychological problems as adults (Robins and Ratcliff 1980). Kandel etal. (1984)
found that truants were more likely than non-truants to have unstable employment
careers, while also being more likely to have broken marriages and more periods of
illness. Truants were also found to have higher debts than non-truants at ages 18 and
32 (Farring ton 1996). Hibbett and Fogelman (1990) and Hibbett etal. (1990) discerned
that truants were more likely than non-truants to experience divorce as adults, to have
more dependent children, to be heavy smokers, to have suffer depression, to have lower
status jobs and higher rates of unemployment.
Recently Carroll (2013), using NCDS data, reported that truancy contributed, in
part, to subsequent social, educational and behavioural difculties within school, but
stressed this was only one part of a multidimensional explanation. Maggs etal. (2008)
reported that truancy at 16 predicted problematic drinking at 42 for both genders,
and the quantity of alcoholic units consumed at ages 16, 33 and 42. Hansen (2003) re-
analysing the Young People and Crime survey data found that those who truanted were
more likely to commit property (p.154) and violent crimes (p.156). Powis etal. (1998)
explored school exclusions, rather than truancy; their data revealed that most excluded
pupils came from single parent families; few lived in homes with an adult wage earner;
half were from ethnic minority groups, and most had engaged in both truancy and
some form of offending behaviour. Drug use was also commonplace and excluded pu-
pils often resided in areas of high deprivation (p.254). To summarize, almost all of
the long-term studies of truancy suggest that it is associated with a range of long-term
negative outcomes, such as depression, substance misuse, offending and poor quality
relationships with employers, spouses and offspring.
Critiquing this literature
Although the aforementioned studies rely on high-quality research designs and reput-
able data sources, there are nevertheless gaps in the framework, which we wish to out-
line. The major critiques which we extend, and attempt to respond to, focuson:
FARRALL ET AL.
120
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

(1) the dominant behaviouralist explanations used, and
(2) the absence of explanations which incorporate wider political, social and
economic variables into their account of rates of truancy.
Work by Carlen and colleagues (Carlen 19 92, Carlen etal. 1992, Gleeson 1994) high-
lights the role of policymaking and political discourses in understanding truancy. Their
work is an attempt to throw light on the structural causes of truancy as a corrective to
the more common focus on individual-level failings. As Gleeson notes (1994:16) ‘The
danger is that behaviouralist explanations, which purport to explain truancy in psycho-
logical terms, do little more than pathologise such stereotypes, xing them in popular
myth’. Indeed, the overall tenor of their study (Carlen etal. 19 92) is to argue that psy-
chological and behaviouralist explanations ignore (to quote Gleeson) ‘the political,
economic and educational consequences of government policy which condition such
behaviour’ (emphasis in original, Gleeson 1994:16). In so doing, they emphasize that
previous research in this area has overlooked the effects of recession, unemployment
and social security cuts on the labour market, communities, schools, parents and pu-
pils, favouring as it does a more atomistic approach. In keeping with this, we seek not
to replace the psychological and behaviouralist explanations, but rather to illuminate
the wider background and social-structural causes which motivate truancy (or, per-
haps, more accurately, demotivate school attendance). As such, our contribution is to
re-emphasize the structural processes along with the individual-level factors.
Scholars before us have identied the challenges of integrating history, politics, cul-
ture and the local environment in criminological research. Although the importance of
these interaction effects and hierarchical relationships has been recognized, few studies
have been able to operationalize a multidimensional approach. High-quality long-term
data are scarce, as are small-area data that are sufciently sensitive. Drawing upon a
range of individual and ecological approaches, the Edinburgh Youth Transitions study
explores individual offending histories in relation to the social and physical structure
of neighbourhoods, and the dynamics of local communities (Smith and McVie 2003).
ResearchDesign
As we were keen to explore changes in the economy, alienation from school, truancy
and subsequent engagement in crime at the individual level over time, we required data
sets with very specic research designs. Although no data sets would ever be perfect for
this, the BCS70 and NCDS make appropriate vehicles with which to study the impact
of dramatic economic restructuring on successive cohorts of school-age children (we
outline both in more detail later). The BCS70 cohort members were born in 1970, and
grew up during the 1980s, during which they would have been subject to changes in
economic, social welfare, housing and schooling policies. The NCDS were born in 1958,
growing up when the welfare state was expanding. Such a research design is described
by Elder and Giele (2009:16) as the ‘pairing [of] strategically related longitudinal sam-
ples’. Hence, by using two cohort studies with respondents born 12years apart, we aim
to highlight ‘variations and differences within and between individuals as they develop
in multidimensional social-historical contexts’ (Almeida and Wong 2009:142).
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
121
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

Outlining the NCDS andBCS70
The NCDS had an initial sample of 17,414, all of whom were born in therst week of
March 1958. Data were collected about and from the sample members in 1958 (birth),
1965 (aged 7), 1969 (11), 1974 (16), 1981 (23), 1991 (33), 2000 (42) and at various points
since. The sample has maintained very good retention rates, with 9,100 (52%) cohort
members being re-interviewed when the survey was last elded (2013). The BCS70
had a slightly smaller sample size (16,135), all of whom were born in therst week of
April 1970. Data were collected about the cohort members in 1970 (birth), 1975 (aged
5), 1980 (10), 1986 (16), 1996 (26), 2000 (30) and again since at various points. The
sample has generally good response rates, although these were lower at age 16 when the
National Union of Teachers was on strike (when about a third of respondents were not
reached). Around two-thirds of cohort members have been interviewed at sweeps since
2000, and the sample remains representative of the original births (Gerova 2006:7). We
outline the specic survey questions upon which we rely when discussing the models
we develop later.
Thatcherism and the Dramatic Economic and Social Changes of the1980s
Let us take a step back and locate the lives of the members of these two cohorts in a
wider social and economic context. There is little doubt that the legislation enacted
during the 1980s, not just relating to education, but to housing, social security, indus-
trial relations and the economic policies pursued, had very profound effects upon the
United Kingdom, both at the time and in the years and decades since. Between 1971
and 1985, some four million jobs were lost from the manufacturing sector. The ofcial
document Social Trends for 2007 (Ofce for National Statistics 2007:47) reportsthat:
Over the last 25years the UK economy has experienced structural change. The largest increase in
employee jobs has been in the banking, nance and insurance industry, where the number of em-
ployee jobs has doubled between June 1981 and June 2006 from 2.7 million to 5.4 million. There
were also large increases in employee jobs in public administration, education and health (up by 40
per cent) and in the distribution, hotels and restaurants industry (up by 34 per cent). In contrast,
the extraction and production industries, made up of agriculture and shing, energ y and water,
manufacturing, and construction showed a combined fall of 43 per cent from 8.2 million jobs in 1981
to 4.7 million jobs in 2006. Manufacturing alone accounted for 81 per cent of this decline, with the
number of employee jobs in this sector nearly halving from 5.9 million in 1981 to 3 million in 2006.
The radicalism of this restructuring extended into the education sector (Hay and
Farrall, 2 011). As Carlen notes, mid-19th century discourses of juveniles being cor-
rupted by poverty and poor parental control were displaced by a discourse that em-
phasized the role of pathological, feckless families that produced delinquent children
during the late-1980s (1992:254). Ahead of the Thatcher governments’ efforts to cut
state-funded school places, a series of right-wing critiques (including the Black Papers
and articles in right-wing newspapers) had started to challenge the existing educa-
tion system. The decline of spending on books in the early 1980s and the reduction
of preschool places for 3–4year olds were documented by Timmins (2001:380) and
Riddell (1985:151). The March 1980 expenditure White Paper projected a 6.9% fall in
expenditure on education in real terms between 1978 and 1979 and 1982–83 (Riddell
FARRALL ET AL.
122
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

1985:151). Staff–student ratios went up, contributing to greater disruption in classes,
more exclusions, greater levels of staff absenteeism, not helped by the reduction in the
status of teachers fostered by government discourse (Gleeson 1994:16; Jones 2003:134 –
5). Timmins (2001:424) notes that in the period between 1984 and 1987, during which
teachers were involved in industrial disputes, the government created a demoralized
pool of teachers whose loyalty to the job was damaged for years to come. After the
Teachers’ Pay and Conditions Act (1987), the teachers’ trade unions were unable to mount
effective resistance to the 1988 Education Reform Act. The 1988 Act introduced, among
other things, the marketization of schools.
Education and schooling policies interacted with other economic and social policies,
such as the 1980 Housing Act (and the various Acts which followed it), and the impact
of economic recession, which eventually saw poorer households concentrated in the so-
cial rented sector, and within that, concentrated within particular areas of the United
Kingdom’s urban spaces (Farrall etal. 2016). Because schools draw (in the rst instance)
from local catchment areas (which the Conservatives, it must be acknowledged, tried
to challenge), this resulted in schools increasingly bifurcating into ‘decent’ schools
(serving relatively afuent areas) and those which served communities with higher
than average rates of deprivation (Gleeson 1994:17). In terms of the wider economy, un-
employment rose from 4.1% in 1979 to 4.8% in 1980 to 8.0% in 1981 (Thomas 2001:52).
After that, it rose (but slightly less steeply) to 9.5% in 1982, and stabilized at around
11% for the 5years from 1983 to 1987. One of the rst things which the Thatcher gov-
ernment did upon gaining power in 1979 was to increase interest rates, which had the
unintended consequence of weakening the United Kingdom’s manufacturing sector
(Thompson 2014:38–9), producing a sharp fall in manufacturing output between 1979
and 1981 (Thompson 2014:38). The economy experienced negative growth for much
of the early 1980s (Thompson 2014:39). So damaging were their economic policies that
by March 1981, the Conservatives had abandoned their monetarist ideals. However, the
UK economy’s troubles were not over and widespread economic disruption and the un-
employment associated with it persisted for many years (Farrall et al. 2 017).
Truancy From School: Trends OverTime
What does the empirical evidence suggest was happening in terms of truancy in
schools in the United Kingdom at this time? In this section of our article, we review
the evidence on trends in truancy and exclusion using the best available data. It is
impossible to nd a consistent source of data charting truancy rates going back to the
1970s, or even the mid-1980s. However, there have been a few studies which give some
insight into rates of truancy across the United Kingdom. Carlen etal . (1992:64) report
a study by the Association of Chief Education Welfare Ofcers in 1973 which sug-
gested that truancy rates were 4%–7%. Astudy in Shefeld (Galloway etal. 1985:54)
suggested that rates were between 0.3% and 7.5% for 1974–76, with a mean of 2%.
Scottish truancy rates were reported to be 14% in Edinburgh and 17% in Glasgow in
1974 (Carlen etal. 1992:64). For 1975, Carlen etal. (1992:64) report a slightly lower
truancy rate of 10% for schools in England and Wales, rising to 15% in 1977 for
schools in Bolton, (Carlen et al. 1992:13 9). Raffe’s (1986) study of truancy among
school leavers in Scotland suggests declining rates between the 1975/76 school year
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
123
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

and the 1981/82 school year, with around 38% admitting to missing ‘a day here and
there’ in 1975/76 to reducing 27% in 1981/82. The Youth Cohort studies (YCS; which
only covered England and Wales) followed students for the nal 3years of their com-
pulsory school education (Table 1). The rst survey started in 1985, and so the nal
year of compulsory study for that cohort was in 1987 (1year after that for the BCS70
cohort). For the YCS 1987 leavers, 48% had truanted at some point during their nal
year of secondary education. Table 1 suggests that around half of the children at
school truanted to some extent in the late-1980s to mid-1990s, but that this dropped
to about a third by the turn of the century.
What these data suggest are low rates of truancy in the 1970s (somewhere around
10% of children missing some proportion of their schooling would appear to be a
reasonable estimation, although the data do uctuate and the denitions of truancy
may vary considerably). In the 1980s, using the YCS data, we see much higher rates
(more than 40%), which declines from the mid- to late-1990s. Against this back-
ground, let us turn to examine how these rates are reected in the two cohorts we
are studying.
Analyses of the NCDS and BCS70 Cohorts’ Experiences of Truancy
Previous analyses of the NCDS and the BCS70 have suggested that truancy was equally
common among the 1958 and 1970 cohorts (Bynner and Parsons 2003:286–7). However,
these analyses did not go beyond descriptive statistics. We start our investigations, nat-
urally, with the 1958 cohort, reporting on their experiences of truancy.
How many of the NCDS cohort truanted from school when they were growingup?
Questions about the number of half days of school missed (rather than truancy per se)
were asked of head teachers at various points during the NCDS eldwork. At age 7, 54%
of the NCDS cohort had missed at least one half day, at age 11 this was 60% and 57%
at age 16. Given that these data could refer to days of illness or authorised absences as
well as truanting, it is hard to draw many rm conclusions from this data. However,
when the cohort was aged 16, their teachers were asked if truanting ‘did not apply’,
‘applied somewhat’ or ‘certainly applied’ for them. This data suggests that at age 16
some 80% of the NCDS cohort members had not truanted, 12% of teachers selected
the ‘somewhat applies’ option whereas only about 8% had truanted to the degree that
their teachers selected the ‘certainly applied’ option (Table 2).
T 1 Rates of truancy 1987–2005 (Youth Cohort Study)
Year 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05
% ever truanted
during nal
year of
education
48 49 50 –48 –50 48 –42 –36 –35 –32 – – 35
FARRALL ET AL.
124
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

The NCDS were asked in 2000 (when they were aged 42)about their recollections of
truanting at school (see Table 3). This suggested that 46% had truanted at some point
when they had been at school.1
Let us now turn to the later cohort and repeat the analyses undertaken for the NCDS.
How many of the BCS70 cohort truanted from school when they were growingup?
At age 10, teachers were asked to account for why cohort members had been absent
from school (Table 4). At this point less than 1% of the BCS70 appeared to have tru-
anted, with only 87 instances of truanting being cited as the explanation of absence at
age10.
By age 16 this gure had increased (the question was asked of the cohort member
directly, it ought to be noted); now some 44% said that they had been absent from
school without having been ill (Table 5) or because they were fed up with school (53%,
data not shown). These gures (for 1986)are broadly comparable to YCS data for 1987,
which suggested that 48% of children had truanted in the pastyear.
Like the NCDS, truanting was asked of the BCS70 cohort in 2000 when they were
aged 30. This suggested that 51% had truanted at some point while they had been at
school (Table 6).2
Thus it would appear that about ve percentage points more of the BCS70 cohort
(51%) than the NCDS (46%) were truanting while at school. This suggests that tru-
ancy had become slightly more prevalent in the intervening 12 or so years and that the
percentages of those truanting ‘all of the time’ had gone up from 3% to 4%. Although
these differences are not large, we argue that they nevertheless represent an important
increase in truancy. Indeed, both a Mann–Whitney U test and a chi-square test based
on a crosstabulation table found that the 1970 cohort was signicantly more likely to
have truanted than the 1958 cohort (both p < 0.000). It is also the case that suspension
1A crosst abulat ion of the teachers’ dat a from 1974 (Table 2) and the cohort members’ recall of their truanc y (Table 4) sug-
gested a ver y high degree of association. The chi-squa re value w as 903.484, p <0.00 0.
2This data, when crosstabulated it with the d ata from age 10, did show a strong positive assoc iation w ith the age 30 data , sug-
gesting that the a ge 30 data were a reliable source of information about truanting 20yea rs earlier. The chi-square v alue was
71.195 , p <0.000.
T 2 Rates of truancy 1974 (NCDS, teacher reports)
Does not apply 9, 911 (8 0%)
Somewhat applies 1, 515 (12%)
Certainly applies 95 8 (8%)
TOTAL 12 ,384 (100%)
T 3 Rates of truancy at school (NCDS, self-report)
Never 6,14 8 ( 55%)
Some of the time 4,779 (43%)
Most of the time 31 6 (3%)
TOTAL 11,243 (100%)
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
125
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

rates increased; twice as many of the 1970 cohort reported being suspended than was
the case for the 1958 cohort (Bynner and Parsons 2003:287). All of our subsequent ana-
lyses are based on the data as reported at age 30 for the BCS70 and 42 for the NCDS
(in keeping with Bynner and Parsons’ analyses, 2003).
Making Sense of Truancy and Economic Change in the Life courses of Those Born 1958
and1970
Our theorizing draws heavily upon the French sociologist Emile Durkheim (1858–
1917). Writing a century before the period which we are chiey concerned with, it was
Durkheim who coined the term ‘anomie’ to refer to the weakening of the social norms
of society and the sense of ‘dislocation’ which this brought about for individuals (1897).
It was, however, an American sociologist (Merton 1938) who adapted Durkheim’s
thinking in such a way as to make it operationalizable in empirical studies. Although
Merton’s essay was initially little used, interest in his thinking grew after the end of the
Second World War (Messner and Rossenfeld 2000:10). Merton’s use of Durkheim’s con-
cept drew on Marxist theories of crime causation, coupled with his own observations of
1930s US society, economy and (recorded) crime rates. Merton reimagined anomie as
a socially based set of discontents which act, over time, to routinely generate deviancy
(including crime) as a by-product of everyday activities which promised economic suc-
cess to all, but which systematically denied success to a great many members of society
(Rock 2007:45). Like many other researchers, we follow Merton’s ‘underlying premise
that the motivations for crime do not result simply from the aws, failures or free
choice of individuals’ (Messner and Rosenfeld 2000:10). Alternatively, and in keeping
with structuralist accounts in general, we believe that the causes of crime are related
to the cultural and structural processes in which individuals are located and which
they need to adapt their behaviours and responses. To summarize, structural-level
T 4 Rates of truancy 1980 (BCS70, teachers report)
Has not tr uanted 11,75 0 (9 9%)
Has tru anted 87 (1%)
TOTAL 11,8 37 (100%)
T 6 Rates of truancy at school (BCS70, self-report)
Never 5, 0 03 (4 9%)
Some of the time 4, 867 (47 %)
Most of the time 405 (4%)
TOTAL 10, 275 (100%)
T 5 Rates of truancy 1986 (BCS70, self-report)
Absent from school but not ill 2,491 (44%)
Not absent from school 3, 229 (5 7%)
TOTAL 5,720 (100 %)
FARRALL ET AL.
126
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

processes block (or at very least seriously impede) the legal opportunities for social and
economic advancement. As a result of this, (some) individuals resort to illegal activ-
ities to achieve success and/or status. In some cases, individuals may express their frus-
tration at nding their routes to advancement ‘blocked’ through criminal behaviour
(Agnew 1985). Agnew revised Durkheim’s and Merton’s thinking, arguing that feelings
of anomie could also be provoked by perceptions that one was ‘trapped’ in aversive
situations. Either way, our argument is that structural-level processes prevent individ-
uals from achieving what they have been encouraged into desiring, and motivate the
use of illegal activities to achieve these goals or to simply express frustration. Hence,
national and regional crime rates are not simply the ‘aggregating up’ of individual-level
action, but rather the outcome of the social forces that shape and mediate individual
actions and their context. Governments, therefore, ‘produce’ variations in crime rates
through their impacts upon the processes which shape those factors which drive crime.
In this way, abrupt and sustained changes in processes which drive crime in turn mo-
tivate processes (such as truancy) which are associated with offending at the individual
level. We support Merton’s initial thinking that the pressure towards anomie was so-
cially structured, being greatest among the lower social strata (as their chances for ad-
vancement are weaker). Accordingly, we argue that the United Kingdom’s experience
during the 1980s meant that the lower social strata were most affected by the social and
economic changes unleashed by Thatcherite policies. From this perspective, theories
of anomie offer an avenue to increase understandings of how dramatic social and eco-
nomic change impact upon society, it’s organization and the crime rates it experiences.
Our thinking is supported not just by structural sociology, but also by research by
psychotherapists on individual loss. The concept of the assumptive world refers to those
beliefs that ground, secure, stabilize or orient people and that accordingly give them a
sense of purpose and meaning to their lives as well as providing feelings of belonging
and connection to others. Parkes writes that the assumptive world ‘is the only world
we know and it includes everything we know or think we know. It includes our inter-
pretation of the past and our expectations of the future, our plans and our prejudices’
(1971:102). Beder argues that the assumptiveworld:
is an organised schema reecting all that a person assumes to be true about the world and the self
on the basis of previous experiences; it refers to the assumptions, or beliefs that ground, secure, and
orient people, that give a sense of reality, meaning and purpose to life. (2004:258)
Most accounts of the assumptive world stress the importance of the notions of safety,
control and justice in the assumptive world. The assumptive world is terribly mundane;
such assumptions lead individuals to the belief that their life has a structure which is
‘knowable’ to themselves and (largely) rewarding and satisfying. The world is under-
standable, predictable, manageable and largely benign. Alongside these assumptions
come the assumptions that oneself is a worthy individual that others care for, and that
others are trustworthy. In short, our assumptions about our social worlds make us think
that the world is understandable, worth caring about and investing in, and unthreat-
ening to ourselves.
Applying this thinking (derived from sociological structuralism and psycho-
therapy) to economic restructuring and truancy, we argue that economic restruc-
turing produces a sense of anomie in pupils at school, and serves to motivate truancy,
especially if it involves widespread, long-term parental unemployment and the loss
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
127
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

of career pathways which would, had they continued to exist, have helped individ-
uals to navigate the transition from school to work, and which (via work) would have
provided the basis for independent living, marriage and family formation. Truancy
itself is associated with later offending. Drawing upon recent sociogenic theories of
desistance (Sampson and Laub 1993) we also argue that employment and marriage
reduce engagement in crime. In this way, our theorizing seeks to explain how and
why economic restructuring provokes truancy, but avoids falling foul of the tendency
to only be able to explain increases in rates of offending, a problem which plagued
many classical theories of offending (Mat za 1964). We wanted to explore the extent
to which national-level social and economic changes may have played a part in tru-
ancy among those at school in the 1980s. To this end we use the NCDS and the BCS70
to explore the ways in which economic policies shaped schooling experiences for
some young children in the 1980s. The NCDS would have completed their education
during the 1970s, whereas the BCS70 would have been at school throughout most of
the 1980s (and until at least 1986). Our hypothesis is that some of the children in
the BCS70 cohort, as they started to think about their lives after school, and started
to become increasingly aware of the economic fortunes of their communities and
neighbourhoods, may have started to become despondent about both schooling and
their abilities to secure a useful role in society and the labour market. This model is
presented in diagrammatic form in Figure 1.
Our model suggests that area-level economic restructuring and the widespread loss of
jobs among men working in heavy industry and mining in the United Kingdom(especially
during the 1980s) will signal to children in the areas affected (even if they are not the chil-
dren of miners, steelworkers, railway employees and those in allied trades) that the assump-
tive world which they had thought was there, has gone forever. This may lead some of these
children to become alienated from school by their mid-teens, and hence to start to truant
from school. Truancy as the literature reviewed earlier makes clear, will be associated with
contact with the criminal justice system into adulthood. However, although area-level eco-
nomic restructuring will reduce the chances of being in employment in one’s mid-20s, for
those who are fortunate enough to secure work, this employment will be associated with
marriage/cohabitation. Both employment in the mid-20s and marriage/cohabitation will
reduce the chances of being in contact with the criminal justice system in laterlife.
Areal
economic
Restructuring
in childhood
Alienaon
from school
at age 16
Truanng
from school
at age 16
Police Arrest
in 30s/40s
Working in
mid-20s
Mid-20s
marriage/co-
habitaon
++
+
+
F. 1 Theoretical model of economic restructuring and truanting
FARRALL ET AL.
128
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

Exploring and testing this model empirically
Area-level economic restructuring was measured by summing two variables from the UK
censuses. These were the proportion of the economically active population employed in
coal mining in each county and the proportion of economically active males who were
unemployed in that same area at the subsequent census. We were unable to simply use
the proportion of the economically active working-age population employed in mining
in later censuses because by the 1981 census coal mining was aggregated with other
primary industries, such as energy and water, so it was not comparable after this date.
Counties were based on 1974–96 counties, and censuses for 1961 and 1971 were geocoded
from smaller areas to these same counties. Error in this geocoding was estimated to be
less and 5%. Figures 2 and 3 show our area-level economic restructuring measure for the
years 1961–71 and 1971–81, respectively.3 In our modelling, Disadvantaged Area (1961–71)
is our measure of area economic restructuring for the area in which the NCDS cohort
member was living in 1974. Similarly, Disadvantaged Area (1971–81) is our measure of
areal economic restructuring for the area in which the BCS70 cohort member was living
in 1986. We choose data for those working in coal mining in 1961 and 1971 as these are a
good barometer of industrial strength in the United Kingdom, whereas unemployment
rates in the same area 10years later is a good measure of the loss of such work. As such,
the proportion of people working in coal mining is used as a proxy for employment in
other heavy industries, as coal mining was frequently co-located with steel production
and processing in South Wales, South Yorkshire, Central Belt Scotland and Teesside,
and shipbuilding (in and around Glasgow in particular), and the maintenance of loco-
motives and railway distribution in centres in Derby, Doncaster, Nottingham, Shefeld,
York and Central Belt Scotland.4 In 1960 there were approximately 607,000 people
(mainly men) working in 698 UK mines, whereas in 1970 these gures had reduced to
290,000 people working in 293 mines.5 Thus our composite measure records for each
county, a combined score composed of the following:
(1) the proportion of people in each county who were employed in mining in 1961
(or 1971 for the 1971–81 analyses), and
(2) the proportion of economically active male employees (traditionally the
‘breadwinner’ in working class households at that time) who were unemployed
in 1971 (or 1981 for the 1971–81 analyses).
These variables, therefore, measure change in local employment patterns, tracking
shifts in the rapid loss of male employment in mining (and related) industries at two
points of time. Although there were other social changes which took place alongside
3Furt her details on the development and use of this variable ca n be provided by the authors on request .
4There were major railway marshalling yard s, e.g. at Toton and Colw ick (both near Notti ngham, Nottinghamshire), Bescot
(Birm ingham, West Midlands), Tees (Middlesborough, Teesside), Mossend (Nort h Lana rkshire, Scotla nd), York (North
Yorkshire), Healy Mills (West Yorkshire), Tyne (Newcastle, Tyne and Wear), Port Talbot and Severn Tunnel (both in Sout h
Wales), Doncaster (S outh Yorkshire), Crewe and War rington (bot h Cheshire), Carlisle (Cu mbria), and in both Tin sley and Wath
(near Shefeld, South Yorkshire). There were large locomotive production, repair and m aintena nce works in Doncaster (South
Yorkshire), Crewe (Cheshire), York (North Yorkshire), Derby (Derbyshire), Stratford (East London) and Glasgow (Scotland).
The UK ca r manufactur ing was centred on the West Midlands. Steel pro duction was centred in South Wales, Central Belt
Scotla nd, Teesside, Shef eld and (albeit to a lesser degree) Corby (Nort hamptonshire).
5Our dat a come from: https://ww w.gov.uk/government/statistical-data-sets/historical-coal-data-coal -production -availability-
and-consumption-1853-to-2011. Last accessed January2019.
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
129
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

F. 2 Areal Economic Restructuring index score of mining (1961) and unemployment (1971)
FARRALL ET AL.
130
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

F. 3 Areal Economic Restructuring index score of mining (1971) and unemployment (1981)
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
131
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

these processes, such as the greater inclusion of females in the labour market, for many
individual households these developments were in part a response to the loss of trad-
itional forms of (male) employment. Many such communities lived and worked closely
together such that local state housing estates (‘council houses’) were dominated by fam-
ilies who derived their household incomes from the same employer (or interdependent
employers), meaning that when coal production declined or ceased altogether in one
community, so the livelihoods of whole estates were impacted upon. In order for readers
to ‘locate’ the parts of Britain most heavily affected by the economic changes between
1961–71 and 1971–81, Figure 2 provides a map of Britain which shows the levels of areal
economic disadvantage using this measure plotted by county using the 1961–71 data. In
Figure 2 one sees that the North-East shoulder of England stands out as an area which
experienced economic disadvantage between 1961 and 1971. Three other areas (also
marked with a rectangle) are also worth a mention. These are Central Belt Scotland,
Central England and the South Wales Valleys. Figure 3 repeats this using data for 1971
and 1981. What one sees rst is that there is a lot of economic disadvantage generally
(especially outside of South-East England). Again, the same four areas stand out as
having experienced higher levels of disadvantage. Second, however, these areas are
also slightly ‘larger’ in terms of their geographical coverage (especially Central Belt
Scotland, the North-East shoulder of England and Central England—which now forms
a ‘belt’ of disadvantage running from the Mersey in the west to the Humber in theeast).
For the 1958 cohort, the census data used were the 1961 and 1971 census (so when
the cohort members were aged 3 and 13). This meant that the measure of economic
restructuring used for the 1958 cohort captured processes of change, from an indus-
trially based local economy (using the 1961 census data) to one in which there was
a degree of male unemployment 10years later (using data from the 1971 census)
while the 1958 cohort were in their formative years. This, when repeated for the 1970
cohort, used the proportion of people working in mining in the local area in 1971
summed with the proportion of economically active males who were unemployed in
that same area in 1981 (when the cohort was aged 1 and 11—again during their for-
ma t ive yea r s).
Operationalizing themodel
School alienation
Both the NCDS and BCS70 cohort members were asked a series of questions about their
feelings towards school when aged 16 (in 1974 and 1986, respectively). Both cohorts
were asked how much the following statements were true for them: I feel school is largely a
waste of time; I am quiet in the classroom and get on with my work; I think homework is a bore; I
nd it difcult to keep my mind on my work; I never take work seriously; I don’t like school; I think
there is no point in planning for the future—you should take things as they come, and, nally;
I am always willing to help the teacher). When factor analysed (separately for each cohort)
these items produced one factor, which we use as our measure of school alienation at
age 16. Higher scores on these measures mean that the respondent was more alienated.
(See Appendix for factor loadings etc. for both cohorts.)
Truanc y was asked about in 2000, when the NCDS were aged 42 and the BCS70 were
aged 30. We use this data as it was consistently worded in both surveys, and because
FARRALL ET AL.
132
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

most other analyses on these data sets have relied on the data collected at this sweep.
Respondents were asked Thinking back to when you were at school, did you ever play truant,
that is, stay away from school when you should have been there? (The codes offered were those
shown in Tables 3 and 6 ea rlier).
Contact with the criminal justice system (labelled Offending in Figures 4 and 5):
Again, the same questions were asked of both cohorts in 2000. Herein we rely on two
questions combined so as to produce a continuous measure. The rst question was
worded Have you ever been arrested by a police ofcer and taken to a police station since [previous
inte r view]. If respondents said that they had been arrested, they were then asked How
many times has this happened? Thus respondents who had not been arrested were coded
0, whereas those who had been arrested were coded with the number of times they had
been arrested. Employment in early adulthood (Employed) was recorded at age 23 for
the NCDS in 1981, and at age 26 for the BCS70 in 1996 (we use derived variables for
both cohorts). Marriage/cohabitation (Li ving w/par tner) was asked about at age 23 for
the NCDS (1981), when respondents were asked who they lived with (we treated this as
a binary with those living alone as one group, and those living as married or actually
married as the other). At age 26 in 1996, the BCS70 were asked: Which of these best de-
scribes your current living situation? with the codes offered being living with your husband
or wife; living as a couple with someone; living alone or in some other arrangement. We used
the rst of these two codes as the measure cohabitation, and the third as indicating
non-cohabitation.
Results
We begin by examining the model for the NCDS cohort (Figure 4). (The standard-
ized coefcients are listed on the path lines between the variables; bolder lines indi-
cate statistical signicance of p=<0.05). The model nds that living in an area which
was experiencing economic restructuring between 1961 and 1971 was associated with
higher levels of school alienation (p< 0.000). However, areal economic disadvantage
was not statistically signicantly associated with truancy at school or offending while
aged 16–42. Areal economic disadvantage was associated with employment at 23, such
that those people living in areas which had experienced economic restructuring be-
tween 1961 and 1971 were less likely to be employed in 1981. School alienation is, as one
might imagine, strongly related to truancy (p<0.000), which in turn is strongly related
to offending (p<0.000). Being in work at age 23 in 1981 was strongly related to cohabit-
ation at age 23 (p<0.000). Of these two variables, only being in work was statistically
signicantly related to offending, such that those in work were less likely to have been
arrested (p<0.000). Overall, the model explained only 3% of the variance in Offending.
The t of the model with the data was reasonable, but below the standard measures of
acceptability; the CFI was .829 (ideally one would want this above .9). The RMSEA was
at a much more satisfactory level (of .048, ordinarily one wants this to be below .08,
and ideally below .05). So, overall the data analyses suggest that the data ts the model
moderately well, but that the economic restructuring thesis is not well supported in that
the model only explains 3% of offending, and the CFI is lower than is ideal. In short,
economic restructuring between 1961 and 1971 did not appear to be related to either
truancy in 1974 or offending between 1974 and2000.
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
133
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

F. 4: Structural Disadvantage, School Alienation, Truancy, Life-Course Transitions and
Offending (NCDS)
F. 5: Structural Disadvantage, School Alienation, Truancy, Life-Course Transitions and
Offending (BSC70)
FARRALL ET AL.
134
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

Turning now to the BCS70 data (see Figure 5), we nd stronger relationships between
the variables (again, the standardized coefcients are listed on the path lines between
the variables) than was the case for the NCDS model. All but one of the paths (that be-
tween cohabiting and offending) were statistically signicant. The CFI was acceptable (at
.906, one wants this to be above .9, recall), and the RMSEA was .041. Overall, the model
t statistics suggest that the data ‘ts’ the model well, and that area-level economic disad-
vantage thesis is appropriate to measure within the given framework. The model explains
11% of Offending. In short, economic restructuring between 1971 and 1981 does appear to
be related to both truancy in 1986 and offending between 1986 and 2000. Interestingly,
the paths between School Alienation and Truanc y , and between Truanc y and Offending are
stronger for the BCS70 model than they were for the NCDS model, suggesting that these
relationships have become more important overtime. Although being employed was
negatively associated with offending for both cohorts, this again becomes stronger for the
BCS70, suggesting that the relationship between employment and offending has grown
stronger and that being out of work is more associated with offending than it used to be.
Drawing upon recent theories of desistance, we found that employment in the mid-20s re-
duced engagement in crime. However, being in employment itself was strongly (and nega-
tively) related to area-level experiences of economic restructuring; those individuals who
were living in areas which had experienced greater levels of this when they were in their
teenage years were less likely to be working in their mid-20s. This is suggestive of a con-
tinued differential impact of wider economic restructuring. For the BCS70 cohort, being
in employment in their mid-20s was much more strongly associated with lower levels of
engagement in crime than it was for the NCDS, suggesting that the role of employment
as a route out of crime is contingent on historical period.
Discussion
Let us start by discussing the limitations of our study. Because the data sets we employ
did not collect any qualitative data, we are unable to assess the meaning of truancy for
the children themselves. However, Willis’ and Carlen et al.’s studies (1977 and 1992,
respectively) suggest that boredom, and a sense of hopelessness or pointlessness were
a motivating factor in truancy. The items used to measure School Alienation, and which
were strongly associated with truancy, speak to these same feelings. The strengths of
our article, on the other hand, are the use of national-level data sets of the highest
quality from two highly respected studies, and which enable us to examine the un-
folding of differential regional impacts of economic restructuring on school attend-
ance. Furthermore, the two cohorts we have studied (as opposed to the more commonly
used unicohort studies which are often drawn from one town or city, and as such do not
permit analyses of regional differences) are both national samples (as opposed to more
locally based samples), and number cases in the thousands (rather than hundreds).
Using two birth cohort studies we have demonstrated that radical economic restruc-
turing in the United Kingdom during the 1980s (which resulted in high levels of re-
gional unemployment and the destruction of key industries) affected those who were
growing up at the time. The BCS70 cohort was more likely than the cohort who grew
up ahead of them to disengage with school and become alienated from education. The
long-term impact of their experiences was an increase in adult offending. Although it is
not simple to locate and incorporate suitable area-level measures into statistical analyses
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
135
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

of this kind, our exploration suggests there is considerable potential and meaning in
doing so. We found that, alongside other individual characteristics, the socio-economic
conditions in which the children were raised, could have a long reach over their en-
gagement in crime as adults. The example described here demonstrates the value of
criminologists fostering ever-closer links with political and economic history.
From our perspective, the value of this analysis is not simply about expanding our appre-
ciation of the ‘wider contexts’. Instead, we argue that political and economic conditions are
fundamental to understanding why and how people offend (Jennings et al. 2017). De spite
the undoubted quality of the research into criminal careers in general, and as it pertains
to the relationship between schooling and offending, it remains the case the much of the
quantitative research has tended to tackle the causal processes of offending in a largely in-
dividualised manner. This ranges from a near-total emphasis on individual-level processes
(Gottfredson and Hirschi 1990), to individual–institutional interactions (Merton 1938;
Hirschi 1969) to more ecological models (Shaw and McKay 1942). This observation has led
Robert Sampson to note that society and the idea of social change was one of the key elem-
ents which was missing from current research into criminal careers (2015:278–9). Similar
observations about life-course research have been made by those working outside of crim-
inology. For example, Mayer noted that the ‘unravelling of the impacts of institutional con-
texts and social processes . . . on life-courses has hardly begun’ (2009:426), addingthat
we know next to nothing about how the internal dynamics of life-courses and the interaction of de-
velopmental and social components of the life-course var y and how they are shaped by the macro
contexts of institutions and social policies.
Thus, although life-course criminology, we would argue, has focussed on what one might
call ‘proximate institutions’ (families, schools, employers and communities most obviously),
those institutional arrangements and the discourses and policies which surround and ow
from more distal institutions and the ideas they promulgate (political parties, governance
structures, discourses about schooling and ‘truants’, ideological stances on education, eco-
nomic policies and the thinking underpinning the funding of social services which support
communities) have not received very much attention at all. In short, the current approaches
adopted by life-course criminologists tend to encourage the construction of ‘the offender’
in individualistic terms—and this it would appear extends to research into truancy and
offending. The inuence of the wider policy agenda has been overlooked.
Our article started with a critique of current approaches to truancy and offending—
which share much in common (in terms of their behaviouralist thinking and measure-
ment) with criminal careers research. We argued that the current thinking has tended
to ignore the structural drivers of truancy (and indeed, changes in these structural con-
ditions) and in so doing have focused thinking and policy initiatives at the individual
level. This has the unfortunate side effect of pathologizing the individuals (and their
families) and focusing on attributes of the school staff at the expense of a wider and, we
feel, more nuanced understanding of the drivers of truancy. Our study, it ought to be
acknowledged, still nds a relationship between truancy and offending, but, moreover,
it nds that the role of structural-level variables (here economic restructuring over a
10-year period) is a more powerful predictor of truancy during periods of dramatic
economic change. As such, the causal relationship is strong and variant. In other words,
when economies shift from (in this case) an industrial base to a post-industrial base, we
nd stronger relationships between economic change and truancy and offending. This
FARRALL ET AL.
136
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

means that the processes associated with truancy and offending, while at the individual
level may appear to be invariant over time, may themselves be associated with changes
at the structural level which are far less invariant, and, at least in the period we have
examined, due to government economic policies. This returns us to the point, afore-
mentioned, made by Mayer (2009); almost all of the previous assessments of the ‘school-
to-prison-pipeline’ have relied upon one cohort of schoolchildren. Our approach of
combining two strategically related samples has enabled us to explore how wider social
and economic structures shape individual life courses with regards to schooling, and
how relationships between key variables can emerge or strengthen over time. This has
important ramications for those studying and theorizing both life courses and crim-
inal careers as many existing studies do not permit an examination of the role of chan-
ging structures, and in so doing may be overlooking important components needed to
explain key individual-level processes and outcomes.
Funding
This work was supported by the ESRC (as award ES/P002862/1).
A
We wish to thank Pat Carlen and John Bynner for their comments and support while we
were conducting our analyses, and which encouraged us that our thinking was along the
correct lines. Brian Dodgeon was of tremendous assistance in the early stages of our ana-
lyses in helping us understand the full complexity of the data sets. The BJC’s anonymous
reviewers also assisted in improving the quality of the opening section of the article.
Colleagues at the Universities of Shefeld and Derby have been most supportive of our
endeavours, and we thank them all. Errors in analyses remain the responsibility of the
authors, however.
Appendix
Factor loadings, KMO and eigenvalue for each cohort’s School Alienation score
Item NCDS BC S70
Waste of time .654 .635
Quiet in class −.760 −.419
Homework a bore .551 .585
Concentrate at school .449 .546
Take work seriously .589 .660
Don’t like school .699 .678
No point planning .340 .404
Willing to help −.345 −.427
KMO .831 .847
Eigenvalue 3.0 51 3.116
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
137
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

R
A,R. (1985), ‘A Revised Strain Theory of Delinquency’, Social Forces, 64: 151–67.
A,D. and W,J. (2009), ‘Life Transitions and Daily Stress Processes’, in G.Elder
and J.Giele, eds., The Craft of Life Course Research, 141–62. Guilford Press.
A,G. and C    ,  P. (2006), ‘Truancy in Secondary School Pupils’, Research Papers
in Education, 21: 467–84.
B,J. (2004), ‘Loss of the Assumptive World’, OMEGA: Journal of Death and Dying, 50:
255–65.
B,J. and P, S. (2003), ‘Social Participation, Values and Crime’, in E.Ferri,
J.Bynner and M.Wadsworth, eds., Changing Britain, Changing Lives, 261–94. Institute of
Education.
C     ,  P. (1992), ‘Pindown, Truancy, and the Interrogation of Discipline’, Journal of Law
and Society, 19(2): 251–70.
C,P., W,J. and G,D. (1992), Tr u ancy. Oxford University Press.
C  ,H . (2013), ‘The Social, Emotional and Behavioural Difculties of Primary School
Children With Poor Attendance Records’, Educational Studies, 39: 223–34.
D, J. G. (1990), Adolescents at Risk. Oxford Univ Press.
D,E. (1 8 9 7/19 51), Suicide. The Free Press.
E,G. and G,Z. (2009), ‘Life Course Studies: An Evolving Field’, in G.Elder, and
J.Giele, eds., The Craft of Life Course Research, 1–24. Guilford Press.
F,S., G ,  E ., J, W. and H,C. (2016), ‘Thatcherite Ideology, Housing
Tenure, and Crime’, British Journal of Criminology, 56: 1235–52.
F, S., J , W., G, E. and H, C. (2017), ‘Thatcherism, Crime and the
Legacy of the Social and Economic Storms of the 1980s’, Howard Journal of Crime and
Justice, 56: 220–43.
F  ,D. (1980), ‘Truancy, Delinquency, the Home and the School’, in I.Berg and
L.Hersov, eds., Out of school, 49–63. Wiley.
F  , D. (1996), ‘Later Life Outcomes of Truants in the Cambridge Study’, in
I.Berg and J.Nursten, eds., Unwillingly to school (4/e), 96–118. Gaskell.
G   , D. , M,R. and W,B. (1985), ‘Persistent Absence from School and
Exclusion from School’, British Educational Research Journal, 11: 51–61.
G  , E. (1996), Trua n c y . Ofce of Juvenile Justice and Delinquency Prevention, U.S.
Department of Justice.
G , V. (2006), BCS1986 Data Note 1: Enhancing the Head-teacher Dataset. Centre for
Longitudinal Studies, Institute of Education.
G,D. (1994), ‘Wagging, Bobbing and Bunking Off’, Educational Review, 46: 15–19.
G,S. and G,E. (195 0), Unravelling Juvenile Delinquency. Commonwealth Fund.
G,M. and H,T. (1990), A General Theory of Crime. Stanford University.
H,K. (2003), ‘Education and the Age-Crime Prole’, British Journal of Criminology, 43:
141 – 68 .
H, C. and F,S. (2011), ‘Establishing the Ontological Status of Thatcherism by
Gauging Its “Periodisability”’, British Journal of Politics and International Relations, 13:
439–58.
H     , A ., and F       ,K . (1990), ‘Future Lives of Truants’, British Jour nal of Educational
Psycholog y, 60: 171–9.
FARRALL ET AL.
138
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

H,A., F ,K. and M,O. (1990), ‘Occupational Outcomes of Truancy’,
The British Journal of Educational Psychology, 60: 23–36.
H,T. (1969), Causes of Delinquency. University of California Press.
J,W., G  , E ., F,S. and H,C. (2017), ‘Penal Populism and the Public
Thermostat’, Governance, 30: 463–81.
J,K. (20 0 3), Education in Britain: 1944 to the Present. Polity Press.
K,D., R    , V. and K   , P. (1984), ‘Continuity in Discontinuities: Adjustment
in Young Adulthood of Former School Absentees’, Youth & Society, 15: 325–52.
M,J., P   , M . and F,L. (2008), ‘Childhood and Adolescent Predictors
of Alcohol Use and Problems in Adolescence and Adulthood in the National Child
Development Study’, Addiction, 103: 7–22.
M,D. (196 4), Delinquency and Drift. Wi ley.
M ,K. (2009), ‘New Directions in Life-Course Research’, Annual Review of Sociology, 35:
413–33.
M,R. (1938), ‘Social Structure and Anomie’, American Sociological Review, 3: 672–82.
M,S. and R,R. (2000), Crime and the American Dream, 3/E. Wadsworth.
O  N S (2 0 07), Social Trends No. 37. Palgrave Macmillan.
P,C. (1971), ‘Psycho-Social Transition’, Social Science and Medicine, 5: 101–15.
P , B., G        ,  P. , G ,  M., L , C. and S , J, (1998), ‘Drug Use and
Offending Behaviour Among Young People Excluded from School’, Drugs, 5: 245–56.
R    ,D . (1986), ‘Unemployment and School Motivation: The Case of Truancy’, Educational
Review, 38: 11–19.
R,K. (2008), ‘Behaviour and Attendance: The National Picture; a Synopsis’, Educational
Review, 60: 333–44
R,A.J. (1951), ‘Delinquency as the Failure of Personal and Social Controls’, American
Sociological Review, 16: 196–207.
R,P. (1985), The Thatcher Government, 2/E. Basil Blackwell.
R,L. and R, K . (1980), ‘The Long-Term Outcome of Truancy’, in L.Hersov
and I.Berg, eds., Out of school: Modern perspectives in truancy and school refusal, 65–83. John
Wi ley.
R,P. (2007), ‘Sociological Theories of Crime’, in M.Maguire, R.Morgan and R.Reiner,
eds., The Oxford Handbook of Criminology, 4th edn Oxford University Press.
R, M., J, W., P, A., O , T. and F, D. (2017), ‘The
Importance of School Attendance’, Crime and Delinquency, 63: 592–612.
S, R . (2015), ‘Crime and the Life-course in a Changing World’, Asian Journal of
Criminology, 10: 277–86.
S,R. and L,J. (199 3), Crime In The Making. Harvard University Press.
S,C . and MK,H. (1942), Juvenile Delinquency and Urban Areas. University of Chicago
Press.
S,D. and MV,S. (2003), ‘Theory and Method in the Edinburgh Study of Youth
Transitions and Crime’, British Journal of Criminology, 43: 169–95.
S, P. S. and L, N. P. (2014), ‘Graduation Outcomes for Truant Students’,
Children and Youth Services Review, 43: 138–44.
T ,R. (2001), ‘UK Economic Policy’, in S.Savage and R.Atkinson, eds., Public Policy
Under Blair. Palgrave.
ROLE OF R ADIC AL EC ONOMIC RE STRUCT URI NG IN TRUA NCY
139
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020

T,H (2014), ‘The Thatcherite Economic Legacy’, in S.Farrall and C.Hay, (2014),
eds., The Legacy of Thatcherism, 33–68. Oxford University Press.
T,N., (2001), The Five Giants. Harper Collins.
V ,M., M      , B ., S-W,C., P     ,B . and A    , A . (2013), ‘Prevalence
and Correlates of Truancy in the US’, Journal of Adolescence, 36: 767–76.
W,P. (197 7), Learning to Labour. Saxon House.
FARRALL ET AL.
140
Downloaded from https://academic.oup.com/bjc/article/60/1/118/5530423 by guest on 23 December 2020