Criminal Justice and Behavior

http://cjb.sagepub.com Crime in Adult Offspring of Prisoners: A Cross-National Comparison of Two Longitudinal Samples
Joseph Murray, Carl-Gunnar Janson and David P. Farrington Criminal Justice and Behavior 2007; 34; 133 DOI: 10.1177/0093854806289549 The online version of this article can be found at: http://cjb.sagepub.com/cgi/content/abstract/34/1/133

Published by:
http://www.sagepublications.com

On behalf of:
International Association for Correctional and Forensic Psychology

Additional services and information for Criminal Justice and Behavior can be found at: Email Alerts: http://cjb.sagepub.com/cgi/alerts Subscriptions: http://cjb.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations (this article cites 15 articles hosted on the SAGE Journals Online and HighWire Press platforms): http://cjb.sagepub.com/cgi/content/refs/34/1/133

Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

CRIME IN ADULT OFFSPRING OF PRISONERS A Cross-National Comparison of Two Longitudinal Samples

JOSEPH MURRAY
University of Cambridge

CARL-GUNNAR JANSON
Stockholm University

DAVID P. FARRINGTON
University of Cambridge

Studies from several countries suggest that parental criminality is a strong predictor of childrens own criminal behavior. Recently, the authors found that parental incarceration predicted boys delinquency in an English cohort, even after controlling for parental criminality and other childhood risks. The present study uses data from Project Metropolitan (Sweden) on 15,117 children born in the same year as the English cohort (1953) to test whether results in England were replicated in Sweden. In Sweden, parental incarceration predicted childrens own criminal behavior, but unlike in England, the effects of parental incarceration disappeared after statistically controlling for the criminality of the parent. This cross-national difference may have been the result of shorter prison sentences in Sweden, more family friendly prison policies, a welfare-oriented juvenile justice system, an extended social welfare system, and more sympathetic public attitudes toward crime and punishment. Keywords: intergenerational crime/delinquency; parental incarceration/imprisonment; prison; cross-national research; longitudinal research

tudies from several countries show that parental criminality predicts childrens own criminal behavior in later life (Farrington, Jolliffe, Loeber, Stouthamer-Loeber, & Kalb, 2001; Fergusson, Horwood, & Nagin, 2000; Glueck & Glueck, 1950; McCord, McCord, & Zola, 1959; Robins, West, & Herjanic, 1975; Smith, 1991; West & Farrington, 1977). However, there has been little investigation of whether parental incarceration affects children in addition to the effects of parental criminality. Recently, Murray and Farrington (2005) found that parental incarceration predicted antisocial behavior and delinquency among working-class males in London, even after controlling for parental criminality. Of boys who experienced parental incarceration in their first 10 years of life, 48% were convicted as adults, compared with 25% of boys who were separated from their parents for other reasons. Although the effects of parental incarceration were reduced after controlling for parental criminality and other childhood risk factors,
AUTHORS NOTE: We are grateful to Sten-ke Stenberg for his cooperation and help in using Project Metropolitan data. Carl-Gunnar Janson is grateful to the Bank of Sweden Tercentenary Foundation for the unprecedented decision to fund Project Metropolitan for a 21-year period. David Farrington is grateful to the Home Office, which has funded the Cambridge Study in Delinquent Development. Joseph Murray thanks the Economic and Social Research Council for funding his research. Lynne Murrays comments on a draft of the article were much appreciated. Correspondence concerning this article should be addressed to Joseph Murray, Institute of Criminology, Sidgwick Avenue, Cambridge, CB3 9DT, United Kingdom; e-mail: jm335@cam.ac.uk.
CRIMINAL JUSTICE AND BEHAVIOR, Vol. 34, No. 1, January 2007 133-149 DOI: 10.1177/0093854806289549 2007 American Association for Correctional and Forensic Psychology

133
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

134

CRIMINAL JUSTICE AND BEHAVIOR

parental incarceration remained significantly predictive of antisocial-delinquent outcomes, even up to age 32. To date, there has been no other prospective longitudinal study of the effects of parental incarceration on offspring offending. These results need replication. Also needed is investigation of the effects of parental incarceration at different developmental stages, for girls as well as boys, for children of different social classes, and for children in different social and penal cultures. In this study, we examined the effects of parental incarceration on children born in Sweden in the same year as the English cohort (1953). The term incarceration refers to any form of custodial confinement, including open or closed prisons, local or training prisons (in the United Kingdom), regional or national prisons (in Sweden), and jail or prison (in the United States). Parental incarceration was relatively rare in Sweden in the 1950s: 4% of children had a parent who was incarcerated by the time the child was 19 in Project Metropolitan. It is not known how many children have incarcerated parents in Sweden today. However, in other jurisdictions, parental incarceration is a growing phenomenon. Between 1991 and 1999, the number of minor children with a parent in state or federal prison in the United States rose from 1 million to 1.5 million (Mumola, 2000). It is critical to inform public policy makers of the effects of incarceration on prisoners families and children, as well as on prisoners themselves. Farrington (2002) suggested six mechanisms that might link parent and offspring criminality: (a) intergenerational exposure to risk, (b) assortative mating (male and female offenders tend to cohabit or marry, and children with two criminal parents are likely to be disproportionally antisocial), (c) imitation and teaching of crime, (d) mediation through environmental risks, (e) genetic mechanisms, and (f) official (police and/or court) bias. Murray and Farringtons (2005) finding that parental incarceration affected children over and above parental criminality suggests either that these mechanisms operate more strongly when a parent is incarcerated or that other mechanisms link parental incarceration and child delinquency. For example, parental incarceration might amplify child delinquency because of parent-child separation during incarceration. For detailed reviews of literature on prisoners children, see Murray (2005) and Johnston (1995). Small-scale, mainly qualitative, research suggests that the incarceration of a parent can affect children indirectly through reduced family income, home and school moves, traumatic prison visits, disrupted relationships between prisoners and those who care for their children, stressed parenting by outside caretakers, inadequate explanations given to children, stigma, shame, and decreased social support (Boswell & Wedge, 2002; Hounslow, Stephenson, Stewart, & Crancher, 1982; Lowenstein, 1986; Richards et al., 1994; Sack, 1977; Sack, Seidler, & Thomas, 1976; Shaw, 1987; Stanton, 1980). Assuming that parental incarceration does affect children through these environmental mechanisms, we hypothesize that parental incarceration during childhood is a stronger predictor of offspring criminality than parental incarceration before childrens birth. Murray and Farrington (2005) were unable to test whether parental incarceration had different effects on boys at different ages or whether girls or children from different social classes were similarly affected. Previous small-scale studies produced mixed findings regarding effects of parental incarceration on boys and girls. Gabel (1993) and Sack (1977) both found worse antisocial reactions among boys, but Friedman and Esselstyn (1965) reported

Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS

135

worse effects for girls. Because most incarcerated parents are fathers (with whom boys might identify more) and because boys appear to react worse than girls to other types of family disruption, such as parental divorce (Rodgers & Pryor, 1998), we hypothesize that males have worse reactions to parental incarceration than females. We examine whether offspring react differently to parental incarceration according to their developmental stage at the time of incarceration: early childhood (0 to 6) compared with midchildhood and adolescence (7 to 19). Although attachment theory (Bowlby, 1969) suggests that experiences of separation might be more harmful for younger children, other effects of parental incarceration, such as social stigma and school moves, might be worse for older children. Hence, our analyses of developmental differences are essentially exploratory. It is also unclear how social class might affect the relationship between parental incarceration and offspring delinquency. For middle- and upper-class children, parental incarceration might be associated with more social stigma than for working-class children. However, the fact that middle-class families have more resources might protect children from some consequences of parental incarceration. In one study, Anderson (1966) suggested that prisoners wives experienced similar levels of crisis, regardless of their social class. In this article, we investigate whether findings from England are replicated in Sweden and whether there is any evidence of age, sex, or social class differences in the relationship between parental incarceration and offspring criminality in Sweden.
METHOD
PARTICIPANTS

Project Metropolitan (Sweden). Project Metropolitan is a prospective longitudinal survey of 7,719 males and 7,398 females born in 1953 and registered as living in the Stockholm metropolitan area in 1963. The project has studied a wide range of social phenomena (including class structures, family socialization practices, educational achievement, mental disorders, and crime) primarily using official records of cohort members and their families. Project Metropolitan has been directed by Carl-Gunnar Janson since its inception in 1964. Detailed descriptions of the data archive and major findings have been published in more than 40 research reports from the University of Stockholm, Department of Sociology (for overviews of the project, see Hodgins & Janson, 2002, chapters 2 and 3; Janson, 2000). The Cambridge Study in Delinquent Development (England). To make the comparison of English and Swedish results more exact, we reanalyzed data from the Cambridge Study in Delinquent Development (the Cambridge Study). The Cambridge Study is a prospective longitudinal study of delinquency among 411 working-class males, first contacted in 19611962. The sample was chosen by taking all the boys who were then aged 8 to 9 and on the registers of six state primary schools within a 1-mile radius of a research office that had been established in South London. Hence, the most common year of birth of these males was 1953. Most of the males were White (97%) and of British origin. The study was originally directed by Donald J. West, and it has been directed since 1982 by David P. Farrington, who has worked on it since 1969. For details of methods and major findings of the Cambridge Study, see Farrington (1995, 2003).

Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

136

CRIMINAL JUSTICE AND BEHAVIOR

PROCEDURE

Because of ethical and political difficulties with conducting randomized experiments concerning incarceration and the practical difficulties of prospectively identifying prisoners, the most promising method for studying the effects of parental incarceration on children is to analyze data collected in large, longitudinal, community surveys. In this study, we used police records of children in Project Metropolitan, and criminal records of their parents, to investigate the relationship between parental incarceration and offspring offending in Sweden. Unfortunately, most large-scale child development studies have not collected information on parental incarceration. Project Metropolitan represents a rare opportunity to study this topic because criminal record data were collected on two generations and because it was based on such a large sample, including several hundred offspring of prisoners. First, we compared crime among Swedish children according to whether their parents were incarcerated or convicted and received other sentences (such as a fine or probation). Second, we investigated effects of parental incarceration on different types of crime. Third, we tested whether a dose-response relationship existed between parental incarceration and offspring offending (i.e., whether the number of offspring offenses continuously increased according to the number of times parents were incarcerated). Fourth, we examined sex and class differences in the effects of parental incarceration. Fifth, using logistic regression, we examined the effects of parental incarceration on offspring offending controlling for parental criminality. Sixth, we compared the effects of parental criminality and incarceration on offspring offending in Sweden and England.
MEASUREMENT

Parental criminality, parental incarceration, and offspring offending variables were mostly dichotomized for these analyses. As well as simplifying the presentation of results, dichotomization equates the sensitivity of measurement of all variables, does not necessarily cause a decrease in measured strength of association, and makes it possible to compare the predictive strengths of explanatory variables (Farrington & Loeber, 2000). Analyzing dichotomous variables in Project Metropolitan also facilitated more direct comparison with results found in England. Data on parental criminality and incarceration (Sweden). Parental convictions and incarceration (up to when children were age 19) were identified from the criminal records of childrens fathers or if the information about the father was meager, from criminal records of the childs mother (Janson, 1975, p. 15).1 If the participant was adopted at birth, then the adoptive parents records were searched. However, adoption was very rare in Sweden in the 1950s; before reaching adulthood, about 1% of the cohort members were adopted. Parental criminal records were collected from the National Police Board (Rikspolisstyrelsen) for three time periods separately: before 1953, 1953 to 1959, and 1960 to 1972, corresponding to before cohort childrens births, when they were 0 to 6, and when they were 7 to 19. For each time period, there is information about how many times the parent (usually the father) was sentenced unconditionally (incarcerated) or given alternative sentences following conviction. Cases of pretrial incarceration could not be identified from these records, but rates of pretrial incarceration in Sweden in the 1950s were very low. In 1957, 417 persons were remanded in custody in the whole of Sweden; of these, only 17 were still in custody at the end of the year.
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS

137

During childhood (0 to 19), 2% (n = 283) of the Project Metropolitan cohort had a parent incarcerated, and 2% (n = 245) of the cohort had a parent incarcerated before their birth (and not again up to age 19). There are two main limitations to the data on parental incarceration in Project Metropolitan. First, criminal records of both parents were not searched for each child, and it is not known exactly how many records pertain to biological fathers, biological mothers, or adoptive parents. Second, it appears that the records only refer to parents identified at birth. However, changes in family structure were relatively rare in Sweden in the 1950s; by 1968, when the children were about 15 years old, only 9% of childrens parents had divorced or separated. Data on offspring crime (Sweden). Cohort members criminal behavior was measured from police records of offenses collected each year until the end of 1983 (when cohort members were aged 30). Records covered offenses committed in the whole of Sweden, including stealing, violent crimes, vandalism, fraud, traffic crimes, narcotic crimes, and other (see Wikstrm, 1987). Traffic offenses were excluded from our analyses to make them comparable with analyses from England. Dead cohort members were not excluded from figures on criminal convictions, but this only has a minimal effect on criminal career patterns in these cohorts (Farrington & Wikstrm, 1994). Chronic offenders in the Swedish cohort were defined as the 5% of cohort members with the most offenses between ages 19 and 30. Data on social class (Sweden). To compare results from Sweden and England, we followed Farrington and Wikstrms (1994) procedure of using social-class data measured when offspring were age 10. In Sweden, social class was measured according to the occupational status of cohort members families in the 1963 population register and usually referred to the father. The family referred to those with whom the cohort member lived in 1963, whether they were registered as living with their parent(s) or other people. In this study, working class refers to working-class occupations (n = 5,722) and to categories of no occupation and no information on occupation (n = 412). Middle-upper class refers to lower-middle-class occupations (n = 6,396) and upper and upper-middle-class occupations (n = 2,587). Data on parental criminality and incarceration (England). Parental criminality and incarceration (up to when boys were aged 18) were measured by searching criminal records of study males biological fathers and mothers at the Criminal Record Office in London. Social workers files were also used to identify further cases of parental incarceration for minor offenses or on remand (up to the boys 15th birthdays). In previous analyses (Murray & Farrington, 2005), cases of parental incarceration were excluded if parents were first incarcerated when boys were aged 11 to 18. However, for this article, they were included to compare with results from Sweden. Accordingly, 7% (n = 29) of the cohort had a parent incarcerated at any time during childhood (0 to 18), and 4% (n = 17) had a parent incarcerated before their birth (and not after). Of study males whose parents were incarcerated during their childhood, 24 boys had a father incarcerated, 4 boys had a mother incarcerated, and 1 boy had both his father and mother incarcerated. Therefore, for both studies, parental incarceration refers overwhelmingly to paternal incarceration.
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

138

CRIMINAL JUSTICE AND BEHAVIOR

Data on offspring crime (England). Cohort members criminal behavior was measured using records from the central Criminal Record Office in London. These data exclude minor crimes such as common assault, traffic infractions, and drunkenness. The most common convictions were for thefts, burglaries, and unauthorized taking of vehicles, although there were also quite a few convictions for violence, vandalism, fraud, and drug use. In England, chronic offenders were identified as the 7% of study males with the most number of convictions between 19 and 30. One difference in the definition of crime between England and Sweden in the postwar years was that in Sweden, acts of joyriding, in which cars were used but then abandoned, were not defined as thefts but as less serious unlawful procedures (Janson, 2004). In England, all acts of joyriding were recorded as criminal offenses. Another difference is that in Sweden, criminal records referred to offenses, although records in England were of convictions. Specifically, if a Swedish offender was caught for one offense and admitted four others, he would be recorded for five offenses. If an offender in England was convicted for one offense and admitted four others, it is likely that he would be convicted only for one offense and have the other four taken into consideration. Offenses taken into consideration were not included as convictions in England. However, Farrington and Wikstrm (1994) found that criminal careers of working-class males in England and Sweden were very similar, despite these differences in measurement. Data on social class (England). Study males were all living in a working-class inner-city area of South London at the time they were first contacted in 1961-1962. In nearly all cases (94%), their family breadwinner at that time (usually the father) had a working-class occupation (skilled, semiskilled, or unskilled manual worker). Therefore, in the comparative analyses, we compared the English cohort with working-class male offspring in Sweden.
RESULTS
INTERGENERATIONAL CRIME AND EFFECTS OF PARENTAL INCARCERATION (SWEDEN)

Having a convicted parent predicted offspring crime in Sweden (Table 1), consistent with many previous studies of continuity in crime across generations. However, parental incarceration was generally a stronger predictor of offspring crime than parental conviction without incarceration. For example, of offspring who had a parent incarcerated when they were aged 0 to 6, 25% offended between ages 19 and 30, compared with 18% of offspring who had a parent convicted (but not incarcerated). The odds ratios for offending for these two groups were 2.4 (95% Confidence Interval [CI] = 1.6, 3.5) for offspring of prisoners and 1.6 (CI = 1.1, 2.2) for offspring of convicted parents (using offspring of unconvicted parents as the comparison group). These two odds ratios were significantly different (p < .05).2 Parental incarceration (ages 0 to 6) also predicted higher rates of chronic offending than parental conviction. Odds ratios for chronic offending were 2.9 (CI = 1.8, 4.7) for offspring of incarcerated parents and 1.9 (CI = 1.2, 2.9) for offspring of convicted parents (using offspring of unconvicted parents as the comparison group). These two odds ratios were nearly significantly different (p = .07). Parental incarceration during childhood was a slightly stronger predictor of offspring offending than parental incarceration occurring before childrens births. For example, the odds ratio for offspring offending (ages 19 to 30) was 2.4 (CI = 1.6, 3.5) for parental incarceration when offspring were aged 0 to 6, compared with 1.8 (CI = 1.3, 2.4) for parental incarceration
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

TABLE 1: Effects of Parental Criminality and Incarceration by Offspring Age (Sweden)

Child 0 to 6 Child 7 to 19

Before Child Born

Offspring Offending
310 12.3 1.8* (1.3, 2.6) 19.4 1.8** (1.3, 2.4) 9.4 1.9* (1.3, 2.8) 5.3 Reference group 9.4 1.9** (1.2, 2.9) 12.2 Reference group 17.9 1.6* (1.1, 2.2) 24.8 2.4** (1.6, 3.5) 13.9 2.9** (1.8, 4.7) 7.2 Reference group 11.7 1.7* (1.1, 2.6) 18.2 2.9** (1.9, 4.5) 14,757 223 137 14,516

Parent Not Convicted

Parent Convicted Only

Parent Convicted and Incarcerated

Parent Not Convicted

Parent Convicted Only

Parent Convicted and Incarcerated

Parent Not Convicted

Parent Convicted Only

393

Parent Convicted and Incarcerated

208

14,217

590

Offended up to 19 (%) Odds ratio (95% CI)

7.1 Reference group

9.8 1.4* (1.1, 1.9)

Offended 19 to 30 (%) Odds ratio (95% CI)

11.9 Reference group

20.8 1.9** (1.6, 2.4)

12.0 Reference group 5.0 Reference group

21.6 2.0** (1.6, 2.6) 13.7 3.0** (2.2, 4.0)

26.0 2.6** (1.9, 3.5) 14.9 3.3** (2.2, 4.9)

Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Chronic offender (%) Odds ratio (95% CI)

5.2 Reference group

9.3 1.9** (1.4, 2.5)

Note. = offended up to age 19 is not shown for parental convictions when offspring were aged 7 to19 because we wanted explanatory variables to be genuinely predictive of outcomes; CI = confidence interval. *p < .05. **p < .001.

139

140

CRIMINAL JUSTICE AND BEHAVIOR

only before offspring were born, although the difference was not significant (p = .10). The timing of parental incarceration during offsprings lives did not seem important. For offspring whose parents were incarcerated when they were aged 0 to 6, the odds ratio for offending (ages 19 to 30) was 2.4 (CI = 1.6, 3.5), compared with 2.6 (CI = 1.9, 3.5) for offspring whose parents were incarcerated when they were aged 7 to 19 (the two odds ratios were not significantly different: p = .70). Because parental incarceration had similar effects in different periods of childhood, for the remainder of the analyses we examined effects of parental incarceration occurring at any time during childhood (ages 0 to 19).
EFFECTS OF PARENTAL INCARCERATION ON DIFFERENT TYPES OF CRIME (SWEDEN)

We investigated whether parental incarceration (ages 0 to 19) predicted some types of crime better than others in Sweden. Parental incarceration predicted different types of crime (from 19 to 30) with similar strength, reflecting the versatility of criminal careers. For example, the odds ratio associated with parental incarceration from 0 to 19 (compared with no parental incarceration ever) was 2.7 (CI = 1.7, 4.1) for violence and 3.0 (CI = 2.1, 4.2) for theft. This concurred with Murray and Farringtons (2005) finding that parental incarceration predicted self-reported delinquency and violence with similar strength. Odds ratios associated with parental incarceration for five other types of crime were 3.2 (CI = 2.1, 4.9) for fraud, 2.5 (CI = 1.7, 3.5) for traffic offenses, 2.7 (CI = 1.6, 4.6) for drug offenses, 1.6 (CI = 0.8, 3.2) for vandalism, and 2.3 (CI = 1.6, 3.4) for other offenses.
PARENTAL INCARCERATION AND OFFSPRING OFFENDING: A DOSE-RESPONSE RELATIONSHIP? (SWEDEN)

To investigate whether there was a dose-response relationship between parental incarceration and offspring offending, we examined the mean number of offenses of offspring according to the number of times their parents had been incarcerated. In Sweden, the more times offsprings parents were incarcerated (before offspring were age 19), the more offenses offspring were likely to commit (ages 19 to 30; Figure 1). The correlation coefficient between the number of times parents were incarcerated and the number of childrens offenses was significant (p < .001) but small (r = .06). However, the different distributions of the two variables might have accounted for this low correlation coefficient (13,241 children had no criminal record and 14,589 children had parents who had not been to prison). We calculated that the maximum possible r was .10 given these two distributions.3 Dividing the actual r (.06) by the maximum r (.10) gave an adjusted r value of .62, which perhaps better represents the nature of the dose-response relationship. Farrington and Loeber (1989) showed that in 2 2 tables, the actual r divided by the maximum possible r was mathematically identical to the widely accepted measure of strength of relationship termed Relative Improvement Over Chance.
SEX AND CLASS DIFFERENCES IN THE EFFECTS OF PARENTAL INCARCERATION (SWEDEN)

We examined whether the effects of parental incarceration differed according to offsprings sex and social class in Sweden. Parental incarceration predicted offending among females and males and among working- and middle-upper-class offspring (Table 2). For the two social classes, the effects of incarceration were similar (there was no significant difference between
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS

141

5 Mean offences of offspring (aged 19 to 30)

4.98

3 2.29 2.07 2 1.62

0.77

0 0 (n = 14,589) 1 (n = 316) 2 (n = 86) 3-4 (n = 66) 5+ (n = 60)

Number of times parents incarcerated (before offspring 19) Figure 1: Dose-Response Relationship Between Parental Incarceration and Offspring Offenses (Sweden)

TABLE 2: Effects of Parental Incarceration by Offspring Sex and Class (Sweden)

Offspring Offending n (parent incarcerated) n (parent not incarcerated) Offended 19 to 30 Odds ratio (95% CI) Chronic offender Odds ratio (95% CI)

Male
136 7,459 2.7** (1.9, 3.9) 3.0** (2.0, 4.5)

Female
147 7,130 3.3** (1.9, 5.7) 5.5** (2.7, 11.2)

Working Class
180 5,806 2.1** (1.5, 2.9) 2.4** (1.6, 3.6)

Middle-UpperClass
103 8,783 2.3** (1.4, 3.8) 3.1** (1.6, 5.8)

Note. Odds ratios compare offspring whose parents were incarcerated (offspring aged 0 to 19) and offspring whose parents were never incarcerated. CI = confidence interval. *p < .05. **p < .001.

odds ratios). However, the effects of parental incarceration were greater for females than for males. The odds ratio associated with parental incarceration was 5.5 (CI = 2.7, 11.2) for female chronic offending, compared with 3.0 CI = 2.0, 4.5) for male chronic offending (the difference between the two odds ratios was significant: p < .05). This finding was contrary to our hypothesis that males would be affected more by parental incarceration than females. It is possible that females appeared to be affected more because of the low number of female offenders. Of females in Project Metropolitan, only 4% offended between ages 19 and 30, compared with 21% of males. Perhaps the few female offenders were more likely to have
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

142

CRIMINAL JUSTICE AND BEHAVIOR

TABLE 3: Effects of Parental Incarceration Before and After Controlling for Parental Convictions (Sweden)

Entering Prison Second Step 1 Parent Convictions Offspring Offending

Offended 19 to 30 Chronic offender

Entering Convictions Second Step 1 Parent Prison LRCS

35.01 31.07 p < .001 < .001

Step 2 Parent Prison LRCS

1.43 2.06 p .232 .152

Step 2 Parent Convictions LRCS

19.88 14.39 p < .001 < .001

LRCS
53.46 43.40

p < .001 < .001

Note. LRCS = likelihood ratio chi-square. Parent prison variable = parent incarcerated (offspring aged 0 to 19) compared with parent never incarcerated. Parent convictions = number of parental convictions before offspring 19.

deviant backgrounds (including incarcerated parents) than the more common male offenders, thereby producing a stronger association between parental incarceration and offending for females.
PARENTAL INCARCERATION: AN INDEPENDENT PREDICTOR OF OFFSPRING OFFENDING? (SWEDEN)

Although parental incarceration predicted offsprings own offending, this might have been because incarcerated parents were highly criminal, not because incarceration had specific effects on offspring. We tested whether parental incarceration predicted offsprings offending independently of parents convictions, using stepwise logistic regression. In the first step, we entered the number of parents convictions, which significantly improved the prediction of offspring offending (Column 1, Table 3). In the second step, we added the parental incarceration variable. In fact, parental incarceration did not significantly improve the prediction of offspring offending after parental convictions were controlled for (Column 2, Table 3). By contrast, when parental incarceration was entered into the model first and then parental convictions, parental convictions significantly improved the prediction of offspring offending (Columns 3 and 4, Table 3).4 Together, these results suggest that parental incarceration did not directly affect offspring offending; rather, parental criminality explained the link between parental incarceration and offspring offending in Sweden.
COMPARING EFFECTS OF PARENTAL INCARCERATION IN SWEDEN AND ENGLAND

Parental incarceration in Sweden did not predict offspring offending after controlling for parental convictions. This contrasted with Murray and Farringtons (2005) previous finding that parental incarceration in England was an independent predictor. Next, we compared, more closely, the effects of parental incarceration in England and Sweden. In both countries, we compared offspring who experienced parental incarceration during childhood with offspring whose parents were incarcerated only before their births. Using offspring whose parents were incarcerated only before their births as the control group should isolate effects of the experience of parental incarceration (which we hypothesize has specific effects on children) from background factors. We also controlled for the number of parental convictions using logistic regression (because parents incarcerated only before childrens births had fewer offenses than parents incarcerated after childrens births).
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS TABLE 4: Effects of Parental Incarceration on Male Offspring Offending in Sweden and England

143

Sweden POR Controlling for Parental Convictions

1.3 (0.7, 2.6) 1.4 (0.4, 4.6)

England POR Controlling for Parental Convictions

3.2* (0.9, 12.0) 2.6 (0.2, 27.8)

Offspring Offending
Offended 19 to 30 (95% CI) Chronic offender (95% CI)

OR
1.3 (0.7, 2.5) 2.0 (0.7, 6.0)

OR
4.1** (1.1, 14.5) 5.1 (0.6, 45.6)

Note. OR = odds ratio comparing parental incarceration when offspring aged 0 to 19 with parental incarceration before offsprings births; POR = partial odds ratio; CI = confidence interval. *p < .10. **p < .05.

We conducted the same comparison in England and Sweden, matching offspring as closely as possible on sex (male), social class (working class), timing of parental incarceration (any time during childhood, compared with before birth), and the timing of offsprings outcomes.5 In the English cohort, 29 offspring had a parent incarcerated during childhood, and 17 had a parent incarcerated only before their birth (out of 411). In the Swedish cohort, 92 offspring had a parent incarcerated during childhood, and 72 had a parent incarcerated only before their birth (out of 3,191 working-class males). The effects of parental incarceration during childhood were much stronger in England than in Sweden (Table 4). For example, in England, the odds ratio for offspring offending was 4.1 (CI = 1.1, 14.5), compared with 1.3 (CI = 0.7, 2.5) in Sweden (the difference between the size of the odds ratios was significant ( p < .01). Even after controlling for the number of parental convictions, the effects of parental incarceration on offspring offending were stronger in England than in Sweden (the difference between the partial odds ratios was significant: p < .05). In conclusion, parental incarceration was a stronger risk factor for offending in England than in Sweden.
EXPLAINING DIFFERENT EFFECTS OF PARENTAL INCARCERATION IN SWEDEN AND ENGLAND

Why was parental incarceration a stronger risk factor for offending in England than in Sweden? The difference might be explained either by mechanisms linking parent and offspring criminality (Farrington, 2002) or by mechanisms more specific to parental incarceration (such as parent-child separation or distressing prison visits). If our results are partly explained by mechanisms linking parent and offspring criminality, there should also be a difference in the effects of parental conviction (without incarceration) between England and Sweden. To examine this hypothesis, we compared the effects of parental conviction (without incarceration) on offspring offending in England and Sweden. The effects of parental conviction (without incarceration) also appeared stronger in England than in Sweden (Table 5). For chronic offending, the odds ratio associated with parental conviction in England was significantly higher than the odds ratio in Sweden ( p < .001). Therefore, the difference between the effects of parental incarceration in England and Sweden might partly be explained by mechanisms linking parent and offspring criminality.
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

144

CRIMINAL JUSTICE AND BEHAVIOR

TABLE 5: Effects of Parental Conviction (Without Incarceration) on Male Offspring Offending in Sweden and England

Offspring Offending
Offended 19 to 30 (95% CI) Chronic offender (95% CI)

Sweden OR
1.8* (1.4, 2.3) 1.8* (1.2, 2.7)

England OR
2.3* (1.3, 3.9) 6.0* (2.3, 15.1)

Note. OR = odds ratio comparing parental conviction (without incarceration) when offspring aged 0 to 19, with parents never convicted; CI = confidence interval. *p < .01.

The different effects of parental incarceration in Sweden and England might also have occurred because parents were incarcerated for shorter lengths of time in Sweden than in England, and shorter sentences may have less harmful effects on offspring than longer ones. Farrington, Langan, and Wikstrm (1994) found that average sentence lengths and time served were generally longer in England than in Sweden (and generally longer in the United States than in England). In Sweden, a large proportion of people who were incarcerated from 1950 to 1970 were sentenced for drunk-driving offenses and received relatively short sentences. For example, of those who were received into Swedish penal institutions in 1970, 57% were admitted for less than 2 months (Ward, 1972). Data were not available in Project Metropolitan on the length of parental incarceration, but they were in the Cambridge Study. We compared male offspring in England whose parents were incarcerated for long periods (more than 2 months when offspring were 0 to 19, n = 17, M = 18.8 months, SD = 21.4) and short periods (less than 2 months, n = 14, M = 1.1 months, SD = 0.36) to test whether longer periods in custody had worse effects on offspring.6 Of offspring whose parents were in custody for less than 2 months, 7% became chronic offenders, compared with 35% of offspring whose parents were held in custody for more than 2 months (n = 31, 2 = 3.48, df = 1, p = .062). The trend was in the same direction for offspring offending (50% and 77%, respectively, offended between ages 19 and 30), although the difference was not significant. These figures support the hypothesis that parental incarceration had stronger effects in England because parents were incarcerated for longer periods. In addition, the finding supports our dose-response hypothesis. However, the finding could be attributable to a selection effect, if longer sentence prisoners were more antisocial than shorter sentence prisoners.
DISCUSSION

This study investigated the intergenerational transmission of criminality and the effects of parental incarceration on children in Sweden and England using prospective longitudinal data from Project Metropolitan and the Cambridge Study in Delinquent Development. In Sweden, parental criminality and parental incarceration predicted offspring offending through the life course. There was a dose-response relationship between the number of times parents were incarcerated and the number of times offspring offended in adulthood. However, the most striking finding was that in Sweden, the effects of parental incarceration disappeared after controlling for parental criminality, but the effects remained strong in England. Why might parental incarceration predict offspring offending over and above associated risks in England but not in Sweden?
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS

145

First, we investigated the hypothesis that parental incarceration was less harmful in Sweden than in England because prison sentences were shorter in Sweden. Supporting this hypothesis, there was some evidence that longer periods of parental incarceration predicted higher rates of offending in England. However, other explanations for the difference in the effects of parental incarceration between Sweden and England are possible. For example, Swedish prison policies might have mitigated the effects of parental incarceration on offspring. In the 1960s, approximately one third of Swedish prisoners were held in open prisons; prisoners had a right to home leave every 3 or 4 months after serving a fixed proportion of their sentence, prisoners had the opportunity to communicate with their family by telephone and uncensored mail, and although the institution of conjugal visits was not officially prescribed, private visits commonly occurred in Swedish prisons (Morris, 1966; Ward, 1979). By contrast, in England and Wales, only 3% of the sentenced population was granted home leave in 1959 (Kelemenis, 1999), and there was (and still is) no practice of conjugal or private visiting. Assuming that parental incarceration causes offspring offending partly because of parent-child separation and disrupted prisoner-family relationships, family-friendly prison policies in Sweden might have protected offspring from some of the adverse effects of parental incarceration. However, parental conviction without incarceration also conferred more risk on offspring in England than in Sweden. Therefore, there might be wider social reasons for the differences found between Sweden and England. As a third explanation, the Swedish justice system might have prevented the escalation of criminal careers following parental incarceration. Juvenile justice systems were very different in Sweden and England in the 1950s and 1960s. In Sweden, child welfare, rather than punishment, was the paramount concern in cases of child delinquency (see Janson, 2004, for an overview of the Swedish youth justice system). Delinquency cases in Sweden were handled by an extensive system of Child Welfare Committees, each of which included a member of the poor-law board, a clergyman, a schoolteacher, and a physician. By contrast, in England until the late 1960s, juvenile delinquents were dealt with in courts that were similar to adult criminal courts (Bottoms, 2002). (For a comparison of Swedish and U.S. justice systems at that time, see Marnell, 1972.) Therefore, it is possible that the effects of parental incarceration were mitigated by the welfare-oriented juvenile justice system in Sweden or even exacerbated by the juvenile justice system in England. Also, it is possible that there was more bias in the English justice system. Perhaps, in England, offspring were more likely to be arrested, prosecuted, or convicted on the basis of having criminal parents than in Sweden. Fourth, it is possible that social welfare policies in Sweden in the 1950s and 1960s protected offspring from some of the adverse effects of parental incarceration. Particularly in the postwar period, Sweden was one of the most egalitarian states, where the basic needs of all Swedish citizens for a minimum income, job security, satisfactory health care, educational opportunities, adequate housing, good public transportation, and other social services have been met (Ward, 1979, p. 152). In Sweden in the 1950s, the lower end of the income distribution was almost truncated, and families with children received a general child allowance and housing allowance (Janson & Wikstrm, 1995; Olsson, 1993). Notably, although working-class families might have fewer resources to cope with the costs of parental incarceration than middle-class families, in our study of Swedish children, the effects of parental incarceration were similar across social classes. Swedens social welfare system might have protected offspring of prisoners (particularly working-class offspring) from economic and social adversities and mitigated the effects of parental incarceration on offspring.
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

146

CRIMINAL JUSTICE AND BEHAVIOR

A fifth possible explanation for the difference between the two countries is that in Sweden, public opinion on crime and punishment might have reduced the effects of parental incarceration on offspring. Although this is not based on quantified public surveys, visitors to Sweden in the 1960s and 1970s were often impressed by the sympathetic attitudes toward crime and punishment they found there. In 1966, Norval Morris noted, the Swedish criminal or prisoner still remains a Swedish citizen meriting respect, continuing properly to enjoy a quite high standard of living and remaining part of the community (p. 5; see also Friday, 1976; Janson & Wikstrm, 1995). Unlike in England, the identity of Swedish offenders was almost never revealed in media accounts of trials and convictions (Walker, 1980, pp. 151-152). Moreover, by 1972 prisoners had the right to vote in general elections, unlike in England and the United States (Marnell, 1972). Incarceration might also have been associated with less social stigma in Sweden because many prisoners were incarcerated for drunk-driving offenses, and consequently, middle-class people were more at risk of imprisonment in Sweden than in England. Sixth, the different effects of parental incarceration in Sweden and England might be explained by different levels of exposure to delinquent peers. Parental incarceration might represent a risk for offspring only to the extent that they are exposed to delinquent peers. If delinquency were more widespread in England than in Sweden, this could account for different effects of parental incarceration between the two countries. However, the same proportion of working-class males offended in the Swedish cohort (26%) as in the English cohort (25%). Different exposures to delinquent norms are unlikely to account for our results. It is also possible that the different findings in Sweden and England reflect different measurement of parental incarceration in the two studies. Criminal records were searched for both parents in England. In Sweden, usually only the fathers criminal records were searched, although some mothers were included. Other limitations of our study were that legal definitions varied somewhat between England and Sweden, we were unable to control for the seriousness of parents crimes, we could not test whether effects on offspring differed according to whether mothers or fathers were incarcerated, we were not able to test why parental incarceration appeared to affect females more than males, and cohorts in both countries were born just more than 50 years ago. We need replication of these results among todays prison populations, continuing research on the mechanisms linking parental incarceration and offspring criminality, and systematic comparison of the effects of maternal and paternal incarceration. Resiliency research suggests that children are protected from adversity by individual, family, and wider environmental factors, such as above average IQ, parental attachment and bonding, temperament and behavior factors, and positive peer relations (Garmezy, 1985; Luthar, 2003; Rutter, 1990). It is important for future research to identify protective factors among offspring of prisoners. In conclusion, there are several possible reasons why parental incarceration was a stronger predictor of offspring offending in England than in Sweden. However, to date, there has been no rigorous test of the mechanisms linking parental incarceration and adverse child outcomes. Until these mechanisms are better known, we can only speculate about why the effects of parental incarceration might differ between jurisdictions. Bronfenbrenner (1979) argued that child development can be enhanced by the adoption of public policies and practices that create additional settings and societal roles conducive to family life (p. 7). In Sweden, the combination of shorter prison sentences, prison policies that encouraged family contact, the advanced social welfare system, welfare-oriented juvenile justice policies, and sympathetic
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS

147

public attitudes toward criminal offenders could have protected children from harmful effects of parental incarceration. Future research could analyze the effects of parental incarceration in other jurisdictions according to their social and prison policies. For example, in the United States, effects of parental incarceration could be compared between states with different policies on prisonerfamily contact, average length of sentence, and social support provided to prisoners families. Given the huge number of children experiencing parental incarceration in the United States today (more than 1.5 million in 1999; Mumola, 2000), it is particularly urgent to investigate the effects of parental incarceration on these children. Parental incarceration might have even more damaging effects on children in the United States than in England because average time served is generally longer in the United States than in England (Farrington, Langan, & Tonry, 2004). The question this study provokes is whether children might be protected from negative effects of parental incarceration today by using social and prison policies from Sweden in the 1950s.
NOTES
1. Unfortunately, data on the sex of the parent whose criminal records were searched are not held in Project Metropolitan archives. Also, Project Metropolitan data have been deidentified. Therefore, it would be impossible to establish the sex of parents whose criminal records were searched. Our best interpretation of the notes kept on parental criminality in Project Metropolitan is that the vast majority of criminal records referred to fathers, but a few mothers were also included. 2. Comparisons of odds ratios throughout this paper are based on the formula: Z = (LOR1 - LOR2) / spooled where LOR1 = Ln (1st odds ratio), LOR2 = Ln (2nd odds ratio), and spooled = the pooled standard error of LOR1 and LOR2. The pooled standard error (spooled) is derived fromthe formula: S pooled =
(n1 1)s12 + (n2 1)s22 . , where n1 = sample size of group 1, n2 = n1 + n2 2

sample size of group 2, s1 = standard error for LOR1, and s2 = standard error for LOR2. 3. The maximum possible r was calculated by cross-tabulating the number of parental incarcerations and the number of offspring offenses. As many cases as possible (given the column and row totals) were placed on the diagonal of the table (0,0; 1,1; 2,2, etc.). Other cases were placed, unambiguously, as close to the diagonal as possible (given the remaining row and column totals). The maximum r possible was calculated from this distribution. 4. In these analyses, the parental convictions variable was ratio level (number of convictions), and the parental incarceration variable was dichotomous (parents were incarcerated or not). To test whether these results merely reflected more sensitive measurement of convictions, we also conducted the analyses with a dichotomous parental convictions variable (split at the same proportion as the parental incarceration variable). The results were similar: Parental convictions predicted child delinquency controlling for parental incarceration, but not vice versa. 5. In these analyses, chronic offenders in Sweden were identified as the 7% of working-class male offspring with the most convictions (ages 19 to 30) to compare with the 7% chronic offenders in London. 6. Two cases previously excluded because parents were only incarcerated for 1 day were included in these analyses.

REFERENCES
Anderson, N. N. (1966). Prisoners families: A study of family crisis. Minneapolis: University of Minnesota Press. Boswell, G., & Wedge, P. (2002). Imprisoned fathers and their children. London: Jessica Kingsley. Bottoms, A. (2002). The divergent development of juvenile justice policy and practice in England and Scotland. In M. K. Rosenheim, F. E. Zimring, D. S. Tanenhaus, & B. Dohrn (Eds.), A century of juvenile justice (pp. 413-504). Chicago and London: University of Chicago Press. Bowlby, J. (1969). Attachment and loss Vol. I: Attachment. London: Hogarth Press and the Institute of Psycho-Analysis. Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press. Farrington, D. P. (1995). The development of offending and antisocial behaviour from childhood: Key findings from the Cambridge Study in Delinquent Development. Journal of Child Psychology and Psychiatry, 36, 929-964. Farrington, D. P. (2002). Families and crime. In J. Q. Wilson & J. Petersilia (Eds.), Crime: Public policies for crime control (pp. 129-148). Oakland, CA: Institute for Contemporary Studies Press.
Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

148

CRIMINAL JUSTICE AND BEHAVIOR

Farrington, D. P. (2003). Key results from the first forty years of the Cambridge Study in Delinquent Development. In T. P. Thornberry & M. D. Krohn (Eds.), Taking stock of delinquency: An overview of findings from contemporary longitudinal studies (pp. 137-183). New York: Kluwer Academic/Plenum. Farrington, D. P., Jolliffe, D., Loeber, R., Stouthamer-Loeber, M., & Kalb, L. M. (2001). The concentration of offenders in families, and family criminality in the prediction of boys delinquency. Journal of Adolescence, 24, 579-596. Farrington, D. P., Langan, P. A., & Tonry, M. (Eds.). (2004). Cross-national studies in crime and justice. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Farrington, D. P., Langan, P. A., & Wikstrm, P.- O. H. (1994). Changes in crime and punishment in America, England and Sweden between the 1980s and the 1990s. Studies on Crime and Crime Prevention, 3, 104-131. Farrington, D. P., & Loeber, R. (1989). Relative improvement over chance (RIOC) and phi as measures of predictive efficiency and strength of association in 2 2 tables. Journal of Quantitative Criminology, 5, 201-213. Farrington, D. P., & Loeber, R. (2000). Some benefits of dichotomization in psychiatric and criminological research. Criminal Behaviour and Mental Health, 10, 100-122. Farrington, D. P., & Wikstrm, P.- O. H. (1994). Criminal careers in London and Stockholm: A cross-national comparative study. In E. G. M. Weitekamp & H.-J. Kerner (Eds.), Cross-national longitudinal research on human development and criminal behavior (pp. 65-89). Dordrecht, Denmark: Kluwer. Fergusson, D. M., Horwood, L. J., & Nagin, D. (2000). Offending trajectories in a New Zealand birth cohort. Criminology, 38, 525-551. Friday, P. C. (1976). Sanctioning in Sweden: An overview. Federal Probation, 40, 48-55. Friedman, S., & Esselstyn, T. C. (1965). The adjustment of children of jail inmates. Federal Probation, 29, 55-59. Gabel, S. (1993). Characteristics of children whose parents have been incarcerated. Hospital and Community Psychiatry, 44, 656-660. Garmezy, N. (1985). Stress-resistant children: The search for protective factors. In J. E. Stevenson (Ed.), Recent research in developmental psychopathology. Journal of Child Psychology and Psychiatry book supplement (Vol. 4, pp. 213-233). Oxford: Pergamon. Glueck, S., & Glueck, E. (1950). Unraveling juvenile delinquency. Cambridge, MA: Harvard University Press. Hodgins, S., & Janson, C.- G. (2002). Criminality and violence among the mentally disordered: The Stockholm Metropolitan Project. Cambridge, UK: Cambridge University Press. Hounslow, B., Stephenson, A., Stewart, J., & Crancher, J. (1982). Children of imprisoned parents. Sydney: Ministry of Youth and Community Services of New South Wales. Janson, C.- G. (1975). Project Metropolitan Research Report No. 2: A description of its data archive. Stockholm, Sweden: University of Stockholm, Department of Sociology. Janson, C.- G. (2000). Project Metropolitan. In C.- G. Janson (Ed.), Seven Swedish longitudinal studies in behavioral sciences: Vol. 8 (pp. 140-171). Stockholm, Sweden: Forskningsradsnamuden. Janson, C.- G. (2004). Crime and justice, Vol. 31. Youth justice in Sweden. In M. Tonry & A. N. Doob (Eds.), Youth crime and youth justice: Comparative and cross-national perspectives (pp. 391-442). Chicago and London: University of Chicago Press. Janson, C.- G., & Wikstrm, P.- O. H. (1995). Growing up in a welfare state. In J. Hagan (Ed.), Delinquency and disrepute in the life course. Vol. 4 in current perspectives on aging and the life cycle (pp. 191-215). Greenwich, CT: JAI Press. Johnston, D. (1995). Effects of parental incarceration. In K. Gabel & D. Johnston (Eds.), Children of incarcerated parents (pp. 59-88). New York: Lexington. Kelemenis, I. (1999). Between captivity and freedom (doctoral dissertation). Oxford, UK: Oxford University Press. Lowenstein, A. (1986). Temporary single parenthood - the case of prisoners families. Family Relations, 35, 79-85. Luthar, S. S. (Ed.). (2003). Resilience and vulnerability: Adaptation in the context of childhood adversities. Cambridge, UK: Cambridge University Press. Marnell, G. (1972). Comparative correctional systems: United States and Sweden. Criminal Law Bulletin, 7, 748-760. McCord, W., McCord, J., & Zola, I. K. (1959). Origins of crime: A new evaluation of the Cambridge-Somerville Youth Study. New York: Columbia University Press. Morris, N. (1966). Lessons from the adult correctional system of Sweden. Federal Probation, 30, 3-13. Mumola, C. J. (2000). Incarcerated parents and their children. Washington, DC: Bureau of Justice Statistics. Murray, J. (2005). The effects of imprisonment on families and children of prisoners. In A. Liebling & S. Maruna (Eds.), The effects of imprisonment (pp. 442-492). Cullompton, Devon, UK: Willan. Murray, J., & Farrington, D. P. (2005). Parental imprisonment: Effects on boys antisocial behaviour and delinquency through the life-course. Journal of Child Psychology and Psychiatry, 46, 1269-1278. Olsson, S. E. (1993). Social policy and welfare state in Sweden. Lund, Sweden: Arkiv. Richards, M., McWilliams, B., Allcock, L., Enterkin, J., Owens, P., & Woodrow, J. (1994). Occasional Papers No. 2. The family ties of English prisoners: The results of the Cambridge Project on Imprisonment and Family Ties. Cambridge, UK: Centre for Family Research, Cambridge University. Robins, L. N., West, P. A., & Herjanic, B. L. (1975). Arrests and delinquency in two generations: A study of black urban families and their children. Journal of Child Psychology and Psychiatry, 16, 125-140.

Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008

Murray et al. / CRIME IN ADULT OFFSPRING OF PRISONERS

149

Rodgers, B., & Pryor, J. (1998). Divorce and separation: The outcomes for children. York, UK: Joseph Rowntree Foundation. Rutter, M. (1990). Psychosocial resilience and protective mechanisms. In J. Rolf, A. S. Masten, D. Cicchetti, K. H. Nuechterlein, & S. Weintraub (Eds.), Risk and protective factors in the development of psychopathology (pp. 181-214). Cambridge, UK: Cambridge University Press. Sack, W. H. (1977). Children of imprisoned fathers. Psychiatry, 40, 163-174. Sack, W. H., Seidler, J., & Thomas, S. (1976). The children of imprisoned parents: A psychosocial exploration. American Journal of Orthopsychiatry, 46, 618-628. Shaw, R. (1987). Children of imprisoned fathers. Bungay, Suffolk, UK: Richard Clay. Smith, W. R. (1991). Project Metropolitan Research Report No. 33: Social structure, family structure, child rearing, and delinquency: Another look. Stockholm, Sweden: Department of Sociology, University of Stockholm. Stanton, A. M. (1980). When mothers go to jail. Lexington, MA: Lexington. Walker, N. (1980). Punishment, danger and stigma: The morality of criminal justice. Oxford, UK: Blackwell. Ward, D. A. (1972). Inmate rights and prison reform in Sweden and Denmark. Journal of Criminal Law, Criminology and Police Science, 63, 240-255. Ward, D. A. (1979). Sweden: The middle way to prison reform? In M. E. Wolfgang (Ed.), Prisons: Present and possible (pp. 89-167). Lexington, MA: Lexington. West, D. J., & Farrington, D. P. (1977). The delinquent way of life. London: Heinemann. Wikstrm, P.- O. H. (1987). Project Metropolitan Research Report No. 24: Patterns of crime in a birth cohort. Age, sex and social class differences. Stockholm, Sweden: Stockholm University, Department of Sociology.

Downloaded from http://cjb.sagepub.com at Uni Transilvania Brasov on September 19, 2008