The Juvenile Sex Offender

Abstract

In this article, we investigate whether the life events of marriage, parenthood, and employment were associated with general offending for a Dutch sample of 498 juvenile sex offenders (JSOs). In previous empirical studies, these life events were found to limit adult general offending in the population as well as high-risk samples. A hybrid random effects model is used to investigate within-individual changes of these life events in association with general offending. We also investigated whether the findings differed for child abusers, peer abusers, and group offenders, as they have distinct background profiles. We found that JSOs make limited transitions into the state of marriage, parenthood, and employment, showing overall stagnating participation rates. For the entire sample of JSOs, employment was found to be associated with a decrease in offending. Group offenders benefited most from employment. Marriage and parenthood were not associated with the general offending patterns, whereas for child abusers, parenthood was associated with an increase in offending. We conclude that policies aimed at guidance toward employment, or inclusion into conventional society, may be effective for JSOs.

Keywords

juvenile sex offender life-course criminology sex offender typology transitions sex offender policies

Numerous life-course studies have demonstrated that offending follows a typical pattern across age. On average, a peak in offending occurs in adolescence, and a gradual decline sets in soon afterwards in early adulthood (Farrington, 1986; Piquero, Farrington, & Blumstein, 2003). This decline has been explained through life-changing events taking place in early adulthood, such as leaving the parental home, getting married, becoming a parent, and finding employment (Arnett, 2004). Assuming such adult roles generates embeddedness in conventional society, increasing responsibility for the self as well as others (Sampson & Laub, 1993). Delinquency then becomes less appealing as it jeopardizes the newly gained responsibilities and status in society (see also Hirschi, 1969). Thus, life events associated with the transition to adulthood can be regarded as a protective factor in adulthood, which can reduce offending. Several studies found empirical support for the function of these life events, both in population-based and high-risk samples; the individuals in this sample are considered at high risk of re-offending due to adverse background characteristics (e.g., Blokland & Nieuwbeerta, 2005; Sampson & Laub, 1993; Savolainen, 2009; van der Geest, 2011). To date, no studies have examined how these life events affect the adult criminal career of individuals convicted for a sexual offense as juveniles.

A number of recent studies have investigated the adult criminal careers of (juvenile) sex offenders (e.g., Caldwell, 2002; Letourneau & Miner, 2005; Lussier & Blokland, 2014; Lussier, van den Berg, Bijleveld, & Hendriks, 2012; van den Berg, Bijleveld, & Hendriks, 2011; Zimring, Jennings, Piquero, & Hays, 2009; Zimring, Piquero, & Jennings, 2007). These studies mostly revealed that the criminal career patterns of juvenile sex offenders (JSOs) resemble those of other offenders, with little sexual recidivism and a gradual decline in offending after adolescence. Therefore, it seems reasonable to presume that the decrease found for general offending in JSOs’ adult criminal careers can be explained by the same protective factors as were found for population-based as well as high-risk samples.

Studying this assumption is important as sex offenders are put under increasingly strict surveillance and registration regimes (Boone, 2011; Lasher & McGrath, 2012; Tewksbury, 2005), despite the base rate for sexual re-offending being low. These criminal justice responses ostracize (juvenile) sex offenders and generate difficulties for them in finding housing, employment, and possibly also a romantic partner (Tewksbury, 2005). Therefore, policies intended to protect society may in fact be counter-productive if (juvenile) sex offenders are unable to benefit from experiencing certain life events. However, it is unknown whether such life events affect JSOs in the same way they affect offenders in population-based and high-risk samples. This study aims to expand our knowledge of the development of JSOs over the life course by examining three protective adult life events (marriage, parenthood, and employment) and the association with adult general offending.

Theoretical Background and Empirical Findings

For most offenders, the decline in offending sets in during the same period that adult “turning points” occur, such as work, marriage, and parenthood. Therefore, these events have been hypothesized as a potential explanation for the reduction of, and desistance from, offending (Sampson & Laub, 1993). However, the underlying mechanisms for this hypothesized association differ per theory and life event.

Marriage

In line with the age-graded theory of social control, Sampson, Laub, and Wimer (2006) proposed several processes underlying the association between marriage and offending. First, in marriage, there is always the responsibility for a spouse. Second, this spouse heightens direct and indirect supervision and control. Third, daily routines are altered by marriage as spouses spend time together, causing less interaction with deviant peers, and finally a cognitive transformation may occur that will change one’s identity into that of a spouse.

The current body of empirical literature does provide fairly convincing evidence for the association between marriage and general offending (e.g., Bersani, Laub, & Nieuwbeerta, 2009; Blokland & Nieuwbeerta, 2005; Laub & Sampson, 2003; Sampson et al., 2006; Warr, 1998). Using a rich official registration data set which controlled for individual characteristics as well as time-varying covariates, Zoutewelle-Terovan, van der Geest, Liefbroer, and Bijleveld (2014) found that marriage promoted desistance from serious offending. Horney, Osgood, and Haen Marshall (1995) also used a within-individual analysis to model the month-to-month changes in the life circumstances reported by previously incarcerated offenders and found that living with a spouse significantly reduced the likelihood of committing an offense. Using stratified marriage-age cohorts, Theobald and Farrington (2009) found that marriage before the age of 25 was associated with lower levels of self-reported offending. Only a small number of studies did not find any association of marriage on general offending (e.g., Giordano, Cernkovich, & Rudolph, 2002; Knight, Osborn, & West, 1977). This may be attributable to studies investigating more recent cohorts, where cohabitation has overtaken marriage as a dominant first relationship type in many Western countries. For example, Skardhamar and Lyngstad (2009) found no decline in official offending rates after marriage, yet they found that a gradual decline already sets in during the period prior to marriage. The authors suggest that the decline was due to the offenders being in stable relationships or cohabitating prior to marriage.

Parenthood

The mechanisms underlying the association between parenthood and offending are comparable with those assumed for marriage. Becoming a parent brings new responsibilities (Sampson & Laub, 2003), changes daily routines, and limits time spent with deviant peer groups (Warr, 1998). Parenthood is also associated with cognitive transformation, as the identity change associated with becoming a parent is often not compatible with the old identity that facilitated offending (Giordano et al., 2002).

The existing empirical research shows contradictory results for the association between parenthood and general offending. A number of studies have found that parenthood is linked to increased general offending. Farrington and West (1995) suggested that becoming a parent out of wedlock increases offending, and that any decrease would be attributable to the presence of a partner. In line with this finding, Zoutewelle-Terovan et al. (2014) found that what they phrased the “full family package” (marriage and parenthood) had a strong negative influence on official registered offending, indicating that the association of parenthood and offending may be conditional on marriage. However, other studies found very limited evidence for the influence of parenthood on general offending desistance (Blokland & Nieuwbeerta, 2005). Skardhamar and Lyngstad (2009) found that in the period prior to parenthood (or marriage), officially registered offending was already lower. Yet in the period following parenthood, an increase in offending for cohabiting and married men was found.

Employment

Many general criminological theories postulate that employment reduces offending (for an overview, see Uggen & Staff, 2001). The routine activities theory assumes that employment limits time and opportunities to offend (Cohen & Felson, 1979). Furthermore, Sutherland and Cressey (1978) hypothesized that employment reduces offending through social values learned from co-workers. In line with age-graded theory of social control by Sampson and Laub (1993), some theories assume a more gradual influence of employment on crime by stating that the more time a person spends being employed, the more he or she identifies with a non-criminal lifestyle (Crutchfield & Pitchford, 1997; Maruna, 2001).

Again, the empirical evidence for the association between employment and general offending is mixed. Sampson and Laub (2003) found stable employment to be one of the most salient life events in reducing general offending. However, qualitative studies found no evidence for desistance or a reduction of offending due to stable employment (see Visher, Winterfield, & Coggeshall, 2005). Visher et al. (2005) conducted a meta-analysis on the effect of employment programs aimed at reducing recidivism, and concluded that employment did not reduce offending. Longitudinal studies generally found employment to reduce general offending (Savolainen, 2009; Thornberry & Christenson, 1984; van der Geest, Bijleveld, & Blokland, 2011; Verbruggen, Blokland, & van der Geest, 2012). Some found the employment-offending association to be conditional on different factors, such as age (Paternoster, Bushway, Brame, & Apel, 2003; Uggen, 2000), employment stability (Crutchfield & Pitchford, 1997), job quality (Apel, Paternoster, Bushway, & Brame, 2006; van der Geest et al., 2011), or offense type (Piquero, MacDonald, & Parker, 2002). A small number of studies found no relationship between self-reported offending and employment (Horney et al., 1995; MacKenzie & De Li, 2002). Skardhamar and Savolainen (2012) however concluded that employment is a consequence of natural desistance rather than a cause. This could possibly explain the mixed empirical evidence found for the employment–offending association.

To our knowledge, employment is the only life event for which desistance of sexual offending has been examined. Kruttschnitt, Uggen, and Shelton (2000) found that in a sample of male adult sex offenders, the only aspects associated with a reduction in general recidivism rates were stable employment and sex offender treatment. Hanson and Harris (1998) found that the sex offenders with the highest re-offense risk were those who lacked stable employment. For JSOs, van den Berg, Bijleveld, Hendriks, and Mooi-Reci (2014) found a negative relationship between regular employment and general offending.

JSOs

For lack of empirical data on all three life events for JSOs, it can only be assumed that life events function in a similar way for JSOs as for other offenders. However, there are three reasons why such a generalization is unwarranted. First, it has been well documented in the literature that sex offenders suffer from more psychological and psychiatric disturbances than non-sex offenders (Smallbone, 2006). van Wijk, van Horn, Bullens, Bijleveld, and Doreleijers (2005) found JSOs to be significantly more neurotic and less extroverted compared with juvenile delinquents convicted for a non-sexual offense. In addition, high levels of anxiety, depression, and personality problems have been reported in juvenile sexual offenders (Seto & Lalumière, 2010). Moreover, several studies have found that the social skills of JSOs are much less well developed than those of non-sex offenders (Seto & Lalumière, 2010; van Wijk et al., 2005). Given that JSOs often have such “negative” traits, it is possible that these individuals experience problems finding a romantic partner and employment. Therefore, the participation of JSOs on the labor and marriage “market” may be limited due to these adverse background characteristics.

Second, despite evidence for a negative association between employment and offending, finding employment may also be difficult, because in many countries (including the Netherlands, where the current study was conducted), juvenile and adult sex offenders are by law excluded from numerous kinds of jobs (requiring a certificate of good conduct) in which they might spend unsupervised time with children or persons dependent on their care. This rule comprises a very broad range of jobs such as taxi driver, bus driver, concierges in schools, caregiver jobs, and the like. Sex offenders are therefore excluded from a wide pool of jobs, with fewer employment opportunities than for other offenders.

Third, there are indications that sex offenders experience difficulties taking on adult roles, because of the stigma attached to sex offending. In spite of empirical evidence to the contrary, policy makers, law-makers, and the general public believe that sex offending is a marker for sexual deviance and an ongoing high risk for sexual recidivism. Although it is unknown to what extent this stigma would complicate employment, it is likely that it would complicate relationship formation. For example, Beijers and Bijleveld (2012) conducted a vignette study in which they asked respondents to rate the attractiveness of romantic partners of different types of offenders. Sexual offenders were consistently rated as significantly less attractive romantic partners than perpetrators of (domestic) violence or theft.

There is reason to assume that the aforementioned three reasons apply more strongly for some sex offenders than for others. The prevalence of disorders has been reported to be highest among juveniles who offended against children (Hendriks, 2006; van Wijk, 1999; van Wijk et al., 2005). Laws that exclude sex offenders from certain kinds of jobs are most stringent for those who offended against a child (Boone, 2011). In addition, some of those who offended sexually against children may have sexual feelings (even if not exclusively) for children, which may make it harder to maintain a (sexually) satisfactory relationship with an adult. Also, the stigma against sex offenders appears to pertain most strongly to those who sexually offended against a child. Because of the higher prevalence of “adverse” background characteristics in child abusers, it is possible that these offenders would benefit least from the life-changing events of marriage, parenthood, and employment. In contrast, it is possible that the influence of adult roles may be most beneficial for those who committed the sexual abuse in a group. Juvenile (sexual) group offending is often contextually determined, whereby group processes, group pressure, and group normalization play a large role in the etiology of the offense (Bijleveld, Weerman, Looije, & Hendriks, 2007). In addition, previous research has demonstrated that juveniles who committed the sexual offense with a group were least burdened in terms of internal problems (Bijleveld & Hendriks, 2003). Therefore, because of their greater susceptibility to group influences, it is possible that these offenders would benefit most from life events.

Current Focus

The present study examines the role of life events in desistance from offending using a large sample of JSOs. The sample was prospectively followed up well into adulthood. Complete union and family formation, offending, and employment data were obtained from official registries.

Research by van den Berg et al. (2014) suggested that employment is associated with a reduction of offending. However, their research only examined employment. In this study, we analyze three important live events—marriage, parenthood, and employment—combined. We use these three life events together, because in recent decades, the events are more frequently occurring at later ages (approximately mid-20s to early 30s) and more close together (Shanahan, 2000). Possibly, the temporal togetherness of these life events indicates an association between the three life events. Thus, for a comprehensive view on the life course of a JSO, exploring the combination of all the three events is important.

First, the prevalence and frequency with which JSOs transition in and out of each life event will be examined. Second, the extent to which these life events are associated with general offending will be examined. This study aims to answer the following research questions:

Research Question 1: How do the marital, parental, employment, and criminal careers of JSOs develop (frequency, onset, prevalence, etc.) over time?

Research Question 2: To what extent are marital status, parenthood status, and employment status associated with offending in adulthood (age 18-28)?

Research Question 3: Do the marital, parental, employment, and criminal careers and their association with offending in adulthood differ for the three types of JSOs (child abusers, peer abusers, group offenders)?

Method

Sample

The current sample consists of 498 male JSOs. All were convicted of at least one contact sex offense (where there has been physical contact between perpetrator and victim), or had confessed to it. All perpetrators had an active role during the offense. Almost half of the sample had offended prior to the sampling offense: 15% had been convicted of a prior sex offense, and 25% had committed a property or violent offense prior to the sampling offense. The sampling offense was lechery assault in 25.5% of the cases, 24.4% of the JSOs were convicted for indecent or sexual assault and about half of the sample (50.1%) committed rape.

Age at the sampling offense, committed between 1988 and 2001, ranged from 10 to 17 years, with an average of 14.4 years (SD = 1.8). Mean follow-up time was 14 years, at which sample members were between the ages of 18 and 40 years, with an average of 28.7 years (SD = 3.9). In response to the sampling offense, 37% (n = 182) were convicted to treatment in an inpatient judicial treatment institution for juveniles and approximately 20% (n = 107) were treated in an outpatient treatment facility. However, the majority of the sample (n = 209, 42%) were only screened at an outpatient treatment facility and did not receive any treatment.

The sample will be classified into three groups (child abusers, peer abusers and, group offenders). The first group contains juveniles who have been convicted for abusing a pre-pubertal child who is at least 5 years younger (52% of the sample). Consistent with previous studies, we will refer to these offenders as “child abusers” (e.g., Gunby & Woodhams, 2010; Hendriks, 2006). The second group, referred to as “peer abusers” (Hendriks, 2006), are youths convicted for a sexual offense against a victim their own age or older (32% of the sample). The last group committed the sexual offense with at least one co-offender (16% of the sample), and will be termed “group offenders” in line with Hendriks (2006).

Before the end of the observation period, 7 individuals had died and 14 had emigrated (according to the Dutch Municipal Personal Records Database [GBA]). No significant differences were found between the 21 censored individuals and other sample members with regard to demographic and study variables.

Measures

Personal information about the offenders, obtained from screening files by the different treatment facilities, was used to retrieve data from the different data sources used in this study. After all information was retrieved, the data were de-identified. Three data sources were consulted in this study. First, judicial documentation was used to construct the dependent variable offending. Second, municipal registration data were used to construct the independent variables of marriage and parenthood. Third, centralized employment records were used to construct the independent variable employment.

Judicial documentation

The judicial documentation originates from the Judicial Documentation Centre in Almelo and contains information about all offenses registered for prosecution in the Netherlands, regardless of the verdict. This documentation holds information about the date of perpetration, offense committed, conviction date, and sentence. Conforming to research practice in the Netherlands, all offenses for which a person was acquitted or when prosecution dropped the case on so-called “technical grounds,” predominantly when the case is expected by the prosecutor to end in acquittal (such as insufficient proof, a wrong suspect identified, etc.) were excluded from the data. Offenses were classified according to the standard classification system of Statistics Netherlands (Centraal Bureau voor de Statistiek [CBS], 2010). For our dependent variable offending, a count of the number of offenses committed per age year for each individual is used. This variable contained the following offense categories (as defined by Statistics Netherlands): sexual offending (non-sexual), violent offending, property offending, and other offending (consisting mainly of drug offenses and offenses as described in the Dutch law on weapons and ammunition). The Dutch Ministry of Security and Justice granted permission for the use of these data.

Marriage and parenthood

Information on marriage and parenthood was collected from the GBA. This system contains information on each inhabitant of the Netherlands, registered with the different municipalities. Data on first marriage and on all registered children were obtained. The registration of parenthood may be incomplete because children are only registered under the name of the father if the father legally acknowledges the child as his. A dichotomous variable for marriage (1 = married, 0 = not married) per age year was constructed. For parenthood, a count or continuous variable for the number of children was used. The Dutch Ministry of the Interior and Kingdom Relations (BPR) granted permission for the use of these data.

Employment

Employment data were collected from the Ministry of Social Affairs and Employment (SUWINET) and the trade register of the Netherlands Chamber of Commerce (in Dutch: Kamer van Koophandel). The SUWINET database holds information on all employment contracts with an employer and social benefits, whereas the trade register contains information about business ownership. The number of days employed was calculated from the start and the end date of a contract as registered in the database, as well as the registration and termination date of a business in the trade register. Permission for the use of these data was granted by SUWINET.

Sample members were considered to be employed if they had regular employment with an employer, was a business owner, was employed by a sheltered workshop for the (mentally) disabled, or employed by a temporary employment agency. For each age year of measurement, an individual was considered employed if he had been employed at least 90 days in that year (van den Berg et al., 2014; Verbruggen et al., 2012).

Analytic Approach

To answer the first (descriptive) research question on the development of the criminal, marital, parenthood, and employment careers, frequency counts were used. To compare the subgroups, simple t tests and ANOVA with post hoc tests were used.

Next, a within-individual regression model was used to answer the second research question, investigating the association between the independent variables of marriage, parenthood, and employment and the number of offenses per age year as the dependent variable. A within-individual model controls for selection (Allison, 2005). This is necessary because it is possible that the same characteristics that select people into marriage, parenthood, and (un)employment also select them into crime. By examining within-individual change, the person in a sense is compared with himself over time, thereby controlling for all time-stable individual confounders. Our dependent variable, offending, is an overdispersed count variable. Therefore, the use of a regular fixed effects negative binomial model is problematic and will result in an overestimation of significant effects (see Allison & Waterman, 2002; Greene, 2005; Guimarães, 2008). For that reason, the usage of a hybrid random effects negative binomial model was necessary. This model is a random effects model with all the time-varying covariates expressed as deviations from the individual-specific means. Therefore, this is still a within-individual regression model (for more information, see Allison, 2005).

Two multivariate models were estimated. The first model examined the effect of marriage, parenthood, and employment on offending for the complete sample (Research Question 2). The second model examined the effect of marriage, parenthood, and employment on offending per sex offender type: child abuser, peer abuser, and group abuser (Research Question 3). For both models, control variables were included for age and incarceration.

Results

Descriptive Results

Offending

JSOs’ criminal activity peaked in adolescence, and overall, they committed more offenses in adolescence than in adulthood (see also van den Berg et al., 2011). After the sampling offense, the recidivism rate for sex offending was 12%. Of the new sex offenses, only 10% were committed in adulthood (i.e., after age 18). Thus, sex offending seems to decline with age for the current sample. The recidivism rate for violent offending was 32%. Over the complete follow-up period, about half of the sample has ever been incarcerated (235 persons, 47.2%).

The criminal careers in adulthood differed significantly between sex offender types (see Table 1). Child abusers committed significantly fewer offenses in adulthood than peer abusers (F = 5.89, p < .01). This group also committed fewer violent offenses (F = 7.73, p < .001) compared with peer abusers. The sex offending rate was on average 0.2 sexual offenses in adulthood, and showed no significant differences between the three offender groups. No sex offender type emerged as specialized in sex offending in the sense that they committed mainly sex offenses. All in all, child abusers seemed to differ significantly from the peer abusers in their overall offending rate (almost 50% lower) as well as in violent and other offending rates. Moreover, child abusers differed significantly from group offenders in other offending (F = 6.87, p < .001).

Table 1.

Descriptive Statistics on Adulthood Offending per Offender Type.

	Sample
		Offender type
	All	Child abuser	Peer abuser	Group offender	Group comparisons (F statistics)
n	498	259	160	79	—
Offending mean 18-28 (SD)	2.3 (4.3)	1.7 (3.5)	3.1 (5.2)	2.6 (4.6)	5.89**
Violent offending 18-28 (SD)	0.7 (1.6)	0.4 (1.4)	1.1 (2.0)	0.8 (1.5)	7.73***
Sex offending 18-28 (SD)	0.2 (0.7)	0.2 (0.6)	0.2 (0.8)	0.1 (0.4)	ns
Property offending 18-28 (SD)	1.2 (3.3)	1.0 (2.7)	1.6 (4.1)	1.2 (3.0)	ns
Other offending 18-28 (SD)	0.2 (0.7)	0.1 (0.5)	0.3 (0.8)	0.4 (0.9)	6.87**

p < .05. **p < .01. ***p < .001.

Marriage and parenthood

In the sample, 14.1% (n = 70) of the JSOs were married at least once. On average, this group married at the age of 25.1 years (SD = 3.5), with the youngest being 19 years. In the follow-up period, 14 (20%) out of the 70 marriages ended in divorce. The average age at break-up was 27.4 years (SD = 3.6). Marital and parenthood development over time are displayed as the percentage of sample members who were married or a parent at a certain age (ages 18-28) in Figure 1. By age 28, 12.1% of the sample was ever married. Compared with average Dutch males¹ (CBS, 2014a, 2014b, www.statline.nl), JSOs were married less frequently at age 28, whereas about 20% of the average Dutch males were married. Furthermore, the marital development of JSOs seemed to be comparable with a Dutch sample of high-risk juveniles. Zoutewelle-Terovan, van der Geest, Liefbroer, and Bijleveld (2012) found for these high-risk youths that about 12% was married at age 28.

Figure 1.

Development of life events over time.

About 20.5% (n = 102) of the sample fathered at least one child. On average, the sample members had their first child at age 24 (SD = 2.8), with the youngest being 16. Sample members had approximately 1.7 children (SD = 0.7) with a maximum of four. Because Statistics Netherlands does not provide information on parenthood patterns over time for males, it can only be established that JSOs were on average younger when becoming a parent than average Dutch males, who on average have their first child at the age of 27 (CBS, 2013, www.statline.nl).

Employment

For a description of the employment development over time, the percentage of sample members who were employed more than 90 days a year, are also plotted (ages 18-28) in Figure 1. Most of the JSOs were employed at some point during the follow-up period. Employment contracts spanned between 1 day and 15 years, and the average contract duration was 192 days. The sample members had approximately 11 job contracts in their employment career. Compared with average Dutch males, duration of employment contracts was very short, because the average contract for Dutch males spans 1.7 years for ages 25 to 29 (Organisation for Economic Co-Operation and Development [OECD], 2012). As can be seen in Figure 1, the percentage of employed JSOs increased gradually from ages 18 to 25. After the age of 25, the employment rate appeared stabilized at about 60%. The individuals in the 40% unemployed were not always the same individuals: as contracts were short, there is much variation. The employment rate at age 20 was similar to that of average Dutch males, with JSOs being slightly more employed than low educated Dutch males (CBS, n.d., www.statline.nl). However, with age, JSOs appeared to increasingly lag behind in their employment, as the increase in percentage employed stagnated. Compared with a Dutch sample of high-risk youths followed up until age 32, the JSOs were somewhat more employed than the high-risk juveniles who appeared to have a stabilized employment rate of 50% (see van der Geest et al., 2011; Verbruggen et al., 2012).

Descriptive statistics per offender type

Using an ANOVA, the number of people married or a parent, and the mean age of marriage or birth of first child were compared between each of the offender types (Table 2). We also included the mean age per offender type to investigate whether the findings are due to age differences between offender types. However, because no significant differences in mean age were found between the groups, none of the findings can be attributed to established age differences between the offender types. Group offenders were married more often, yet their age of marriage was slightly higher than that for the other two offender groups. However, marriage patterns did not differ significantly, χ²(2) = 1.352, p = .509. Roughly the same conclusion can be drawn for parenthood: group offenders fathered a child more often compared with the other two offender types. About 24% of the group offenders had at least one child, compared with 19% for child abusers and 17% for peer abusers, but differences were not significant. The age at which their first child was born differed significantly between the offender types. Child abusers on average had their first child at the age of 23, which was earlier than the peer abusers and group offenders who became a parent at, on average, age 25. In addition, the development over age and per offender type was plotted for marriage, parenthood, and employment.² No significant differences were found between the offender types in marital and parenthood patterns per age year. For employment, significant differences were found from ages 18 to 20, whereby child abusers appeared to be employed more often than the peer abusers and group offenders. At higher ages, the offender types did not differ.

Table 2.

Descriptive Statistics of Marriage and Parenthood per Offender Type (Age 18-28).

	Sample
		Offender type
	All	Child abuser	Peer abuser	Group offender	Group comparisons (F statistics)
n	498	259 (52%)	160 (32.1%)	79 (15.9%)	—
M age at last data collection (SD)	28.7 (3.9)	28.7 (4.0)	28.5 (3.7)	27.9 (4.4)	ns
n married (1st marriage)	70 (10.2%)	25 (9.7%)	16 (10%)	10 (12.7%)	ns
M age marriage (1st marriage) (SD)	23.5 (2.3)	23.2 (2.4)	23.5 (1.8)	24.1 (2.9)	ns
n parenthood (1st child)	95 (19.1%)	49 (18.9%)	27 (16.9%)	19 (24.1%)	ns
M age parenthood (1st child) (SD)	24.3 (2.8)	23.6 (2.9)	25.0 (2.4)	25.1 (2.8)	3.386*
n employed (ever)	448 (89.8%)	232 (89.6%)	147 (91.9%)	69 (87.3%)	ns

p < .05. **p < .01. ***p < .001.

Multivariate Analysis of the Influence of Transitions on Offending

In Figure 1, the percentage of sample members who offended and the average number of offenses per sample member are added. For offending and offending participation (percentage of the sample criminally active in a certain age year), a gradual decrease with age was found. However, this decrease was not as gradual as the increase in sample members who married. About the same conclusion can be drawn concerning the relation between parenthood and offending. Up to age 24, there was a decreasing trend for the number of offenses per person and an increasing trend for employment. Therefore, the development over time appears mirrored: when employment increases, offending decreases and vice versa. This indicates a possible relationship between offending and employment.

Next, a multivariate regression analysis was conducted consisting of two models (Table 3). The first model revealed an association between the status of marriage, parenthood, and employment on offending from age 18 to 28. Employment was associated with a significant decline of 34.2%³ in offending. Marriage and parenthood were not significantly associated with offending. An additional model estimating the influence of the “full family package” of marriage and parenthood status combined (see Zoutewelle-Terovan et al., 2014) on offending was analyzed but yielded no significant relationship.

Table 3.

Within-Individual Model for the Effect of Life Events on Offending.

	Model 1			Model 2
	β	SE	%	β	SE	%
Constant	0.975	0.821	—	0.920	0.779	—
Marriage	0.015	0.365	ns	—	—	—
Parenthood	0.277	0.137	ns	—	—	—
Employment	−0.419	0.133**	−34.2%	—	—	—
Marriage × Child abuser	—	—	—	1.126	0.666	ns
Marriage × Peer abuser	—	—	—	−0.485	0.478	ns
Marriage × Group offender	—	—	—	0.208	1.046	ns
Parenthood × Child abuser	—	—	—	0.514	0.209*	67.1%
Parenthood × Peer abuser	—	—	—	0.171	0.202	ns
Parenthood × Group offender	—	—	—	−0.062	0.381	ns
Employment × Child abuser	—	—	—	−0.250	0.198	ns
Employment × Peer abuser	—	—	—	−0.474	0.197*	−37.7%
Employment × Group offender	—	—	—	−0.943	0.381*	−61.1%
n (No. of individuals)	493			493
Observations	4,243			4,243

p < .05. **p < .01. ***p < .001.

In the second model, we analyzed group-specific effects for the different offender types. Still no association between marriage and offending was found in any of the three offender groups. Child abusers offended significantly more frequently after they had become a parent. For the remaining offender types, no significant associations between parenthood and offending were found. Also, no significant differences were found between the offender groups with regard to parenthood (not shown in Table 3). However, there were significant differences for employment and offending patterns between the sex offender groups. When employed, the chance of offending was reduced by 37.7% for peer abusers and 61.1% for group offenders. Child abusers did not significantly benefit from employment.⁴ Group comparisons demonstrate that these results differ only between child abusers and the group offenders, χ²(2) = 3.88, p < .05.

Conclusion and Discussion

Using a large sample of JSOs who had been followed up well into adulthood, this study aimed to expand the existing knowledge of the life course of JSOs by examining how and to what extent conventional adult roles (i.e., marriage, parenthood, and employment) relate to offending in JSOs. Objective registered information on offending, official employment, marriage, and parenthood was used. Our findings demonstrate that JSOs fulfill adult roles less than average emerging adults do. This is similar to prior findings for samples of high-risk emerging adults. Although their entry on the labor market appeared normative at younger ages and in fact perhaps even somewhat elevated given their generally low educational level, this quickly stagnated. From age 25 onwards, a varying proportion of about 40% was unemployed, a rate much higher than that of average Dutch males and comparable with the rate of men who spent their youth in juvenile treatment institutions (van der Geest et al., 2011). Our findings also demonstrate that JSOs often have short employment contracts, lasting on average no longer than 6 months.

JSOs’ employment careers were interspersed with periods of unemployment, and with age, their labor market participation decreased. However, the reason for this is unclear. The interspersed employment careers at the younger ages may be due to these men still being in school or acquiring diplomas. However, another possibility may be that these JSOs enter the labor market at a time when they are still too young to have educational qualifications. Moreover, with maturation, the legal minimum wages they are entitled to make them too expensive for employers, which may explain why their labor market participation is gradually reduced. Furthermore, perhaps the lack of social skills and other negative characteristics associated with this population cause frequent termination of job contracts, making workforce re-entry increasingly harder as they grow older. It may also be that the laws governing certificates of good conduct (which became stricter over the years that this sample was observed) make acquiring new jobs harder for this group. Finally, the influence of incarceration on employment may explain the decreased employment participation with maturation, as incarceration has been found to complicate re-entry in society. More research is needed to shed light on this. Even though it was expected that child abusers would fare worst on the labor market, no significant differences were found in participation rates compared with the other two groups. Therefore, it is unlikely that the adverse background characteristics, which child abusers were found to possess on average more than the other two groups, are responsible for the frequent job contract terminations and low labor market participation of this group.

For marriage, the same stagnating pattern was found, where, after a certain age, no new marriages were included. JSOs on average had children at younger ages than average Dutch males. Surprisingly, few differences were found between the sex offender types. Child abusers had children at significantly younger ages, an unexpected finding because they have adverse background characteristics that may cause difficulty in finding a partner. It is possible that child abusers marry younger and also have children earlier as a certain “overcompensating” or “overadaptation.” In other words, they possibly feel pressure to conform to conventional society, because of the stigma attached to sex offending.

Another possibility is that the offender wants to present a conventional image of being married and having children to gain the trust of others, which will make selecting victims and avoiding detection more likely as opposed to a person who is not married and has no children (see Lussier, Bouchard, & Beauregard, 2011). The multivariate analyses showed that, of the three transitions that were investigated, only employment was associated with a reduction in offending. The relationship was large, given that the reduction was more than a third overall, ranging from a null effect in child abusers to more than 60% reduction for group offenders. Although differences were significant between group abusers and child abusers, as expected, it was found that group abusers benefit most from employment. One possibility for this finding is that the same characteristics that selected them into group offending made them more receptive for the positive influences of a conventional work environment. Parenthood was significantly associated with an increase in the likelihood of general offending only in child abusers. Because our previous analyses showed that child abusers were significantly younger than the other groups when transitioning to parenthood, this may also be an effect of an “off-time” transition. In other words, these offenders may become a parent before they are ready and experience adverse effects as a result (see Elder, 1998; Thornberry, 2005).

The key finding in this study is that despite the limited transitions that JSOs made to the labor market, with fractured employment careers, employment is associated with a reduction in offending. This suggests that current criminal justice responses and treatment modalities that treat JSOs as a special group of offenders and place them “outside” of conventional society by restricting their access to a limited pool of jobs may not be evidence-based. Nevertheless, supporters of the current ever more restrictive laws and policies could argue that despite JSOs being restricted from certain jobs, the positive offending reduction effect of employment remains. Those who focus more on the empirical evidence might argue that society is doing itself a disfavor by excluding sex offenders from an increasing pool of jobs, as they would possibly offend even less if they had greater access to the labor market, making their own living and costing society less.

Another important finding is that JSOs appeared to be a heterogeneous group. Child abusers differed from the other two offender types as they do not benefit from marriage or employment but offended more when they became a parent. Further heterogeneity within this group of child abusers may be present. Although this group committed significantly fewer non-sexual offenses in adulthood than the other types of JSOs, their sex offending levels were similar to those of the other types. It might be that a particular group of child abusers exists within this group, such as pedophilia and/or paraphilia groups. Further research is needed to examine this group of child sex offenders to determine whether such heterogeneity exists and whether that would warrant specific policies or treatment modalities.

This study has important strengths as well as limitations. First, a prospective, rich data set was used consisting of long-term objective and detailed information on marriage, parenthood, and criminal and employment careers. Second, advanced methods to control for selection bias were used. Several limitations are noteworthy. First, the lack of influence of marriage on offending may be attributable to the fact that the sample is “too young,” given that the average age of marriage for Dutch males in the year of data collection was 36.5 years and the vast majority of this sample were under the age of 30 (CBS, 2014c, www.statline.nl). In addition, as recent studies have suggested, marriage may no longer be a suitable marker for involvement with a romantic partner as many young adults cohabit (in the Netherlands, cohabitating couples are entitled to the same full legal rights as married couples). Moreover, several empirical studies found marriage to influence offending if the marriage is of good quality, but in the current data, no such qualifications of the nature of the relationship are available. Second, the same goes for the quality of the employment and the relationship to a child. Third, the parenthood variable may also be considered inadequate: More men may have fathered children but not acknowledged them legally. In addition, it is unknown whether sample members actually lived with and took care of their children. Fourth, the employment variable captured only official employment registration, and no measures of irregular labor are included. Furthermore, it is unknown whether sex offenders were employed full-time or part-time, although it is likely that most contracts would have been (almost) full-time, given that the employment records showed that very few applied for supplementary benefits that citizens are entitled to if their income is below a certain minimum. Fifth, information about education is not available; therefore, it is impossible to disentangle the effect of education from employment per se.

This study was able to investigate only those life events that made it to official records, and as such were part of the official, registered side of the JSOs’ lives. More in-depth research is needed to establish the mechanisms responsible for the association between employment and offending for JSOs, and the non-effect of marriage and parenthood. For example, Ouimet and Le Blanc (1996) found that cohabitation or marriage at younger age is associated with an increase in offending, whereas after age 21, cohabitation is associated with a reduction. Therefore, future analyses should look at age-specific effects in more depth. Last, future research should incorporate the quality and nature of employment in analyses, as effects may be contingent on the quality of the social bonds they generate.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Allison

P. D.

(2005). Fixed effects regression methods for longitudinal data using SAS. Cary, NC: SAS Institute.

Allison

P. D.

Waterman

(2002). Fixed effects negative binomial regression models. In Stolzenberg

R. M.

(Ed.), Sociological methodology 2002 (pp. 247-265). Oxford, UK: Basil Blackwell.

Apel

Paternoster

Bushway

Brame

(2006). A job isn’t just a job: The differential impact of formal versus informal work on adolescent problem behavior. Crime & Delinquency, 52, 333-369.

Arnett

J. J.

(2004). Emerging adulthood: The winding road from the late teens through the twenties. Oxford, UK: Oxford University Press.

Beijers

Bijleveld

C. C. J. H.

(2012, September) Sexual delinquents and their (un)attractiveness as romantic partners. Presented at the 2012 conference of the European Society of Criminology, Bilbao.

Bersani

B. E.

Laub

J. H.

Nieuwbeerta

(2009). Marriage and desistance from crime in the Netherlands: Do gender and socio-historical context matter? Journal of Quantitative Criminology, 25, 3-24.

Bijleveld

C. C. J. H.

Hendriks

(2003). Differential personality and background characteristics of juvenile group- and solo-sex offenders. Psychology, Crime & Law, 9, 237-245.

Bijleveld

C. C. J. H.

Weerman

Looije

Hendriks

(2007). Group sex offending by juveniles: Coercive sex as a group activity. European Journal of Criminology, 4, 5-22.

Blokland

Nieuwbeerta

(2005). The effects of life circumstances on longitudinal trajectories of offending. Criminology, 43, 1203-1240.

10.

Boone

M. M.

(2011). Judicial rehabilitation in the Netherlands: Balancing between safety and privacy. European Journal of Probation, 3(1), 63-77.

11.

Caldwell

M. F.

(2002). What we do not know about juvenile sexual re-offense risk. Child Maltreatment, 7, 291-302.

12.

Centraal Bureau voor de Statistiek. (2010). Standaardclassificatie Misdrijven 2010 [Standard classification offenses 2010] (Netherlands Statistics). Retrieved from http://www.cbs.nl/nl-NL/menu/themas/veiligheid-recht/methoden/classificaties/2011-standaardclassificatie-misdrijven-politie-pub.htm

13.

Centraal Bureau voor de Statistiek. (2013). Geboorte; leeftijd vader (op 31 december), rangnummer moeder [Birth; age father (on December 31th). Rank mother.] (Netherlands Statistics). Retrieved from http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=81019NED&D1=0&D2=1-10&D3=2-12&VW=T

14.

Centraal Bureau voor de Statistiek. (2014a). Bevolking; geslacht, leeftijd en burgerlijke staat, 1 januari [Population; gender, age and marital status. January 1st] (Netherlands Statistics). Retrieved from http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=7461BEV&D1=2&D2=1&D3=19-29&D4=48-58&VW=T

15.

Centraal Bureau voor de Statistiek. (2014b). Bevolking; kerncijfers [Population; Key figures] (Netherlands Statistics). Retrieved from http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=37296NED&D1=1,10&D2=48-59&VW=T

16.

Centraal Bureau voor de Statistiek. (2014c). Gemiddelde leeftijd huwen en partnerschapsregistraties bij mannen in 2009 [Mean age of marriage and partnership registration for males in 2009] (Netherlands Statistics). Retrieved from http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=37772NED&D1=5,25&D2=59&VW=T

17.

Centraal Bureau voor de Statistiek. (n.d.). Aantal werkende mannen [The amount of employed males in the Netherlands] (Netherlands Statistics). Retrieved from www.cbs.nl/nl-NL/menu/themas/bevolking/cijfers/default.htm

18.

Cohen

Felson

(1979). Principles of criminology. Philadelphia, PA: J. B. Lippincott.

19.

Crutchfield

R. D.

Pitchford

S. R.

(1997). Work and crime: The effects of labor stratification. Social Forces, 76, 93-118.

20.

Elder

G. H.

Jr. (1998). The life course as developmental theory. Child Development, 69, 1-12.

21.

Farrington

D. P.

(1986). Age and crime. Crime and Justice, 7, 189-250.

22.

Farrington

D. P.

West

D. J.

(1995). Effects of marriage, separation, and children on offending by adult males. In Hagan

(Ed.), Current perspectives on aging and the life cycle (Vol. 4, pp. 249-281). Greenwich, CT: JAI Press.

23.

Giordano

P. C.

Cernkovich

S. A.

Rudolph

J. L.

(2002). Gender, crime and desistance: Toward a theory of cognitive transformation. American Journal of Sociology, 107, 990-1064.

24.

Greene

(2005). Functional form and heterogeneity in models for count data. Foundations and Trends in Econometrics, 1, 113-218.

25.

Guimarães

(2008). The fixed effects negative binomial model revisited. Economics Letters, 99, 63-66.

26.

Gunby

Woodhams

(2010). Juvenile sex offenders: Comparisons between the offender and offence characteristics of “child abusers” and “peer abusers.” Psychology, Crime & Law, 16, 47-64.

27.

Hanson

R. K.

Harris

(1998). Dynamic predictors of sexual recidivism. Ottawa, Ontario: Solicitor General of Canada.

28.

Hendriks

(2006). Jeugdige zedendelinquenten: een studie naar subtypen en recidive [Juvenile sex offenders: A study on subtypes and recidvism]. Utrecht, The Netherlands: Forum Educatief.

29.

Hirschi

(1969). Causes of delinquency. Berkeley: University of California Press.

30.

Horney

J. D.

Osgood

Haen Marshall

(1995). Criminal careers in the short term: Intra-individual variability in crime and its relation to local life circumstances. American Sociological Review, 60, 655-673.

31.

Knight

B. J.

Osborn

S. G.

West

D. J.

(1977). Early marriage and criminal tendency in males. British Journal of Criminology, 17, 348-360.

32.

Kruttschnitt

Uggen

Shelton

(2000). Predictors of desistance among sex offenders: The interaction of formal and informal social controls. Justice Quarterly, 17, 61-87.

33.

Lasher

M. P.

McGrath

R. J.

(2012). The impact of community notification on sex offender reintegration: A quantitative review of the research literature. International Journal of Offender Therapy and Comparative Criminology, 56, 6-28.

34.

Laub

J. H.

Sampson

R. J.

(2003). Shared beginnings, divergent lives: Delinquent boys to age 70. Cambridge, MA: Harvard University Press.

35.

Letourneau

E. J.

Miner

M. H.

(2005). Juvenile sex offenders: A case against the legal and clinical status quo. Sexual Abuse: A Journal of Research and Treatment, 17, 293-312.

36.

Lussier

Blokland

(2014). The adolescence-adulthood transition and Robin’s continuity paradox: Criminal career patterns of juvenile and adult sex offenders in a prospective longitudinal birth cohort study. Journal of Criminal Justice, 42, 153-163.

37.

Lussier

Bouchard

Beauregard

(2011). Patterns of criminal achievement in sexual offending: Unravelling the “successful” sex offender. Journal of Criminal Justice, 39, 433-444.

38.

Lussier

van den Berg

C. J. W.

Bijleveld

C. C. J. H.

Hendriks

(2012). A developmental taxonomy of juvenile sex offenders for theory, research, and prevention: The adolescent-limited and the high-rate slow desister. Criminal Justice and Behavior, 39, 1559-1581.

39.

MacKenzie

D. L.

De Li

. (2002). The impact of formal and informal social controls on the criminal activities of probationers. Journal of Research in Crime and Delinquency, 39, 243-276.

40.

Maruna

(2001). Making good: How ex-convicts reform and rebuild their lives. Washington, DC: American Psychological Association.

41.

Organisation for Economic Co-Operation and Development. (2012). Employment by job tenure intervals—Average tenure. Retrieved from http://stats.oecd.org/Index.aspx?DataSetCode=TENURE_AVE

42.

Ouimet

Le Blanc

(1996). The role of life experiences in the continuation of the adult criminal career. Criminal Behaviour and Mental Health, 6, 73-97.

43.

Paternoster

Bushway

Brame

Apel

(2003). The effect of employment on delinquency and problem behaviors. Social Forces, 82, 297-335.

44.

Piquero

A. R.

Farrington

D. P.

Blumstein

(2003). The criminal career paradigm. In Tonry

(Ed.), Crime and justice: A review of research (Vol. 30, pp. 359-506). Chicago, IL: The University of Chicago press.

45.

Piquero

A. R.

MacDonald

J. M.

Parker

K. F.

(2002). Race, local life circumstances, and criminal activity. Social Science Quarterly, 83, 654-670.

46.

Sampson

R. J.

Laub

J. H.

(1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press.

47.

Sampson

R. J.

Laub

J. H.

(2003). Life-course-desisters? Trajectories of crime among delinquent boys followed to age 70. Criminology, 41, 555-592.

48.

Sampson

R. J.

Laub

J. H.

Wimer

(2006). Does marriage reduce crime? A counterfactual approach to within-individual causal effects. Criminology, 44, 465-508.

49.

Savolainen

(2009). Work, family and criminal desistance: Adult social bonds in a Nordic welfare state. British Journal of Criminology, 49, 285-304.

50.

Seto

M. C.

Lalumière

M. L.

(2010). What is so special about male adolescent sexual offending? A review and test of explanations through meta-analysis. Psychological Bulletin, 136, 526-575.

51.

Shanahan

M. J.

(2000). Pathways to adulthood in changing societies: Variability and mechanisms in life course perspective. Annual Review of Sociology, 26, 667-692.

52.

Skardhamar

Lyngstad

T. H.

(2009). Family formation, fatherhood and crime: An invitation to a broader perspective on crime and family transitions (Discussion Paper No. 579). Oslo: Statistics Norway, Research Department.

53.

Skardhamar

Savolainen

(2012). Does employment contribute to desistance? Offending trajectories of crime-prone men around the time of job entry (Discussion Paper No. 716). Oslo: Statistics Norway.

54.

Smallbone

S. W.

(2006). Social and psychological factors in the development of delinquency and sexual deviance. In Barbaree

H. E.

Marshall

W. L.

(Eds.), The juvenile sex offender (pp. 143-163). New York, NY: The Guilford Press.

55.

Sutherland

Cressey

(1978). Principles of criminology. Philadelphia, PA: J. B. Lippincott.

56.

Tewksbury

(2005). Collateral consequences of sex offender registration. Journal of Contemporary Criminal Justice, 21, 67-81.

57.

Theobald

Farrington

D. P.

(2009). Effects of getting married on offending: Results from a prospective longitudinal survey of males. European Journal of Criminology, 6, 496-515.

58.

Thornberry

T. P.

(2005). Explaining multiple patterns of offending across the life course and across generations. The Annals of the American Academy of Political and Social Science, 602, 156-195.

59.

Thornberry

T. P.

Christenson

R. L.

(1984). Unemployment and criminal involvement: An investigation of reciprocal causal structures. American Sociological Review, 49, 398-411.

60.

Uggen

(2000). Work as a turning point in the life course of criminals: A duration model of age, employment, and recidivism. American Sociological Review, 65, 529-546.

61.

Uggen

Staff

(2001). Work as a turning point for criminal offenders. Corrections Management Quarterly, 5(4), 1-16.

62.

van den Berg

C. J. W.

Bijleveld

C. C. J. H.

Hendriks

(2011). Jeugdige zedendelinquenten. Lange termijn criminele carrières en achtergrondkenmerken [Juvenile sex offenders: Long term criminal careers and background characteristics]. Tijdschrift voor Criminologie, 53, 227-243.

63.

van den Berg

C. J. W.

Bijleveld

C. C. J. H.

Hendriks

Mooi-Reci

(2014). The juvenile sex offender: The effect of employment on offending. Journal of Criminal Justice, 42, 145-152.

64.

van der Geest

V. R.

(2011). Working their way into adulthood: Delinquency and employment in high-risk boys to age 32. Amsterdam, The Netherlands: Vrije Universiteit Amsterdam.

65.

van der Geest

V. R.

Bijleveld

C. C. J. H.

Blokland

A. A. J.

(2011). The effects of employment on longitudinal trajectories of offending: A follow up in high-risk youth from ages 18 to 32. Criminology, 49, 1195-1234.

66.

van Wijk

van Horn

Bullens

Bijleveld

C. C. J. H.

Doreleijers

(2005). Juvenile sex offenders: A group on its own? International Journal of Offender Therapy and Comparative Criminology, 49, 25-36.

67.

van Wijk

A. P.

(1999). Een verkennend onderzoek naar jeugdige zedendelinquenten? [An exploratory study into juvenile sex offenders?]. Amsterdam, The Netherlands: Vrije Universiteit Amsterdam.

68.

Verbruggen

Blokland

van der Geest

(2012). Effects of employment and unemployment on serious offending in a high-risk sample of men and women from ages 18 to 32. British Journal of Criminology, 52, 845-869.

69.

Visher

C. A.

Winterfield

Coggeshall

M. B.

(2005). Ex-offender employment programs and recidivism: A meta-analysis. Journal of Experimental Criminology, 1, 295-316.

70.

Warr

(1998). Life-course transitions and desistance from crime. Criminology, 36, 183-216.

71.

Zimring

F. E.

Jennings

W. G.

Piquero

A. R.

Hays

(2009). Investigating the continuity of sex offending: Evidence from the second Philadelphia birth cohort. Justice Quarterly, 26, 58-76.

72.

Zimring

F. E.

Piquero

A. R.

Jennings

W. G.

(2007). Sexual delinquency in Racine: Does early sex offending predict later sex offending in youth and young adulthood? Criminology & Public Policy, 6, 507-543.

73.

Zoutewelle-Terovan

van der Geest

Liefbroer

Bijleveld

(2014). Criminality and family formation: Effects of marriage and parenthood on criminal behavior for men and women. Crime & Delinquency, 60, 1209-1234.