Adolescent Health and Harassment Based on Discriminatory Bias

Abstract

Is harassment based on personal characteristics such as race/ethnicity, sexual orientation, religion, or disability more detrimental than general harassment? In 2 large population-based studies of adolescents, more than one third of those harassed reported bias-based school harassment. Both studies show that bias-based harassment is more strongly associated with compromised health than general harassment. Research on harassment among youths rarely examines the underlying cause. Attention to bias or prejudice in harassment and bullying should be incorporated into programs and policies for young people.

In adolescence, discrimination often takes the form of school harassment or bullying, but studies rarely consider prejudice as a motive underlying bullying.¹ Racial/ethnic discrimination is associated with distress and compromised academic performance.^2–4 Similarly, anti-gay, anti-lesbian, and anti-bisexual harassment is linked to health risks,^5–7 and youths with disabilities report frequent victimization.⁸ Previous studies have highlighted the health-compromising effects of discriminatory harassment. They have not, however, compared bias-based harassment with general harassment or examined multiple forms of bias. We addressed these issues in 2 population-based studies and compared the health implications of bias-based harassment and general harassment.

METHODS

We analyzed data from the 2008–2009 Dane County (Wisconsin) Youth Assessment (DCYA) and the 2007–2008 California Healthy Kids Survey (CHKS). Participants and measures are described fully on the DCYA (http://www.danecountyhumanservices.org/Family/Youth/youth_assessment_2009.aspx) and CHKS (http://chks.wested.org) Web sites. Using logistic regression, we estimated and compared the odds of reported dichotomized health risk indicators separately for youths who reported no harassment, general harassment in the absence of bias, and bias-based harassment. Odds were adjusted for basic demographic characteristics.

Harassment

The DCYA asked youths how often they had been bullied, threatened, or harassed at school in the preceding 12 months because they were perceived as gay, lesbian, or bisexual (GLB) or because of their race/ethnicity. The CHKS asked about these issues and gathered data on harassment based on religion, gender, and physical or mental disability (past 12 months). These data were recoded into binary indicators of bias-based harassment.

With respect to general harassment, the DCYA asked how frequently youths had been verbally and physically harassed at school in the preceding 30 days and how frequently they had been harassed “about how [they] look” or harassed via the Internet or text messages in the preceding 12 months (α = 0.85). The CHKS asked young people questions regarding rumor spreading, physical and sexual harassment at school, and harassment related to their appearance, how they talked, or “any other reason” in the preceding 12 months (α = 0.77). Those who reported these types of harassment but not bias-based harassment were the non-bias-based harassment group.

Outcomes

The DCYA and CHKS assessed substance use in the preceding 12 months and the preceding 30 days, respectively. Youths were asked about smoking and alcohol, marijuana, and inhalant use; CHKS youths were also asked whether they had ever used methamphetamines (0 = never, 1 = ever). In both surveys, youths were asked whether they had driven while they were drunk or had been a passenger in a car driven by an intoxicated adult or teenager in the preceding 12 months. Youths reported whether a boyfriend or girlfriend had ever purposefully hit, slapped, or physically hurt them (in the DCYA) or had done so in the preceding 12 months (in the CHKS).

In both surveys, youths were asked whether, in the preceding 12 months, they felt “so sad or hopeless almost every day for 2 weeks or more that [they] stopped doing usual activities.” The DCYA asked youths whether they had thought seriously about killing themselves in the preceding 30 days and whether they had attempted to kill themselves in the preceding 12 months.

Youths were asked about their grades (0 = mostly Bs and Cs or above, 1 = mostly Cs or below) and how often they had been truant in the preceding year (CHKS) or 4 weeks (DCYA; 0 = never, 1 = ever). The CHKS asked how often on school property youths had been threatened or injured with a weapon and had had property stolen or deliberately damaged in the preceding 12 months (0 = never, 1 = ever).

RESULTS

Among youths reporting harassment, 35.8% (DCYA) and 40.3% (CHKS) reported bias-related harassment. In the DCYA, 15.5% of the sample reported GLB-based harassment, and 15.8% reported race-based harassment; the corresponding figures in the CHKS were 10.2% and 17.7%. Slightly more than 10% of the CHKS respondents reported harassment based on religion or gender, and 6.5% reported disability-related harassment.

Tables 1 and 2 present odds ratios comparing youths who had not suffered harassment with youths who had suffered bias-based and non-bias-based harassment. Young people who had experienced bias-based harassment were more likely than were those who had not experienced harassment to report each of the risk outcomes. Mental health status and substance use levels were worse among youths suffering bias-based harassment than among those suffering non–bias-based harassment. Bias-based harassment was associated marginally with grades but much more with truancy. Young people experiencing bias-based harassment had much higher odds than did those experiencing non–bias-based harassment of being threatened with a weapon and having property damaged at school.

TABLE 1—

Comparison of Youths Who Experienced Harassment Versus Those Who Did Not: Dane County Youth Assessment, 2008–2009

	Type of Harassment
	Non–Bias-Based (n = 5784), OR (SE)	Bias-Based (n = 3232), OR (SE)	GLB-Based (n = 2052), OR (SE)	Race-Based (n = 2152), OR (SE)
Substance use
Smoking (past y)	1.37 (0.15)	2.00 (0.25)	2.13 (0.31)	2.28 (0.34)
Alcohol use (past y)	1.48 (0.13)	1.99 (0.21)	1.90 (0.23)	2.39 (0.31)
Marijuana use (ever)	1.37 (0.14)	1.90 (0.22)	2.05 (0.28)	2.31 (0.31)
Inhalant use (ever)	1.89 (0.37)	3.54 (0.70)	4.02 (0.86)	5.23 (1.12)
Risk behaviors/areas and mental health
Drunk driving (past y)	1.98 (0.15)	2.35 (0.22)	2.33 (0.26)	2.55 (0.29)
Victim of relationship violence (ever)	2.54 (0.40)	4.19 (0.70)	4.39 (0.81)	5.23 (0.98)
Depression (past y)	2.40 (0.26)	4.06 (0.50)	4.61 (0.64)	3.96 (0.57)
Suicidal ideation (past mo)	2.86 (0.42)	5.10 (0.79)	5.82 (0.98)	4.74 (0.84)
Suicide attempt (past y)	2.46 (0.57)	5.06 (1.18)	5.80 (1.45)	5.87 (1.50)
School-related outcomes
Grades mostly Cs or below (past y)	1.37 (0.17)	1.68 (0.23)	1.74 (0.28)	1.69 (0.27)
Truancy (past y)	1.42 (0.16)	1.86 (0.23)	2.08 (0.30)	2.24 (0.32)

Note. GLB = gay, lesbian, or bisexual; OR = odds ratio. The total sample size was n = 17 366 (50.1% boys, 73.8% White, 5.4% GLB, transgender, or questioning), consisting of students in grades 7–12 in 45 public schools. Odds ratios were adjusted for age, race/ethnicity, gender, and sexual orientation. All odds ratios significantly differed from 1 at P < .001.

TABLE 2—

Comparison of Youths Who Experienced Harassment Versus Those Who Did Not: California Healthy Kids Survey, 2007–2008

	Type of Harassment
	Non–Bias-Based (n = 238 290), OR (SE)	Bias-Based (n = 160 791), OR (SE)	GLB-Based (n = 56 965), OR (SE)	Race-Based (n = 98 541), OR (SE)	Religion-Based (n = 57 605), OR (SE)	Gender-Based (n = 57 950), OR (SE)	Disability-Based (n = 36 066), OR (SE)
Substance use
Smoking (past mo)	1.49 (0.03)	2.71 (0.06)	3.73 (0.09)	2.47 (0.05)	3.08 (0.08)	3.49 (0.09)	4.97 (0.14)
Alcohol use (past mo)	1.78 (0.03)	2.37 (0.04)	2.39 (0.05)	1.96 (0.03)	1.91 (0.04)	2.35 (0.05)	2.90 (0.07)
Binge drinking (past mo)	1.67 (0.03)	2.52 (0.05)	2.93 (0.07)	2.15 (0.05)	2.47 (0.06)	2.91 (0.07)	4.10 (0.11)
Marijuana use (ever)	1.55 (0.03)	2.13 (0.04)	2.54 (0.06)	1.86 (0.03)	2.00 (0.04)	2.37 (0.05)	3.31 (0.09)
Inhalant use (ever)	1.95 (0.04)	3.63 (0.08)	4.28 (0.10)	2.93 (0.06)	3.44 (0.08)	3.84 (0.09)	5.17 (0.13)
Methamphetamine use (ever)	1.41 (0.05)	3.34 (0.11)	5.27 (0.18)	3.17 (0.10)	4.56 (0.15)	5.11 (0.17)	7.57 (0.27)
Risk behaviors/areas and mental health
Drunk driving (ever)	1.86 (0.04)	2.55 (0.05)	2.64 (0.07)	2.17 (0.05)	2.21 (0.06)	2.83 (0.07)	3.53 (0.11)
Victim of relationship violence (past y)	2.63 (0.10)	5.98 (0.22)	6.23 (0.20)	4.23 (0.12)	5.17 (0.16)	6.14 (0.22)	7.57 (0.26)
Depression (past y)	2.45 (0.04)	4.96 (0.09)	5.55 (0.12)	3.75 (0.06)	4.14 (0.09)	4.84 (0.10)	5.55 (0.16)
School-related outcomes
Grades mostly Cs or below (past y)	1.32 (0.02)	1.45 (0.03)	1.52 (0.04)	1.32 (0.03)	1.24 (0.03)	1.35 (0.03)	1.52 (0.04)
Truancy (past y)	1.79 (0.30)	2.50 (0.04)	2.52 (0.05)	2.08 (0.03)	2.25 (0.05)	2.44 (0.05)	3.06 (0.08)
Threatened with weapon (past y)	4.60 (0.19)	16.47 (0.65)	13.82 (0.39)	9.97 (0.16)	14.01 (0.40)	16.25 (0.48)	21.76 (0.68)
Property damaged (past y)	5.33 (0.11)	12.12 (0.27)	9.22 (0.19)	7.42 (0.13)	8.48 (0.18)	9.69 (0.20)	11.79 (0.30)

Note. GLB = gay, lesbian, or bisexual; OR = odds ratio. The total sample size was n = 602 612 (49.1% boys, 24.2% White), consisting of students in grades 6–12 in 16 185 public schools. Odds ratios were adjusted for age, race/ethnicity, and gender. All odds ratios significantly differed from 1 at P < .001.

DISCUSSION

Our findings from a pair of US surveys show that youths experiencing bias-based harassment at school are at greater risk for compromised health than are youths experiencing non-bias-based harassment or no harassment. Our results highlight a gap in bullying research: bias-related motives that underlie harassment are underexamined even though they appear to be a major cause of elevated risk. Furthermore, although our study was limited to harassment at school, discriminatory experiences in other contexts merit attention.

Multiple policy implications follow from this work. First, major health-monitoring youth surveys routinely assess harassment but not its motives. Consequently, vulnerability resulting from prejudice is systematically overlooked, and the ability to identify needs and resources is truncated. Given our findings, these surveys should monitor bias-based harassment and other discriminatory experiences. Second, bullying prevention programs must devote attention to prejudice and discrimination. Finally, there is a growing focus on anti-bullying laws and policies. Enumeration of vulnerable groups has been a major point of debate. Given the clear health risks associated with bias-based harassment, laws and policies should focus not only on harassing behavior but also on the bias and prejudice that motivate such behavior.