Coming of Age in the Shadow of the Taliban: Adolescents’ and Parents’ Views Toward Interpersonal Violence and Harmful Traditional Practices in Afghanistan

Abstract

Objectives. To understand Afghan adolescents’ and parents’ attitudes toward interpersonal violence.

Methods. We used a 2-stage sampling method in 6 provinces during 2016; we included 916 adolescents aged 12 to 15 years and 454 parents.

Results. In the abstract, a minority of adolescents or parents endorsed violence; however, specific situations justified most violence (e.g., husbands beating wives, parents hitting children). Both adolescents (48.0%) and parents (39.0%) thought a wife hitting her husband was least justifiable. Endorsement of violence justification did not appear to vary significantly on the basis of parental education or wealth. More fathers rejected all violence rationale than did mothers, and women were more likely to justify wife beating (75.0% vs 58.6%; P < .01), beating of daughters (78.5% vs 60.6%; P < .01), and teachers hitting students (62.9% vs 51.5%; P < .01). Of all respondents, 25% approved of threatening a child if he or she speaks out against harmful traditional practices.

Conclusions. Although it may be socially unacceptable to advocate physical aggression, most Afghans still find numerous conditions that justify it. Without deliberate violence reduction strategies, education alone is unlikely to reduce the high levels of interpersonal violence in Afghanistan.

Since 1996, violence has been recognized as a major public health issue by the World Health Organization.¹ For Afghanistan, the legacy of interpersonal violence is intertwined with decades of armed conflict compounding what is already an undeniable horror. Together with a history of a number of harmful traditional practices, such as marrying off daughters to other families to pay for family debt or the use of boys as sex slaves for older men, the severe health and emotional consequences for women and children is unquestionable.^2–4

One study in Afghanistan estimated that 2300 women and girls attempt suicide annually because of domestic violence and other hardships.⁵ This estimate does not reflect the much larger number of those mentally or physically injured. There appears to be a normative acceptance of interpersonal violence as a way of life.⁶ For example, in 2006, a national report on domestic abuse concluded that among 4700 households in 16 provinces across Afghanistan, almost 90% of women experienced at least 1 form of interpersonal violence and nearly two thirds (62%) experienced more than 1.⁶ In a 2016 Multiple Indicator Cluster Survey (MICS), the United Nations Children’s Fund (UNICEF) reported that 90.2% of Afghan adults believed that wife beating was justified under at least 1 of 5 conditions.⁷ The Afghanistan Independent Human Rights Commission reported more than 4500 self-reported incidences of violence against women in 2014, 98% of which happened within households.⁸ Among the reported cases, physical violence was the most prevalent, accounting for 30.1% of all cases.⁸ In consideration of the socially normalized attitudes of violence against women, it is reasonable to assume that there is a significant amount of underreported violence cases. Although a Bill on Elimination of Violence Against Women was passed in 2009, cases involving violence against women have been constantly reported and are rarely prosecuted.^9–11

In addition to the documented violence against women, violence toward children has been a concern in the society. Multiple studies have shown that children in war-torn societies are at higher risk than elsewhere for exposure to violence inside the family as well as out,^12,13 which reflects the situation for Afghan children. A recent study with children from 3 geographic regions in Afghanistan (Kabul, Jalalabad, and Torkham) revealed that 71.1% of children were subjected to domestic violence (e.g., being pushed, grabbed, or kicked).^14,15 In the same study, 92.3% of parents reported ever using corporal punishment as a method of child discipline.^14,15 Besides physical aggression, harmful traditional practices against children have also been prevalent concerns. For example, child marriage and exchanging daughters to pay family debt are reported to be widespread in Afghanistan.¹⁶ These harmful practices disproportionally affect girls and represent the most severe forms of gender discrimination. Evidence suggests higher prevalence in rural communities.¹⁶ So, too, another study reported that beyond the mental health traumas of war and posttraumatic stress, family violence was strongly associated with the mental health consequences for children in Afghanistan.¹⁷

In addition to violated human rights, a large body of research has revealed the health burden imposed by domestic violence both mentally and physically, including depression, PTSD, substance abuse, and sexually transmitted diseases.^18,19 Early exposure is especially detrimental; growing up in an environment of domestic violence puts children at high risk of experiencing more abuse and adversity later in life.²⁰ Children in Afghanistan have a greater risk of experiencing these health consequences than older populations because of their higher exposure to violence and their developmental stage.

Understanding how adolescents perceive violence and the conditions that they believe justify it provides data needed for developing effective public health interventions. To our knowledge, the present study is the first to explore attitudes toward domestic and interpersonal violence in Afghanistan among adolescents aged 12 to 15 years and their parents. Likewise, to our knowledge, it is the first such analysis since the fall of the Taliban in 2002.

METHODS

The study included adolescents and parents from 6 provinces: Badghis, Bamyan, Daikundi, Kandahar, Logar, and Paktya. Together, these provinces include diverse tribal, geographic, and political sectors. We selected these sites on the basis of the United Nations Child Deprivation Index and the availability of support from organizations for potential implementation of future interventions. The United Nations Child Deprivation Index is an index score calculated on the basis of multiple indicators representing 4 domains—environmental or material conditions, health, nutrition, and education—from which an index score is generated and national as well as provincial ranks established. For the purpose of the present study, we selected the highest ranked provinces indicating the greatest deprivation.

We used a 2-stage sampling design, initially selecting 40 areas proportional to population size of target districts followed by a random selection of 17 households within each area. We defined a household as people eating from the same kitchen or at the same table or tablecloth in each area. Altogether, 922 adolescents were eligible for interview on the basis of age criteria (12 to 15 years) and from whom parental consent and adolescent assent were obtained. We subsequently excluded 6 adolescents who assented to participate because of nonresponse to questions regarding interpersonal violence. We had a final usable sample of 916 adolescents in the analyses with 431 (47.05%) boys and 485 (52.95%) girls. Of this sample, 23.5% of adolescents were from Badghis, 29.5% from Bamyan, 27.1% from Daikundi, 11.2% from Kandahar, 3.5% from Logar, and 5.2% from Paktya. A majority of the interviewed was enrolled in school (73.7%) and unmarried (96.9%; Table 1).

TABLE 1—

Demographic Characteristics of the Study Population (Adolescents and Parents): Afghanistan, 2016

	Males, No. (%), Mean ±SD, or Median (IQR)	Females, No. (%), Mean ±SD, or Median (IQR)	P
Adolescents
Sample size	431	485
Age, y			.46
12	130 (30.2)	146 (30.1)
13	114 (26.5)	109 (22.5)
14	95 (22.0)	123 (25.4)
15	92 (21.3)	107 (22.1)
Province			.76
Badghis	107 (24.8)	108 (22.3)
Bamyan	126 (29.2)	144 (29.7)
Daikundi	109 (25.3)	139 (28.7)
Kandahar	53 (12.3)	50 (10.3)
Logar	14 (3.2)	18 (3.7)
Paktya	22 (5.1)	26 (5.4)
Current school enrollment			< .01
No	78 (18.1)	163 (33.6)
Yes	353 (81.9)	322 (66.4)
Marriage status			< .01a
Never married or engaged	427 (99.1)	461 (95.1)
Engaged to be married	4 (0.9)	18 (3.7)
Married	0 (0)	6 (1.2)
Household educational level			.52
None	242 (56.1)	279 (57.5)
Primary education	128 (29.7)	124 (25.6)
Secondary education	23 (5.3)	39 (8.0)
High school	25 (5.8)	28 (5.8)
Teacher-training college	9 (2.1)	9 (1.9)
University	4 (0.9)	6 (1.2)
Household wealth index			< .01
Index	0.29 ±1.77	–0.26 ±1.63	< .01c
Low (–1.59 ±0.40)	108 (25.1)	198 (40.8)
Medium (–0.38 ±0.39)	149 (34.6)	156 (32.2)
High (1.97 ±1.41)	174 (40.4)	131 (27.0)
Parents
Sample size	198	256
Age, y			< .01
< 30	19 (9.6)	57 (22.3)
30–50	114 (57.6)	169 (66.0)
> 50	65 (32.8)	30 (11.7)
Province			.66
Badghis	52 (26.3)	67 (26.2)
Bamyan	61 (30.8)	77 (30.1)
Daikundi	53 (26.8)	79 (30.9)
Kandahar	17 (8.6)	13 (5.1)
Logar	7 (3.5)	7 (2.7)
Paktya	8 (4.0)	13 (5.1)
Marriage status			.01a
Single	2 (1.00)	3 (1.2)
Married	192 (97.0)	235 (91.8)
Separated	1 (0.5)	0 (0)
Widowed	3 (1.5)	18 (7.0)
Age at first marriage	20 (18–25)	16 (15–19)	< .01b
Education
Never	130 (65.7)	238 (93.0)	< .01a
Elementary	31 (15.7)	11 (4.3)
Secondary	31 (15.7)	6 (2.3)
Beyond secondary	6 (3.0)	1 (0.4)

Note. IQR = interquartile range.

^aFisher exact test.

^bWilcoxon rank-sum test.

^cStudent t test.

For the parental survey, consenting parents were interviewed voluntarily. For example, if 2 parents were both available at enrollment, they decided who would be the one to complete the survey. Only 1 parent per household then consented to participate in the study. A total of 82 parents declined participation, and an additional 17 did not complete the interview. There was a final usable sample of 454 parents in the analyses. Among 198 (43.6%) male and 256 (56.4%) female parents, 76 (16.7%) were aged younger than 30 years, 283 (62.3%) were aged 30 to 50 years, and 95 (20.9%) were older than 50 years. Of the parents, 30.4% were from Bamyan, 29.1% from Daikundi, and 26.2% from Badghis. Less than 10% of parents were from Kandahar (6.6%), Logar (3.1%), or Paktya (4.6%). The majority of parents (81.1%) reported never having had formal education (Table 1). Although many of the adolescents and parents were from the same households, they were not matched for 2 main reasons: (1) adolescents agreed to participate but their parents did not consent for themselves or (2) some adolescents chose not to participate whereas their parents did. Fewer parents than adolescents participated in the study.

Data Collection

A field team of Afghans, all of whom had at least a high-school education (the majority held bachelor’s degrees) collected the data. Most had previous experience with survey data collection conducted by national and international organizations. All data collectors received a 2-week training before the start of data collection. To ensure data quality, double supervision was applied from the team supervisor and one of the field managers. Data collection was done in teams of 2 with a male and female data collector and was completed via individual interview; responses were recorded on tablets. Adolescent interviews lasted on average slightly less than 1 hour and parental interviews approximately 45 minutes.

Questionnaire

Both adolescent and parent questionnaires were developed for the UNICEF–IKEA South Asia Program for Adolescents (Afghanistan Impact Evaluation) with some of the questions generated from previous regional and global surveys.^21–23 The adolescent survey was made up of sections assessing household demographics (e.g., age, gender, marital status, educational level of each family member; a total of 20 questions per family member for up to 30 members); household wealth (10 questions); the largest section assessing adolescent demographics (e.g., education, employment, marital status; 50 questions); beliefs about child marriage (17 questions), interpersonal violence (35 questions) and harmful traditional practices (7 questions); empowerment (45 questions); media access and use (12 questions); behavioral control and decision-making (7 questions); and substance use (8 questions). The parent survey collected similar information except for household wealth and detailed household demographics (survey instruments available upon request).

Key Analytic Variables

Adolescent age was measured as a discrete variable with eligible values from 12 to 15 years. Parent age was categorical (< 30, 30–50, > 50 years). Gender was binary for male and female participants. Household education level was reported as highest level of school completion by either parent: no education, primary education, secondary education, high school, teacher training college, and university. Subsequently, we collapsed all secondary education and higher into 1 category to facilitate analysis. We converted household wealth index to a categorical variable with 3 levels (low, medium, high) from a score calculated by principal component analysis using information from family income resources, access and convenience to clean water and toilets, household lighting, materials used for cooking, and house roofs and floors. We measured attitudes toward interpersonal violence by using 5-point Likert scale response options ranging from strongly disagree and disagree to agree and strongly agree. Responses of strongly disagree and disagree were grouped together to calculate the disapproval rate per type of interpersonal violence, and agree and strongly agree were grouped for agreement. We recorded responses for violence justification categorically as “yes,” “no,” “don’t know,” and “refuse to answer,” and we used responses of “yes” to calculate the proportions of violence justification for each scenario.

Data Analysis

We analyzed baseline data. We first checked the percentage of missing data for both adolescent and parent data, and with the exception of maternal and paternal education in adolescent data, 5% or less of respondent data were missing, allowing us to use k-nearest-neighbor imputation with k = 5 to impute missing values.^24,25 With the imputed data, we tabulated the frequency of adolescents’ as well as parents’ attitudes for interpersonal violence, violence justification, and harmful traditional practices. For both sets of data (adolescents and parents), we compared categorical responses for attitudes by gender with the χ² test. We conducted categorical variable comparisons by gender for demographic characteristics with the χ² test or Fisher exact test as appropriate; we compared continuous variables with the student t test if assumptions of normality and homoscedasticity were met. If not, we conducted a Wilcoxon rank-sum test. We considered P values less than .05 and .025 statistically significant for single comparison and multiple comparison, respectively. We further used logistic regression to study gender differences in violence attitudes among adolescents, with control for family wealth. We used Stata version 14.2 (StataCorp LP, College Station, TX) and R version 3.3.3 (R Project) for statistical analysis.

RESULTS

We first studied adolescents’ and parents’ perceptions toward interpersonal violence, focusing on physical aggression between couples, parents and their children, teachers and students, and among peers. Second, we investigated the circumstances that justified such violence. We also evaluated attitudes toward harmful traditional practices.

Interpersonal Aggression

Perceptions of adolescents.

When asked about interpersonal aggression in general, most adolescents in the study disapproved of interpersonal violence involving husbands hitting wives, parents beating children, teachers hitting students, and physical aggression among their friends. Specifically, more than half of adolescents rejected as “normal” any of the following violent behaviors: a husband beating his wife (75.1%), parents hitting their children (71.8%), teachers hitting students (60.8%), and boys or girls hitting friends (77.1% and 80.8%, respectively). Disapproval rates were comparable between boys and girls for aggression from a husband to his wife (76.8% vs 73.6%; P = .27), parents hitting children (72.6% vs 71.1%; P = .62), and teachers beating students (59.6% vs 61.9%; P = .49). Boys expressed greater disapproval than girls for violence against their friends. For example, 83.3% of boys versus 71.6% of girls disapproved of boys hitting their friends (P < .01) and 84.9% versus 77.1% of girls rejected violence among girls (P < .01; Table 2). To account for confounding, we adjusted for family wealth to study gender differences, and the results confirmed these findings (odds ratios not shown). Disapproval rates for each type of violence did not vary by household education level or maternal or paternal education, respectively (P ≥ .05).

TABLE 2—

Percentage of Disapprovals for Interpersonal Violence and Traditional Harmful Practices by Study Adolescents: Afghanistan, 2016

	Boys (n = 431), %	Girls (n = 485), %	P
Interpersonal violence
Most husbands hit their wives	76.8	73.6	.27
Most parents hit their children	72.6	71.1	.62
Most teachers hit their students	59.6	61.9	.49
Most boys hit their friends	83.3	71.6	< .01
Most girls hit their friends	84.9	77.1	< .01
Traditional harmful practice
Exchanging a daughter for family debt	99.1	98.4	.34
Older men forcing a boy to have sex	99.8	100	.47a
Marrying a child off before he or she is born	98.8	97.5	.14

^aFisher exact test.

Perceptions of parents.

In general, approximately 70% of parents said they disagreed with any type of violence (slightly lower at 61.2% for teachers physically punishing students). More fathers than mothers said they disapproved of teachers hitting students (67.7% vs 56.3%; P = .01) and also boys (81.8% vs 64.5%; P < .01) or girls (85.4% vs 69.9%; P < .01) hitting their friends. Comparable percentages of mothers and fathers disagreed with husbands hitting wives (78.3% vs 73.8%; P = .27) or parents hitting children (66.7% vs 69.9%; P = .46).

Justifications for Interpersonal Violence

Perceptions of adolescents.

When we looked more specifically as to the conditions under which adolescents endorsed interpersonal violence, we saw variable responses. Specifically, for each type of interpersonal aggression (e.g., spouses hitting each other and parents hitting children), we explored 5 conditions under which such physical aggression might be perceived to be justified and 3 conditions for justifying teachers hitting students (Table 3). Overall, adolescents justified interpersonal violence under at least 1 scenario with 71.2% endorsing a husband hitting his wife, 79.7% parents hitting daughters, and 83.6% parents hitting sons, and 70.3% endorsed corporal punishment in schools. A wife hitting her husband was the least justifiable form of physical aggression endorsed by 48.0% of adolescents. Violence endorsement did not appear to vary by household education, maternal or paternal education, or family wealth index (P ≥ .05). However, adolescents with mothers with no formal education were more likely to endorse a wife beating her husband than those whose mothers completed secondary education or higher (49.8% vs 31.8%; P = .02).

TABLE 3—

Percentage of Approval Conditions Justifying Interpersonal Aggression Among Adolescents Interviewed: Afghanistan, 2016

	Boys, %	Girls, %	P
Conditions justifying wife beating
If she goes out without telling him	65.6	49.9	< .01
If she neglects the children	51.2	44.4	.05
If she argues with him	53.6	54.5	.80
If she refuses to have sex with him	45.3	27.1	< .01
If she burns the food	21.2	24.7	.23
Conditions justifying husband beating
If he goes out without telling her	8.1	10.7	.19
If he drinks too much	54.5	38.9	< .01
If he is mean to her	21.6	19.6	.46
If he forces her to have sex	21.5	10.7	< .01
If he complains about the food	5.8	6.0	.90
Conditions justifying beating of daughters
If she goes out without taking the consent of parent(s)	63.8	62.8	.76
If she neglects the younger siblings	52.6	52.9	.94
If she argues with parent(s)	62.7	70.1	.02
If she neglects the household chores	53.9	50.6	.33
If she spends a lot of time chatting or playing with other girls	46.1	41.6	.19
Conditions justifying beating of sons
If he neglects his household or farm chores	56.4	53.4	.37
If he neglects his studies	72.4	59.0	< .01
If he is mean to his brothers or sisters	59.7	46.9	< .01
If he argues with his parents	72.0	68.3	.23
If he goes out without telling his parents	63.8	57.8	.07
Conditions justifying beating of students
If he or she has not done his or her homework	73.2	60.6	< .01
If he or she comes late to school	64.2	58.1	.07
If he or she refuses to stay after school to meet with the teacher alone	33.9	41.2	.04

Significantly more boys than girls endorsed at least 1 scenario to justify a husband beating his wife (76.8% vs 66.2%; P < .01), a wife beating her husband (58.0% vs 39.2%; P < .01), parents physically punishing sons (88.6% vs 79.2%; P < .01), and teachers hitting students (79.1% vs 64.6%; P < .01). Justification for parents physically punishing daughters did not differ between boys and girls (81.2% vs 78.4%; P = .28).

When aggression from husband was justified, the primary reason given was a wife going out without telling her husband, and boys were more likely to see this as a justification than were girls (65.6% vs 49.9%; P < .01). In addition, a greater proportion of boys considered wife beating justifiable if she refused to have sex with her husband (45.3% vs 27.1%; P < .01). Conversely, when the adolescents considered it to be justified for a wife to beat her husband, the primary reason given was his excessive drinking, and boys were more likely to see this as a justification than were girls (54.5% vs 38.9%; P < .01). Likewise, boys were more likely to endorse a wife beating her husband if the husband forced her to have sex (21.5% vs 10.7%; P < .01). Arguing with parents was the primary justification adolescents gave for parental beatings with more girls than boys seeing it as justification when it was a daughter arguing (70.1% vs 62.7%; P = .02). On the other hand, when it was a boy arguing with a parent, both boys and girls saw it as a significant justification for beating and there was no significant gender difference (72.0% vs 68.3%; P = .23). Likewise, boys were more likely than girls to justify parental aggression if a son neglected his studies (72.4% vs 59.0%; P < .01) or was mean to siblings (59.7% vs 46.9%; P < .01). Failure to finish homework was the most common justification for a teacher beating a student, with 73.2% of boys and 60.6% of girls endorsing that scenario (P < .01).

Again, we adjusted for family wealth to explore gender differences, and again results aligned with data presented previously except that, compared with boys, girls were less likely to justify a husband hitting his wife if she neglected the children (adjusted odds ratio [AOR] = 0.74; 95% confidence interval [CI] = 0.56, 0.97) and to endorse a teacher beating a student for being late for school (AOR = 0.73; 95% CI = 0.55, 0.97). We observed no gender differences for justifying a teacher beating a student for his or her refusal of meeting after school (AOR = 1.24; 95% CI = 0.92, 1.67).

Perceptions of parents.

Similarly, the majority of parents endorsed a husband beating his wife (67.8%), parents hitting daughters (70.7%) and sons (71.2%), and teachers beating students (57.9%) under specific situations. As was true for adolescents, parents felt that a wife hitting her husband was the least justifiable form of interpersonal aggression (39.0%). Husbands were more likely than wives to justify a wife hitting her husband (43.9% vs 35.1%; P = .04) whereas wives were more likely to justify wife beating (75.0% vs 58.6%; P < .01). When asked about the conditions that justify a wife hitting her husband, the number-1 reason given was excessive drinking; however, argumentative behavior was the most common justification for a husband to hit his wife. Wives were more likely than husbands to endorse parents beating their daughters (78.5% vs 60.6%; P < .01) and teachers hitting students (62.9% vs 51.5%; P < .01). Regardless of gender, arguing with parents was the primary justification for parents beating children. “Going out without permission” was the top reason for physically punishing daughters. Endorsement for parental aggression toward children differed by parents’ education level. Compared with those who completed at least secondary education, less-educated parents were more likely to justify physical aggression toward both daughters (75.3% vs 52.3%; P < .01) and sons (73.9% vs 54.6%; P = .02). Nearly half of parents considered it reasonable for teachers to physically punish students if they were late for school.

Harmful Traditional Practices

There is a set of behaviors that has been clustered as harmful traditional practices—most specifically, exchanging a daughter to pay family debt, older men forcing a boy to have sex (bacha bazi), and marrying a child before he or she is born. These were rejected by nearly all adolescents (≥ 98.1%) and parents (≥ 96.3%) surveyed. Disapprovals by adolescents for all practices did not vary by maternal or paternal education (P > .05). However, despite opposition to these practices, about 20% of adolescents and parents both said that it would be appropriate for an adult to threaten a young person if he or she reports experiencing or witnessing any of these practices.

DISCUSSION

Our study is the first, to our knowledge, to explore attitudes toward multiple types of interpersonal violence from a large multiprovincial sample of adolescents and parents in Afghanistan. Two other smaller-scale studies have recently investigated violence against children in Afghanistan: one study evaluated the prevalence of peer violence and its associated factors among schoolchildren, the other assessed violence against children by using a community sample of 149 adolescents aged 12 to 18 years and 104 parents.^14,26 Our study across 6 provinces and 13 districts aligns with the findings in these smaller samples. It has been noted that corporal punishment in schools was associated with peer violence perpetration.²⁶ Conversely, we explored the attitudes of 2 generations of Afghans, not just among in-school adolescents. Although not national, the geographic scope (6 provinces) and size of this study provides insights heretofore unavailable.

Our results suggested that although adolescents and parents generally rejected violence, they endorsed it under numerous specific conditions, whether with spouses, parents, peers, or schoolmates. Consistent with what O’Leary et al. found, home-based corporal punishment was strongly endorsed in the present study. Specifically, O’Leary et al. found that 60% of the parents (n = 104) endorsed physical forms of discipline, while 8% only used positive disciplinary approaches.¹⁴ We did not explore positive parenting but found a wide variety of conditions in which parents beating adolescent children was believed to be justified. Another study (n = 182) based on 3 Afghan communities found a relatively low reported level of awareness of violence against children among religious and community leaders.²⁷ Both our study and other research should raise concerns given that witnessing and experiencing violence perpetuates violence itself; this has been reported in Afghanistan.²⁶

Our data are also consistent with the findings in existing literature reporting that 90.2% of Afghanistan women considered wife-beating justified under certain circumstances.⁷ Our study indicated that 66.2% of girls and 75.0% of female parents thought wife abuse to be acceptable under at least 1 condition. The same study also found that Afghan women, as well as women in several regional countries (e.g., Nepal, Pakistan, Bhutan), were more likely to endorse interpersonal violence against women, consistent with parents’ attitudes in our study and reflected the results suggested by the MICS data of Afghanistan (2015).²⁸ Where the inconsistency appears is the age gradient that endorses this violence. Though not observed in the MICS data, ours suggest that fewer adolescent girls endorsed the violence (66.2%) relative to adult women (75.0%). The discrepancy between MICS and the present study may be partially explained by sample selection—MICS included women from all provinces and thus is more nationally representative. It is worthwhile for future research to explore generational difference in attitudes toward violence with child–parent dyads.

Since the end of Taliban rule, UNICEF and the United Nations have been actively engaged in Afghanistan in reducing violence against children and promoting equality through programming and legislation.^29–34 A multipronged approach has been adopted by government to increase gender equality in health, reported by Samar et al.³⁵ However, cases and small-scale study reports highlight the continued exposure to violence experienced by women and children despite more than 15 years of programs and services.^10,11,36 In addition, although nearly all the participants in our study rejected harmful traditional practices, those practices appear still prevalent in certain regions of the country.¹⁶

Limitations

Although the present study was multiprovincial, it may not be nationally representative as it sampled select provinces, as many within the country are challenged by active conflict. As the sampled group included some of the most disadvantaged provinces in Afghanistan, it is possible that the participants were more economically depressed and educationally constrained. However, the United Nations Child Deprivation Index score reflects deprivation at a provincial level and not at the household, where greater variability was seen. We further used the 2-tiered sampling strategy to minimize bias by randomly sampling households within each district of the target provinces. The inability to match parent with adolescent responses precludes exploring concordance of responses.

Public Health Implications

Since the fall of the Taliban in 2002, there has been a concerted effort by the government and nongovernmental organizations in Afghanistan to reduce domestic violence and improve gender equality.^29–34 The present study suggests that although there is general repudiation of violence, both young adolescents and their parents endorse violence as a solution to interpersonal disagreement under numerous specific conditions. Findings from our study also suggest that adolescent girls and women perhaps justify violence more as a means of dealing with interpersonal issues. With these data there is a need to develop programs and strategies starting in childhood so that subsequent generations of Afghanis are less tolerant of interpersonal violence and less likely to resort to it than the present generations. The data also serve as strong empirical evidence for such programs aiming at changing cultural and social norms that contribute to child maltreatment as recommended by the World Health Organization.³⁷ Moreover, to inform effective programming design and practice, a reliable evidence base is needed.³⁸ Our study serves as an addition to this limited evidence base.