Juveniles and Status Offenses:
The Impact of Hirschi's Bonds on Juvenile Delinquency
Hirschi’s (1969) Social Control Theory has been cited in previous literature to explain deviance, in particular juvenile delinquency, for years. Hirschi (1969) suggests there are four bonds: attachment, commitment, involvement, and belief of which a person must have at least one in order for delinquency to be prevented. In the year 2001, there were 2,273,500 juvenile arrests made (Office of Juvenile Justice and Delinquency Prevention, 2005). Even with the previous studies done which have tested Hirschi’s (1969) theory, there are still questions which should be answered in order to better understand delinquency and its causes so that perhaps the number of juvenile arrests may be reduced.
Previous studies have mainly focused on the effects of attachment, commitment, involvement, and belief of the child with his or her school, peers, and community. The bond of attachment with the family has been looked at in previous studies; however, the results are mixed as to whether it does in fact have an effect on delinquency. Some of the previous studies have confirmed Hirschi’s (1969) theory to support that these bonds do in fact, contribute to juvenile delinquency. Yet, it is still questionable as to whether Hirschi’s (1969) social control theory does play a part in determining the delinquency of juveniles.
The purpose of this research is to further test Hirschi’s (1969) social control theory. The paper will focus on the commitment a child has in school and with goals, the belief a child has in conventional ideas, the involvement the child has in school and community activities, and the attachment with both peers and family. Delinquency will be limited to status offenses because previous research indicates that regardless of age, status offenses are affected by the four bonds (Shoemaker & Gardner, 1988). The status offenses examined will include underage drinking, running away from home, and truancy. This study will compare the attachment of the child with friends compared to the attachment with parents in greater detail than previous studies to find if one form of attachment has a greater impact on status offenses than the other. It will also expand on the previous research testing Hirschi’s (1969) theory to determine whether the theory is a good measure of the causes of juvenile delinquency.
Theoretical Framework and Literature Review
Various researchers have investigated Hirschi’s Social Control Theory and how it may be used to explain various types of juvenile delinquency (Brownfield & Sorenson, 1994; Demuth & Brown, 2004; Johnstone, 1978; Kierkus & Baer, 2002; Knight & Tripodi, 1996; Poole and Regoli, 1979; Rankin & Wells, 1990; Rosenbaum, 1989; Shoemaker & Gardner, 1988; Wiatrowski, Griswold, & Roberts, 1981). There are four parts or bonds to Hirschi’s Social Control Theory: attachment, commitment, involvement, and belief (Hirschi, 1969). Attachment is the relationship which an individual internalizes; it is measured by looking at how the child internalizes the relationship with the parents (Hirschi, 1969; Demuth & Brown, 2004; Kierkus & Baer, 2002; Rankin & Wells, 1990; Knight & Tripodi, 1996). Attachment is the bond the child has with the parents, and it can come from both communication and supervision (Hirschi, 1969). Commitment is the importance an individual feels social activities hold in his or her life; it is measured by looking at the investment an individual has in social activities (Hirschi, 1969; Alston, et.al., 2001). Commitment is operationalized by looking at the importance an individual places on conventional attachments (Junger-Tas, 1992). Involvement is the actual time spent in socially accepted activities; it is measured by looking at the time spent in socially accepted activities (Hirschi, 1969; Alston, et.al., 2001). However, of the four variables this is believed to be the least predictive of delinquency because it does not actually take time to commit (Hirschi, 1969). Belief is what an individual feels are the morals that he or she should follow; it is measured by an individual’s thoughts on moral norms (Hirschi, 1969; Alston, et.al., 2001). The individual must have beliefs which forbid delinquency in order for delinquency to be prevented (Hirschi, 1969).
Data collected from many studies found that family attachment has been an important predictor of delinquency among juveniles (Demuth & Brown, 2004; Kierkus & Baer, 2002; Rankin & Wells, 1990; Knight & Tripodi, 1996). Many studies have found that when looking at familial attachment, it is not the direct relationship of the family on the child, but the internalized relationship the child feels he or she has with the parent which affects delinquency (Demuth & Brown, 2004; Kierkus & Baer, 2002; Rankin & Wells, 1990; Knight & Tripodi, 1996) Various researchers state that a child is less likely to commit a deviant act when the child feels that what the parents think about their actions is important (Demuth & Brown, 2004; Kierkus & Baer, 2002; Rankin & Wells, 1990; Knight & Tripodi, 1996).
Involvement has been used in previous literature also as a measure of delinquency (Shoemaker & Gardner, 1988; Wiatrowski, Griswold, & Roberts, 1981). Shoemaker & Gardner (1988) studied involvement and how it related to the child’s relationship with the school environment and community involvement, and the relationship between school and delinquency was not significant while community involvement was significant in lowering delinquency (1988). Wiatrowski, Griswold, & Roberts measured involvement of the individual by how much school work the child did (1981). In their study, involvement was significant in lowering the level of delinquency (Wiatrowski, Griswold, & Roberts, 1981).
The bond of commitment was also used in previous studies in order to find if there was a correlation between commitment and delinquency (Shoemaker & Gardner, 1988; Wiatrowski, Griswold, & Roberts, 1981). Shoemaker & Gardner measured commitment to school and church, and their findings show that the bond of commitment is statistically significant in lowering delinquency (1988). Wiatrowski, Griswold, & Roberts (1981) measured commitment by looking at occupational and educational aspirations, and their findings show that commitment is statistically significant in lower levels of delinquency.
The bond of belief was operationalized by Wiatrowski, Griswold, & Roberts (1981) by looking at the later stages of belief instead of the development of a belief system. Their study measured belief by using an honesty index measuring truthfulness, cheating, and lying (Wiatrowski, Griswold, & Roberts, 1981). Belief was found to be correlated to lower levels of delinquency (Wiatrowski, Griswold, & Roberts, 1981). Shoemaker & Gardner (1988) found that belief was more consistent in lowering levels of delinquency than the other elements of the bond.
Various methods have been used in order to study Hirschi’s Social Control Theory. Demuth & Brown (2004) and Rankin & Wells (1990) used the method of longitudinal surveys in order to gather their data. Other studies used a cross-sectional method of a survey to gather data (Kierkus & Baer, 2002; Johnstone, 1978; Knight & Tripodi, 1996). Brownfield and Sorenson used triangulation by using both a questionnaire and interviews (1994). Rosenbaum also used past records from the females used in the study in order to gather data (1989).
Shoemaker & Gardner (1988) have also looked at age in connection with juvenile delinquency and they found that parental attachment is more important influencing delinquent acts such as violent crimes and property offenses at a younger adolescent age (1988). As the age of the individual increases, other factors such as school and peers become more important than family attachment in affecting more serious delinquent acts (Shoemaker & Gardner, 1988). Status offenses, however, do not show the same pattern in the decrease of family affect on delinquency as the other offenses (Shoemaker & Gardner, 1988).
Different types of delinquency have also been further investigated to test Social Control theory, in particular the bond of attachment with the family and peers, in order to find if there is a difference in how familial attachment affects particular offenses (Kierkus & Baer, 2002; Johnstone, 1978). Kierkus & Baer (2002) and Johnstone (1978) both found that familial attachment affected delinquent acts such as status offenses or in-home delinquency more than major offenses such as murder (2002; 1978). Johnstone also found that peer attachment seemed to be the most strongly correlated with all types of delinquent acts (1978). Warr (1993) found that family attachment had an impact on whether the children had peer friends who were delinquent, but it did not directly have an effect on delinquency.
Though the previous two studies by Kierkus & Baer and Johnstone found that familial attachment was not as important in predicting more serious delinquent acts, a study by Poole and Regoli found contradicting results (1979). Poole and Regoli found that respondents who had lower levels of family attachment engaged in more frequent and more serious delinquent acts (1979). Their study did find that as the number of peer delinquents went up so did the number and seriousness of delinquent acts (Poole and Regoli, 1979).
A study by Knight and Tripodi (1996) found that parental attachment continued to be important in determining the levels of delinquency as a child aged. However, their findings showed that the respondents had extremely high attachment with parents, but they were still being arrested for serious delinquent acts (Knight and Tripodi, 1996). It was also found that the population in this study had parents who, similar to the respondents, also did not follow the norms of society, and 70 percent of the parents had been in jail sometime throughout their lives (Knight and Tripodi, 1996). Knight and Tripodi’s study found that the children were attached to unconventional parents which complies with Hirschi’s social control theory which states that children attached to unconventional parents will be as delinquent as children with low attachment to conventional parents (1996; Hirschi, 1969).
Some studies on family and its effects on delinquency have looked at different structures of families such as step-parent, single-parent, two biological parents, and other adults living within the family unit to see if this affects delinquency differentially (Demuth & Brown, 2004; Kierkus & Baer, 2002; Rosenbaum, 1989). Size of family has also been studied to find if it has an effect on parental attachment, and it has been found that there is no correlation between size of family and attachment (Brownfield & Sorenson, 1994). Also, gender of parent has also been studied to find if there is a correlation between gender of parental attachment and delinquency, and it has been found that families with only a single-father are often more delinquent than other family living situations (Demuth & Brown, 2004).
Social control theory has also been applied to prosocial behavior instead of being applied strictly to delinquency (Huebner & Betts, 2002). Huebner and Betts showed that attachment and involvement can increase academic achievement as well as lack of either bond can be a contributing factor in delinquency (2002). This study also examined the differences between gender in relation to the bonds of involvement and attachment and findings suggest that involvement is an important predictor in delinquency and academic achievement for both males and females; however, attachment is more prominent for girls (Huebner & Betts, 2002).
A study by Uggen and Janikula (1999) studied where volunteer work fit into Social Control Theory. Their study compared the arrest rates of respondents who were volunteers with respondents who were not volunteers. The results of their study found that volunteer work did play a part in Social Control Theory, and that the individuals who did participate in the conventional activity of volunteering were less likely than non-volunteers to be arrested after high school (Uggen & Janikula, 1999).
In this study, I will extend the previous literature on Hirschi’s Social Control Theory and juvenile delinquency. Many of the previous studies have used indices to measure both the levels of the four bonds and delinquency. In this analysis, I will investigate how well the bonds of attachment, commitment, involvement, and belief of Hirschi’s social control theory can explain the status offenses of underage drinking, running away from home, truancy, and disobeying parents. Another aspect which my research will look at is if family attachment has a greater affect on status offenses than attachment with peers.
The data for this study was secondary analysis which came from the Kentucky Youth Survey. The sample was obtained from public schools in Kentucky. Students given the survey range from grades 6 to 12 (Wilcox and Clayton, 2001). Data was collected in 1996 with students from 22 schools (Wilcox and Clayton, 2001). The sample size was reduced for the purpose of the present study to include only individuals age 17 and younger, with a total sample size of 25, 379 respondents.
Measurement of Variables
Attachment with Family
To measure the attachment an individual has with family (see Table1), a 10-item scale was used to measure the many aspects of family attachment. The questions in the scale range from sharing feelings with parents to doing things with parents. Respondents were asked to choose from a four-item Likert scale which ranged from 1 (“never”) to 4 (“most of the time”). I constructed an additive scale using the four items (α=.709) with higher scores indicating greater attachment. The mean for the scale was 30.9.
Attachment with Peers
To measure the attachment an individual has with peers (see Table1), a nine-item scale was used to measure the many aspects of peer attachment. The questions constituting the scale range from liking to be around best friends to mutual respect. Respondents were asked to choose from a four-item Likert scale which ranged from 1 (“strongly disagree”) to 4 (“strongly agree”). I constructed an additive scale using the four items (α=.786) with higher scores indicative of greater peer attachment; the mean for the scale was 27.9.
Belief in Conventional Goals
Belief was measured using one single-item indicator (see Table1): How important is religion/spirituality in life? Respondents chose from a four-item Likert scale which ranged from 1 (“not at all important”) to 4 (“very important”). The mean was 3.3 for belief in conventional goals.
Commitment to Conventional Goals
Commitment was measured using two single-item indicators (see Table1): 1) amount of school I think I will complete, 2) once I have a goal, I make a plan on how to reach it. On the first question respondents chose from a six-item Likert scale which ranged from 1 (“some high school”) to 6 (“graduate or professional school; the mean for this question was 4.17. On the second question, respondents chose from a four-item Likert scale which ranged from 1 (“never true”) to 4 (“always true”). The mean for the second question was 2.6.
Involvement in Conventional Activities
For involvement, a four-item scale was used to measure the involvement in various conventional activities (see Table1). The questions in the scale also asked about involvement in school activities, sports, and community or other physical activities. Respondents were asked to choose from a five-item Likert scale which ranged from 1 (“never”) to 5 (“everyday”). I constructed an additive scale using the four items (α=.564) with higher scores indicative of more involvement; the mean was 10.1.
Alcohol use was measured using one singe-item indicator (see Table1): best describes how I drink alcohol. Respondents chose from a six-item Likert scale which ranged from 1 (“never had drink”) to 6 (“drink almost daily”). The mean for alcohol use was 2.3.
Skipping school was measured using one single-item indicator (see Table1): during the past week, I skipped school. Respondents chose from a four-item Likert scale which ranged from 1 (“rarely or none of time”) to 4 (“most or all of the time”). The mean for skipping school was 1.2.
Running away from Home
Running away from home was measured using one single-item indicator (see Table1): during the past week, I ran away. Respondents chose from a four-item Likert scale which ranged from 1 (“rarely or none of time”) to 4 (“most or all of the time”). The mean for running away from home was 1.1.
Control variables were included in order to ensure that the results were not spurious. The control variables included sex, age, and race (see Table1).
TABLE 1 Descriptive Statistics for Study Variables
Variable Metric Mean S.D. Range
Attachment with Family (10=low attachment…, 40=high attachment) 30.91 5.22 10-40
Attachment with Peers (9=low attachment…, 36=high attachment) 27.88 4.27 9-36
Belief in Conventional Goals
Importance of religion/spirituality (1=not at all important, 2=not very important, 3.30 .893 1-4
3=somewhat important, 4= very important)
Commitment to Conventional Goals
School I think I will complete (1=some high school, 2= high school/GED, 3=vocational, trade school, 4.17 2.02 1-6
4=some college, 5=college degree, 6= graduate or professional school)
Have goal, make plans to reach (1= never true, 2=occasionally true, 3=mostly true, 4=always true) 2.62 1.00 1-4
Involvement in Conventional Activities (4=low involvement…, 20=high involvement) 10.06 3.75 4-20
Best describes how drink alcohol (1= never, 2=1-4 times in life, 3=3-4 times per year, 4=once per month, 2.31 1.40 1-6
5=1 or 2 times per week, 6=almost daily)
Past week, skipped school (1=rarely or none, 2=some or little, 3=occasionally or moderate amount, 1.23 .667 1-4
4=most or all of time)
Running Away from Home
Past week, ran away (1=rarely or none, 2=some or little, 3=occasionally or moderate amount, 1.13 .526 1-4
4=most or all of time)
Sex (0=female, 1=male) .492 .500 0-1
Race (0= white, 1=nonwhite) .158 .365 0-1
Age (number of years) 13.98 1.86 10-17
Table 2 shows the bivariate relations among the study variables. The relationships between the independent and dependent variables are all significant (p<.01). Family attachment, commitment, involvement, belief, and peer attachment are all significantly related to the status offenses of alcohol use, truancy, and running away from home at the bivariate level. Family attachment has the highest correlation with reducing the status offenses.
To test the effects of the bonds on the status offenses, while controlling for sex, age, and race, I performed ordinary least squares (OLS) regression models for each dependent variable. As seen in Table 3, family attachment was significantly related to all three of the status offenses. The results indicated that family attachment was significantly and negatively associated with alcohol use, truancy, and running away from home. As family attachment increased, the frequency of each of the status offenses decreased regardless of sex, age, and race (see Figure 1).
Table 3 also shows the effects of peer attachment on the status offenses. Peer attachment was significantly and positively related to alcohol use. As peer attachment increased, the frequency of alcohol use also increased. However, peer attachment was not significantly associated with truancy. Peer attachment was significantly and negatively associated with running away, meaning that as peer attachment increased, the frequency of running away decreased (see Figure 2).
The effects of belief on the status offenses are also shown in Table 3. Belief was significantly and negatively associated with truancy and running away. It was negatively related to alcohol use, but this relationship was not significant. As belief increased, the frequency of each of the status offenses decreased regardless of sex, age, and race.
Table 3 also shows the effects of commitment on the status offenses. The first question about commitment on amount of school the individual thinks they will complete was significantly and negatively related to all three status offenses. As commitment increased, the frequency of each of the status offenses decreased regardless of age, race, and sex. However, the question about reaching goals was positively and significantly associated with running away which indicates that as commitment increases, the frequency of running away also increases. This finding was unexpected, and future research needs to be done to find possible reasons for these results. Reaching goals was negatively associated with alcohol use and positively associated with truancy, but neither of these relationships was significant.
Involvement and its effects on the four status offenses can also be seen in Table 3. Involvement was positively and significantly associated with running away. As involvement increases, the frequency of running away also increases regardless of age, race, and sex. This finding was unexpected, and future research should be done to explain why this relationship occurred. Involvement was positively associated with alcohol use and truancy, but these relationships were not significant.
Regression analysis also showed that as age increased, the frequency of each status offense also increased. Males also scored higher on each status offense than females.
TABLE 2 Zero Order Correlations Among Study Variables
1 2 3 4 5 6 7 8 9 10 11
TABLE 3 OLS Regression Models Examining the Effects of Family Attachment, Peer Attachment, Belief, Commitment, and Involvement on Status Offenses
Alcohol Use Truancy Ran Away
Coeff. S.E. Coeff. S.E. Coeff. S.E.
Family Attachment -.061* .002 -.030* .001 -.027* .001
Peer Attachment .027* .002 -.002 .001 -.005* .001
Belief -.211 .011 -.073* .005 -.037* .004
Amount of School I think I will complete -.025* .007 -.031* .004 -.018* .003
Once I have a goal, make a plan to reach it -.013 .010 .009 .005 .028* .004
Involvement .001 .003 .000 .001 .003* .001
Age .285* .005 .043* .003 .000 .002
Race -.029 .027 .025 .013 .041* .011
Sex .155* .019 .090* .009 .062* .007
N 16,938 17,333 17,304
Consistent with Hirschi’s (1969) Social Control Theory, family attachment was significantly related to all three status offenses. In previous studies, family attachment has been found to be significantly related to status offenses and in-home delinquent acts (Kierkus & Baer, 2002; Johnstone, 1978). My replication of these results provides further support for Hirschi’s (1969) Social Control Theory.
Consistent with Hirschi’s (1969) Social Control Theory, peer attachment was significantly related to alcohol use. In previous studies, peer attachment has been found to be significantly related to all types of delinquent acts from status offenses to serious delinquent acts (Johnstone, 1978; Poole & Regoli, 1979). My results were not consistent with the results from previous studies. A possible reason for these results could be that truancy and running away are both acts which an individual does more by him/herself, and peer attachment may not have a large effect on either of these status offenses.
Belief was found to be significantly related to truancy and running away which is consistent with Hirschi’s (1969) Social Control Theory. Previous research done by Wiatrowski, Griswold, and Roberts (1981) found that belief was more consistent in lowering levels of delinquency than the other three bonds. My results found that family attachment was more consistent in lowering levels of delinquency than belief, involvement, or commitment.
Question one on commitment about the amount of school an individual thinks they will complete was found to be significantly related to all three status offenses which is consistent with Hirschi’s (1969) Social Control Theory. Question two on commitment about reaching goals was not consistent with Hirschi’s (1969) Social Control Theory. Previous research found that commitment was significant in lowering levels of delinquency (Shoemaker & Gardner, 1988; Wiatrowski, Griswold, & Roberts, 1981).
Involvement was found to be positively related to the three status offenses, which is not consistent with Hirschi’s (1969) Social Control Theory. Previous research found that involvement was the least predictive of the four bonds (Hirschi, 1969). My results were consistent with previous research on involvement.
My results found that Hirshci’s (1969) Social Control Theory is still important in predicting status offenses. However, results varied for each status offense which indicates that the bonds affect alcohol use, truancy, and running away from home differently. Also, results for commitment are somewhat confusing since each question measuring commitment found different results.
Some limitations to this research are that it was secondary data, and as a result, the measures for some of the variables were not ideal. Another limitation was that this data was not longitudinal. Results would have been more conclusive had respondents been surveyed for several years to find if the effects of the bonds change over time in predicting delinquency.
For future research, primary data should be used to ensure that the measures of the bonds are good measures. Also, it would be interesting for future researchers to look more into varying degrees of delinquency to find if the bonds do affect status offenses and violent offenders differently. Future researchers also should use a longitudinal method of data collection to find if the bonds vary in predicting delinquency over the years.
Despite the limitations of my study, I have provided evidence that attachment, involvement, commitment, and belief do affect the status offenses of drinking underage, running away from home, and truancy to some extent. This offers support for Hirschi’s (1969) Social Control Theory. However, some of my findings seem to question some aspects of this theory. Further research needs to be done to determine whether Hirschi’s (1969) theory is still a good predictor of delinquency.
Alston, Reginald J., Harley, Debra, & Lenhoff, Karen. (2001). Hirschi’s Sociological Control Theory: A Sociological Perspective on Drug Abuse Among Persons with Disablities. Journal of Rehabilitation, 31-35.
Brownfield, David, & Sorenson, Ann Marie. (1994). Sibship Size and Sibling Delinquency. Deviant Behavior. An Interdisciplinary Journal, 15, 45-61.
Demuth, Stephen, & Brown, Susan L. (2004). Family Structure, Family Processes, and Adolescent
Delinquency: The Significance of Parental Absence Versus Parental Gender. Journal of Research in
Crime and Delinquency, 41(1), 58-81.
Hirshci, Travis. (1969). Causes of Delinquency. Las Angeles, CA: University of California Press.
Huebner, Angela J., & Betts, Sherry C. (2002). Exploring the Utility of Social Control Theory for Youth Development. Youth & Society, 34(2), 123-145.
Johnstone, John W.C. (1978). Juvenile Delinquency and the Family: A Contextual Interpretation. Youth & Society, 9(3), 299-313.
Junger-Tas, Josine. (1992). An Empirical Test of Social Control Theory. Journal of Quantitative Criminology, 8(1), 9-28.
Kierkus, Christopher A, & Baer, Douglas. (2002). A Social Control Explanation of the Relationship between Family Structure and Delinquent Behaviour. Canadian Journal of Criminology, October, 425-458.
Knight, Karen Witchcoff, & Tripodi, Tony. (1996). Societal Bonding and Delinquency: An Empirical Test of Hirschi’s Theory of Control. Journal of Offender Rehabilitation, 23(1/2), 117-129.
Office of Juvenile Justice and Delinquency Prevention. 2005. “The Latest from OJJDP.” Retrieved January 30, 2005 (http://ojjdp.ncjrs.org/index.html).
Poole, Eric D., & Regoli, Robert M. (1979). Parental Support, Delinquent Friends, and Delinquency: A Test of Interaction Effects. The Journal of Criminal Law & Criminology, 70, 188-193.
Rankin, Joseph H., & Wells, L. Edward. (1990). The Effect of Parental Attachments and Direct Controls on Delinquency. Journal of Research in Crime and Delinquency, 27(2), 140-165.
Rosenbaum, Jill Leslie. (1989). Family Dysfunction and Female Delinquency. Crime and Delinquency, 35(1), 31-44.
Shoemaker, Donald J., & Gardner, III, Robert LeGrande. (1988). Social Bonding, Age, Delinquencies: Further Explanation. Journal of Sociology, 19, 195-212.
Warr, Mark. (1993). Parents, Peers, and Delinquency. Social Forces, 72(1), 247-264.
Wiatrowski, Michael D., Griswold, David, & Roberts, Mary K. (1981). Social Control Theory and Delinquency. American Sociological Review, 46, 525-541.
Wilcox, Pamela, & Clayton, Richard R. (2001). A Multilevel Analysis of School-Based Weapon Possession. Justice Quarterly, 18(3).