«The Social Control of Childhood Behavior via Criminalization or Medicalization: Why Race Matters DISSERTATION Presented in Partial Fulfillment of the ...»
Race and Criminalized versus Medicalized Trajectories of Social Control Evidence suggests that experiences with school punishment and the use of therapy or medication are fundamentally different for young White and African-American males (Kim, Losen, and Hewitt 2010; Losen and Martinez 2013; Morgan et al. 2013). While young African-American males are at a greater risk of suspension or expulsion, young White males are more likely to be diagnosed and treated for behavior disorders during childhood. For boys under the age of fifteen, the suspension and expulsion rate for African-Americans is three times larger than that of Whites (Losen and Martinez 2013;
Skiba et al. 2013). While rates of suspension and expulsion have increased substantially for African-American males over the past 25 years, they have remained relatively stable for young White boys (Losen and Martinez 2013). African-Americans and Whites are equally as likely to be suspended or expelled for committing serious offenses such as drug or weapon possession; however, African-American males are much more likely to be removed from school for less serious offenses, including tardiness, disrespectfulness, or classroom disruption that White boys (Hirschfield 2008b; Kupchik 2010; Skiba et al.
2013). As a result, the increasingly punitive nature of school discipline has disproportionately fallen upon young African-American males while leaving young White males relatively unscathed.
Unlike racial disparities in school punishment, young White boys are three times more likely to seek and receive therapy or medication than are African-American boys (Morgan et al. 2013). Indeed, although African-American boys report more frequent behavior problems in early childhood, they are less likely than their white peers to seek and receive therapy or medication for behavior problems (Miller, Nigg, and Miller 2009;
Morgan et al. 2013). Further, when African-American boys do seek out medical assistance for behavior problems, it is usually after symptoms have become severe or frequent enough to merit punishment as well (Miller, Nigg, and Miller 2009).
Consequently, even when their behavior problems are diagnosed and treated medically, young African-American males may be unable to avoid school punishment and the accompanying risk of criminalization (Kim, Losen, and Hewitt 2010).
Racial disparities in the social construction and control of childhood misbehavior via school punishment and/or the use of therapy and psychotropic medication have both direct and indirect implications for long-term social control outcomes over the lifecourse. Specifically, racial discrimination and the negative consequences of school punishment are cumulative risk factors which place African-American males in greater jeopardy of experiencing trajectories of criminalized social control characterized by longterm involvement with the criminal justice system than White boys who display similar behavior problems during childhood. On the other hand, by avoiding school punishment during childhood and using therapy or medication to control child misbehavior, young White males avoid criminalization and instead experience trajectories of medicalized social control characterized by routine visits to mental health professionals and long-term use of psychotropic drugs.
Summary and Hypotheses The research discussed above describes how labeling in childhood serves as a transitional event that positions a young man with respect to long-term trajectories of social control. Specifically, labeling through early school punishment and/or the use of therapy or medication for behavior problems presents a package deal when it comes to dealing with child misbehavior (Link and Phelan 2006). While both seek to ameliorate future misbehavior, school punishment relies on retribution through exclusion and isolation while diagnosis relies on restoration through dealing with underlying biological or psychological issues (Conrad 2007; Simon 2007). Consequently, school punishment disrupts a young man’s education and labels him as a rulebreaker at an early age, placing him at a greater risk of following a criminalized life-course trajectory characterized by increased risk of involvement with the legal system during adolescence and young adulthood. On the other hand, the use of therapy or medication during childhood establishes the mental health profession as the primary social control institution and identifies a young man as sick, as opposed to bad, placing him at a greater risk of following a medicalized life-course trajectory characterized by routine visits to mental health professional and psychotropic medication. Finally, evidence demonstrates that individual experiences with school punishment and the use of therapy or medication for behavior disorders are racial patterned. As a result, greater levels of school punishment among young African-American men stigmatize them early in life, increasing the risk of following criminalized life-course trajectories. On the other hand, greater rates of early diagnosis and low rates of punishment among White boys may offer them an extra “buffer” to avoid the pitfalls associated with criminalization, instead placing them on a medicalized social control trajectory.
Drawing from the research discussed above, I develop and test three research hypotheses regarding the ways in which labeling during childhood contributes to racial stratification in social control trajectories during adolescence and young adulthood. These
are delineated below:
H1a: African-American males will be significantly more likely than White males to follow criminalized life-course trajectories of social control.
H1b: White males will be significantly more likely than African-American males to follow medicalized life-course trajectories of social control.
H2a: School punishment will be significantly associated with a greater risk of following criminalized life-course trajectories of social control.
H2b: Early therapy and medication will be significantly associated with a greater risk of following medicalized life-course trajectories of social control.
H3a: Among those who experienced a labeling event during childhood, the probabilities of membership in criminalized trajectories of social control will be highest for African-Americans who experienced school punishment during childhood and lowest for Whites who experienced therapy or medication.
H3b: Among those who experienced a labeling event during childhood, the probabilities of membership in medicalized trajectories of social control will be highest for Whites who experienced therapy or medication and lowest for African-Americans who experienced school punishment.
Data and Methods To examine trajectories of social control across adolescence and young adulthood, I use data from the National Longitudinal Study of Youth, 1979 Cohort – Young Adult Sample (NLSY79-YA). The NLSY79 is a nationally representative, prospective cohort study containing information on 12,686 men and women who were between the ages of 14 and 22 in 1979. Beginning in 1986, a new survey containing information on the children of the original NLSY79 females was initiated, and, in 1994, a young adult survey was created for those children who had reached 15 years of age. For this project, I reduce my sample to African-American and White males who were younger than 15 years old in 1988 and contributed at least two years of data following their fifteen birthday 1. After removing those observations that were missing or unable to contribute data for my dependent or central independent variables, my final sample contained 3,030 respondents 2.
I focus on males for two important reasons. First, males are more likely to be suspended or expelled and diagnosed with behavior disorders than their female peers (Bertrand and Pan 2013; CDC 2012). Second, criminologists and critical race scholars have argued that school punishment in general and criminalization in particularly play important roles in shaping African-American masculinity and identity vis a vis the criminal justice system (Ferguson 2001; Rios 2009).
There were 71 observations who were missing data on childhood labeling. While the methods used are capable of handling missing data for time-varying variables, to handle issues of missing data in the timestable predictors, I employ multiple imputation techniques to generate values for all covariates using the “ICE” command in Stata (Royston 2005). ICE uses a chained equation approach in which a conditional distribution for missing data for each variable using the appropriate specification (e.g., logistic regression Analytic Strategy This paper employs group-based trajectory models using the traj command in Stata 13.0 (Jones and Nagin 2012; 2013 ). Group-based trajectory models identify clusters of individuals whose behavior follow similar pathways along a specific unit of time (e.g. age or year) (Jones and Nagin 2007; 2013). The method is a specialized application of finite mixture modeling using maximum likelihood to estimate the trajectory of each group, the proportion of the entire sample assigned to each group, and the probability of group membership for all observations (Jones and Nagin 2007; Nagin 2005). Unlike growth-curve models, which assume that the entire population shares similar risks and follows similar trajectories over time, group-based models assume that clusters of individuals follow similar and distinct trajectories (Jones and Nagin 2007;
Nagin 2005; Nagin and Odgers 2010). Although the groups are approximations and not real and actual entities, they provide a useful classification scheme for describing how events may unfold over time (Nagin 2005; Nagin and Odgers 2010; Nagin and Tremblay 2005; Petts 2009). For purposes of this project, criminalized and medicalized social control trajectories reflect extremely different experiences in the long-term contact with important institutions of social control, not necessarily labels that define individuals as criminal or mentally ill, respectively. Instead, this project uses these different trajectories for dichotomous variables) and multiple datasets are created using Gibbs sampling techniques (Royston 2005; van Buuren 2012). Following von Hippel (2007), I impute values for all variables in a given model and then delete observations with missing data on dependent and key independent variables before running our regression analyses.
as useful tools for criminologists and medical sociologists to discuss the layering of social control in the lives of individuals (Medina and McCranie 2011; Nagin and Tremblay 2005; Sampson and Laub 2005).
Trajectories of Social Control Instead of discrete events that occur in a single point in time, the dependent variables for this chapter operationalize social control experiences as trajectories that certain individuals may follow which are characterized by their involvement with or avoidance of two important institutions during young adulthood: the criminal justice system and the mental health system. Thus, I operationalize possible criminalized and medicalized life-course trajectories as separate and distinctly different age-graded trajectories. A criminalized trajectory of social control is characterized by repeated or long-term involvement with the criminal justice system during young adulthood. A medicalized life-course trajectory is characterized by the use of mental health professionals or pharmaceutical drugs to help manage behavior problems throughout young adulthood. In operationalizing each social control trajectory, I specify different groups that fundamentally differ from one another in terms of shape and composition.
For example, young men following criminalized trajectories follow different pathways with respect to criminal justice involvement than those that are not following criminalized trajectories. Similarly, those that follow medicalized trajectories follow different pathways with respect to therapy and medication than those that are not following medicalized trajectories.
The first step in identifying different groups is defining the variables that will be used to create the trajectories. To capture involvement with the criminal justice system, I use a dummy variable equal to “1” if the respondent was convicted of or sentenced for a crime, served probation, or spent time in jail or prison since the last interview 3. A similar dummy variable was created to capture involvement with the mental health system.
Specifically, I use a dummy variable equal to “1” if the respondent indicated had: (1) seen a psychiatrist or psychologist for troubles in school or for a violent temper or behavior problems at school or work and/or; (2) taken drugs to control his behavior since the last interview. The variables were measured biannually for each individual after age fifteen. Because criminalized and medicalized trajectories are created from dichotomous variables, I use the binary logistic distribution to estimate the trajectories. This model estimates each trajectory by specifying the link between contact with the criminal justice
or mental health system and age as a polynomial function:
Recently, criminologists have pointed to the ways in which involvement in the criminal justice system encompasses much more than jails and prison. For example, conviction and being on probation can also present barriers to employment. Criminal convictions carry a stigmatizing mark that follows individuals even if they were never incarcerated (Pager 2003; Western 2006). Further, probation carries the stigma of a criminal record (Pager 2003) as well as requires young men to make routine visits to probation officers, drug testing facilities, and community service obligations (Western 2006). Additionally, fines for probation can often surpass any income individuals may be able to find, placing them at risk of further punishment and preventing them from leaving probation (Harris, Evans, and Beckett 2010). Because of these factors, I consider a wide range of possible ways to come into contact with the criminal justice system.