Harmful practices prevail despite legal knowledge: a mixed-method study on the paradox of child marriage in Bangladesh

Abstract

Child marriage is a globally recognised human rights violation that disproportionately affects girls, especially in developing countries. It has serious negative consequences on girls’ physical, mental, sexual, and reproductive health and rights. Although well-pronounced laws against child marriage were enacted in Bangladesh, the practice remains a significant challenge. Lack of law enforcement and persistent social norms ultimately allow child marriage to persist around the country. Social norms have an impact on the prevalent attitudes toward child marriage. Therefore, this mixed-method study aimed to explore the legal knowledge, perception, and practice of child marriage in Bangladesh. This study was part of a broader evaluation of a UNICEF media programme. Adolescent boys and girls aged between 10 and 19 years and their parents were interviewed in three Bangladeshi districts. All the respondents were aware of the legal age of marriage and knew that child marriage is punishable by law. This study illuminated the reasons, including early marriage among boys, poverty, dowry, and sexual harassment. Communities and policymakers need to be engaged to trigger larger structural and cultural changes to remedy the harmful social norm and its practice.

Résumé

Les mariages d’enfants constituent une violation des droits de l’homme internationalement reconnue qui touche de manière disproportionnée les filles, en particulier dans les pays en développement. Cette pratique a de graves conséquences négatives sur la santé physique, mentale, sexuelle et reproductive des filles et sur leurs droits dans ce domaine. Même si des lois se prononçant clairement contre le mariage d’enfants ont été adoptées au Bangladesh, cela reste un grave problème. Le manque d’application de la législation et les normes sociales persistantes autorisent en fin de compte la poursuite des mariages d’enfants dans le pays. Les normes sociales ont des répercussions sur les attitudes prévalentes à l’égard du mariage d’enfants. Par conséquent, cette étude à méthodes mixtes visait à étudier la connaissance juridique, la conception et la pratique du mariage d’enfants au Bangladesh. Elle s’inscrivait dans l’évaluation plus large d’un programme de l’UNICEF relatif aux médias. Des adolescents, filles et garçons, âgés de 10 à 19 ans et leurs parents ont été interrogés dans trois districts du Bangladesh. Tous les répondants connaissaient l’âge légal du mariage et savaient que l’enfreindre était punissable par la loi. Cette étude a mis en lumière les raisons de cette pratique, y compris les mariages précoces chez les garçons, la pauvreté, la dot et le harcèlement sexuel. Les communautés et les décideurs doivent être recrutés pour déclencher des changements structurels et culturels plus vastes de façon à corriger cette norme sociale néfaste et sa pratique.

Resumen

El matrimonio infantil es una violación de los derechos humanos reconocida mundialmente que afecta de manera desproporcionada a las niñas, en particular en países en desarrollo. Tiene graves consecuencias negativas para la salud física, mental, sexual y reproductiva y los derechos de las niñas. Aunque en Bangladesh se promulgaron leyes bien pronunciadas contra el matrimonio infantil, la práctica continúa siendo un reto significativo. En última instancia, la falta de aplicación de la ley y persistentes normas sociales permiten que el matrimonio infantil persista a nivel nacional. Las normas sociales tienen un impacto en las actitudes prevalentes hacia el matrimonio infantil. Por lo tanto, el objetivo de este estudio de métodos combinados era explorar el conocimiento de la ley, la percepción y la práctica del matrimonio infantil en Bangladesh. Este estudio fue parte de una evaluación más amplia de un programa mediático de UNICEF. Niños y niñas adolescentes de 10 a 19 años y sus padres fueron entrevistados en tres distritos de Bangladesh. Todas las personas entrevistadas eran conscientes de la edad legal para contraer matrimonio y sabían que el matrimonio de menores de edad es sancionado por la ley. Este estudio iluminó las razones: matrimonio precoz entre niños, pobreza, dote y acoso sexual. Las comunidades y formuladores de políticas deben participar para desencadenar mayores cambios estructurales y culturales para remediar la dañina norma social y su práctica.

Keywords:

• Bangladesh
• child marriage
• law
• rights
• sexual harassment

Introduction

Child marriage, defined as marriage under the age of 18, is globally recognised as a human rights violation that disproportionately affects an estimated 700 million girls worldwide, especially in developing countries.^1–3 Although the practice of child marriage is not unique to any particular region or culture, rates are highest in South Asia and sub-Saharan Africa, usually in the most impoverished and most rural areas.^2,4,5 Despite some reduction in child marriage prevalence, it remains one of the most significant challenges in many low- and middle-income countries.^6,7 Previous research shows that almost half of all girls married before 18 years were from South Asia.⁸ Child marriage negatively impacts girls’ physical, mental, reproductive, and sexual health.^6,9–11 This practice also endangers personal development, access to education, increases susceptibility to violence and exploitation, and deprives girls of fundamental human rights.¹² Based on several studies, married young girls face problems protecting their sexual and reproductive health and rights and often they do not get appropriate information and services.¹³ The association between child marriage and childhood pregnancy is a crucial concern that results in complications related to pregnancy and childbirth and is one of the leading causes of death among girls aged 15–19 years.¹⁴ Globally, girls aged 10–14 years are five times, and girls aged 15–19 years are twice as likely to die during childbirth as women aged 20–24 years.¹⁵ The harmful effects of child marriage extend well beyond the young woman to her children, family, and society as a whole.³

Child marriage was declared to be illegal in Bangladesh in 1929.¹⁶ Since the 1980s, the minimum age for marriage in Bangladesh is 18 for women and 21 for men.¹⁴ Demographic and Health Survey data for 2014 reported that Bangladesh had the fourth-highest rate in the proportion of young women married before the age of 18, after Niger, the Central African Republic, and Chad.^14,17 Moreover, 22% of women in Bangladesh aged 20–24 years were married before reaching the age of 15.^18,19 The marriage rate among girls under 15 in Bangladesh is the highest in the world.¹⁵ Even though the median age at first marriage among women increased from 14.4 to 16.1 years between 1993 and 2014, it is still below the minimum legal age of marriage.¹⁹ In Bangladesh, the pressure for conceiving children soon after marriage is a cultural phenomenon. In 2010, a reported 27% of young mothers (15–19 years) started their childbearing years during adolescence, resulting in a high maternal death rate.^1,20,21 Various studies showed that child marriage is often the result of entrenched gender inequality. It restricts a girl’s access to education and employment opportunities, making it a clear violation of rights.^18,22 Child marriage is an index of social norms that perpetuate discrimination against girls at every stage of their lives.^8,15 The Bangladeshi government is committed to fulfilling the sustainable development goal target 5.3 of eliminating child marriage, both early and forced, by 2030.^23,24

Many studies have documented drivers of child marriage in Bangladesh, which are often correlated and cyclical.²⁴ Such factors include deeply entrenched cultural beliefs, desire to protect family reputation, poverty, gender inequality, and dowry.^25–27 Parents’ desires to secure children’s economic wellbeing, protect against the economic insecurities of climate change and sexual harassment, and prevailing social norms render communities vulnerable to child marriage.^28,29 The common practice of child marriage is ultimately a group of individual behaviours influenced by social norms that affect individual, family, and community thought-processes.³⁰ In practice, well-intended laws often go unenforced because they conflict with those prevailing social norms.^31,32

Although it is extremely complex to change widely practised social norms that significantly impact the prevalent attitudes toward child marriage, the accumulation of individual actions can influence existing societal norms.^2,22,33,34 Hence, appropriate interventions should be planned and implemented to change social norms. UNICEF Bangladesh launched a multi-tiered social and behaviour change communication programme for adolescent girls’ development in mid-2017. This programme’s cornerstone is a trans-media entertainment-education campaign to trigger widespread public resistance to child marriage. The programme also includes community engagement and social mobilisation in Bangladeshi districts with a high prevalence of child marriage. This programme works with multiple partners to develop and disseminate media content across different platforms. There is still much to learn about community perceptions on the persistence of child marriage despite laws forbidding it. This mixed-method study aims to explore the community’s legal knowledge and perspectives on child marriage practices in Bangladesh.

Methods

Study site

Using the Bangladesh Multiple Indicator Cluster Survey 2012–13 data,³⁵ six high child-marriage-prevalent districts were identified and from these three were chosen randomly for this study: Tangail, Kushtia and Nilphamari. The study site in Tangail is an urban sub-district called Sadar. From the remaining two districts, four rural sub-districts were chosen randomly. Among them, Dimla and Domer are two sub-districts of Nilphamari, and Kumarkhali and Mirpur are the two sub-districts of Kushtia.

Study design

This was a baseline study conducted as a part of a larger evaluation study (pretest/post-test design) measuring the possible change in social norms by implementing a trans-media educational campaign. The baseline study used a mixed-method triangulation design and relied on both quantitative and qualitative approaches. The triangulation design was chosen as the most appropriate mixed-method design since the research question was focused on legal knowledge and child marriage practice. This study is expected to depict all the elements of legality and practice of child marriage, which could not be addressed through one method.

Study population and inclusion criteria

The study selected households consisting of at least two adolescents (one boy and one girl) aged between 10 and 19 years who stayed with both of their parents at the time of the study. The household had to own at least one fully functioning mobile phone and a television, and the household members had to watch the state-owned television channel, namely, BTV, at least three times a week in order to be included in the study.

Sampling technique and sample size

The expected change in social norms against child marriage due to the trans-media entertainment campaign initiative was assumed to be 10%. For an estimated difference of 10% from baseline to end line at 5% statistical significance, 5% level of precision, and 80% power of the test, the sample size was estimated to be 384 households per district. With a probable non-response rate of 10%, this study’s final sample size was 1280 households. In each household, four different members were targeted to interview: (1) eldest adolescent boy, (2) eldest adolescent girl (both the boy’s and girl’s ages needed to be between 10 and 19 years at the time of baseline interview), (3) father of the adolescents, and (4) mother of the adolescents. At the last stage, 1280 sample households were randomly chosen from the list. Although the sample size was 1280 households, only 1164 of those could be surveyed, either because the rest did not meet the inclusion criteria or were difficult to reach. Of the 1164 households, not all members could be interviewed because of their unavailability or refusal to respond. The total number of respondents to the baseline survey was 4434. For the qualitative study, a cohort of households with adolescents and their parents were identified through the baseline quantitative survey across the selected three districts. From each of the five sub-districts chosen for the quantitative study, one village was randomly chosen. From each village, respondents were conveniently selected for focus group discussions (FGDs). Data were collected until the data saturation point was reached. Finally, we carried out 20 FGDs (five each with adolescent girls, adolescent boys, mothers, and fathers) in five villages in three districts.

Data collection

The quantitative and qualitative data were collected in a sequential manner using two independent tools. Quantitative data collection was carried out using electronic data collection protocols (tablets) with a standard data entry template developed in Open Data Kit (ODK) software. A dedicated data server was created for this project, and after each day of data collection, the data was sent electronically to the server for preservation. Further, all collected data was sent to a Google Drive of the assigned supervisor. A group of trained enumerators carried out the data collection process through face-to-face interviews using interviewer-administered questionnaires.

The qualitative data were collected using semi-structured guidelines. The guidelines were divided into five domains to collect adequate information on the targeted areas: child marriage and legal knowledge related to child marriage. Two experienced qualitative researchers from BRAC James P. Grant School of Public Health (JPGSPH), BRAC University, and two trained research assistants conducted the FGDs. The FGDs were held in a comfortable setting, as preferred by the study participants. The overall duration of this study was from mid-January to March 2018.

Data analysis

The quantitative data were analysed using statistical software Stata version 13. Descriptive statistics included mean, standard deviation (SD), frequency, and percentage. Bivariate analysis was done using the chi-squared test of independence to assess possible relationships between the respondents’ normative perception and characteristics. The chi-squared test of homogeneity was used to assess possible differences in perception across the respondent categories (i.e. father, mother, boy, and girl). A statistical significance level of 5% was used to signify any statistically significant finding.³⁶

The qualitative data were audio-recorded, transcribed verbatim in Bengali directly from the audio recordings. Data familiarisation was done by a literal, interpretative and reflexive reading of transcripts by the research team. A codebook was developed based on different codes and sub-codes, which helped to find the data quickly and to summarise the data in a data matrix. The data was managed, organised, and coded manually using Excel 2013. Quotations were retrieved accordingly from the appropriate text segments. Both deductive (a priori codes) and inductive (emerging codes and sub-codes) approaches were applied for coding. Each member of the research team coded the same transcript to ascertain intra-coder reliability. Then data was reduced by clustering and comparing, and, finally, thematic analysis was done.

Ethical considerations

Ethical approval from the Institutional Review Board (IRB) of BRAC JPGSPH, BRAC University, was obtained for the study. Both verbal and written informed consents were taken from each respondent after clarifying the study objective, the voluntary nature of their participation in the study, and the right of withdrawal from the study at any time during the interview. We obtained assent from the adolescent respondents and consent from their parents before data collection. Since four different groups of respondents were interviewed with the same household, an attempt was made to conduct each interview with utmost privacy to prevent possible bias in responses due to peer-influence.

Findings

Socio-demographic profile

A total of 4434 respondents were interviewed from 1164 households randomly chosen from a list of households satisfying the inclusion criteria. The sample’s description is presented in terms of the number of each type of respondent and their distribution across three study sites (Table 1). The background characteristics of the respondents are presented in Table 2. On average, the fathers were in their mid-40s, whereas the mothers were in their mid-30s. The average ages of both the girls and the boys were around 14 years. Among the interviewed adolescents, only 15 (i.e. 7 girls and 8 boys) were married. Almost all the adolescents had received some formal education, although that was not the case for many of their parents. Around one-third of the fathers and a quarter of the mothers did not have any formal education. More than a quarter of the fathers’ and more than one-third of the mothers’ educational status was at the primary/below primary level. On the other hand, many adolescents (63% girls and 57% of boys) had education up to the Secondary School Certificate (SSC). Fifty per cent of the fathers were engaged in wage work as farmers, drivers, shopkeepers, businessmen, or other occupations, and the mothers were mostly homemakers (90%). On the other hand, most adolescents (90% boys and 99% of girls) were students (Table 2). However, about 9.6% of the adolescent girls and 2% of the boys were no longer continuing their education. Among the respondents of the qualitative analysis, the age of adolescents ranged from 12 to 16 years. All of them were unmarried students, mostly between grades 7 and 10, and were not involved in any income-generating activities. The parents’ age ranged from 29 to 55 years, and each had at least two children.

Table 1. Distribution of respondents

Observation units	Number
No. of total households	1164
No. of total respondents	4434
Parents	2241
Fathers	1122
Mothers	1119
Adolescent girls	1100
Adolescent boys	1093
District-wise distribution of the respondents	Frequency (%)
Tangail (sadar)	1089 (24.6)
Nilphamari	1682 (37.9)
Dimla	833 (49.5)
Domar	849 (50.5)
Kushtia	1663 (37.5)
Kumarkhali	828 (49.8)
Mirpur	835 (50.2)

Table 2. Characteristics of the respondents

	Father (n = 1122)	Mother (n = 1119)	Adolescent girl (n = 1100)	Adolescent boy (n = 1093)
Mean (±SD) age (in years)	45.1 (±7.1)	36.6 (±5.3)	13.9 (±2.4)	14.1 (±2.8)
Marital status
Married	–		0.6	0.7
Unmarried			99.4	99.3
Education
No education	33.6	25.0	–	0.3
Primary and below	27.3	34.6	31.6	39.0
Above primary up to SSC	27.8	36.2	63.0	57.0
Higher secondary	5.2	2.1	5.2	3.0
Above higher secondary	6.1	2.1	0.2	0.7
Occupation
Wage worker	50.2	2.0	5.8	0.1
Self employed	35.3	1.9	1.1	–
Service	11.5	3.7	1.2	0.1
Homemakers	0.4	90.4	–	0.2
Student	0.3	0.6	90.4	98.7
Others	2.4	1.4	1.6	1.0
Household socioeconomic features
Monthly household income (in BDT), mean (±SD)	20,118.0 (±15,687.7)
Access to mobile phone	96.4
Access to smart phone	29.0

Note: Unless specified otherwise, % of respondents in each category is reported.

Knowledge of the law against child marriage and approval of the practice

We found that the respondents could clearly state that if a marriage took place before a girl was 18 years old, it was a child marriage. All the respondents in this study were aware of the legal age of marriage, and 98% of the participants were aware that child marriage was punishable by law. However, the respondents opined that people could also get socially sanctioned if they prevented or refused to practise child marriage. About 87% felt that practising child marriage could lead to imprisonment, whereas about 18% believed that people might get threatened if they prevented child marriage (Table 3). Imprisonment and financial penalty were the two most common forms of punishment mentioned by the respondents. Most respondents in the survey expressed disapproval of child marriage, and around 20% perceived that their community approved the practice. Approval of child marriage revealed a statistically significant difference (P < .05) stratified by sex (P = .088) (more father/boys approving than mothers/girls, in general, except when the parents were asked about their approval of child marriage for girls in particular) (Table 4).

Table 3. Consequences of practising and preventing child marriage (respondents’ perception in percentage)

	For parents who practice	For those who prevent
Legal punishments
Imprisonment	86.7	–
Monetary penalty	4.2
Shameful to the society	1.4
Social sanctions
No consequence	87.9	71.5
Social isolation	4.8	4.3
Threat of violence	2.7	17.8
Considered shameful to the society	6.9	9.5

Note: Respondents did not mention any legal punishment for those who try to prevent child marriage.

Table 4. Perception regarding approval of child marriage (respondents’ perception in percentage)

Do you approve child marriage?	Father	Mother		Father	Mother		Girl	Boy
	For girls only			For boys only			For both boys and girls
Approve/neutral	7.3	5.5	P-value^†: .088	6.4	3.8	P-value^†,*: .004	1.6	2.8	P-value^†,*: .039
Disapprove	92.7	94.5		93.6	96.3		98.5	97.2
Do your family, friends, and community approve child marriage?	Respondents’ perception about their family’s approval			Respondents’ perception about their friends’ approval			Respondents’ perception about their communities’ approval
Approve child marriage	1.8			1.5			7.5
Neutral	3.2			6.0			12.1
Disapprove child marriage	95.0			92.5			80.4

*Statistically significant (P-value < .05).

^† Chi-squared test of independence was performed to assess the possible relationship between type of respondent and perception.

The qualitative findings revealed that the continued practice of child marriage was linked to the early age at marriage of boys and the practice of dowry. Fathers stated that boys who pursue academic study usually get married in their early 20s, though some mentioned a few married at 16. Despite overwhelming sentiments expressed against child marriage in the survey, focus group discussions revealed that child marriage was not uncommon. Adolescent boys said that most of the girls’ families married their daughters before they turned 18.

“Usually, parents get their daughters married when they are fifteen to sixteen years old. They (parents) consider it as the appropriate age for marriage. Most of the girls get married at this age.” (FGD Boy, Domar, Nilphamari)

Other participants pointed out that girls get married as early as 13 years of age.

“In our area, girls get married at a very early age, and it happens before they turn 18. A few days ago, a 14-year-old girl got married.” (FGD Mother, Mirpur, Kustia)

While the mothers agreed with this statement, the fathers strongly protested, saying that this practice does not happen because it is illegal.

Perceived reasons for child marriage

A paradox arises, therefore, for the continued practice of child marriage despite overwhelming sentiments against it. The most common reason given for early marriage of girls was receiving a good proposal (from the suitable boy’s family). Reducing financial burden was also an important reason mentioned. The other reasons were protecting the girl from sexual harassment and avoiding criticism from society. There were interesting differences between fathers and mothers, and between parents and the girls themselves in the proportion reporting these reasons (Table 5). On the other hand, when we asked whether parents would marry off their sons if presented with a good proposal, only about 4% agreed or expressed a somewhat neutral view. Parents mostly thought of marrying off their sons at an early age to improve their sons’ sense of responsibility; one-third of the sons themselves agreed with this (Table 5). Close to 90% of the respondents believe that they would prevent any child marriage in their families. However, only 83.2% would intervene to prevent child marriage in their community. Around 30% of all the respondents did not believe that their community expects the marriage registrars to play an active part in preventing early marriage (Table 6).

Table 5. Perceived reasons to marry off adolescents (respondents’ perception in percentage)

	Father	Mother	Adolescents^a
Perceived reasons to marry off adolescent girls early
Avoid criticism from society	10.4	11.3	14.8
Protect from sexual harassment	7.7	11.1	14.3
Reduce financial burden	13.5	9.5	12.4
Good marriage proposals	13.4	15.1	7.4
Perceived reasons to marry off adolescent boys early
Good marriage proposals	4.0	3.7	3.8
Improve sense of responsibility	23.3	27.6	33.8

Note: The percentages within the table represent proportions that gave the mentioned responses.

^a Adolescent girls were asked about girls’ marriage and boys were asked about boys’ marriage.

Table 6. Prevention of child marriage (respondents’ perception in percentage)

Will you prevent child marriage around you?	In my family				In my community
Yes, I will stop child marriage	87.9				83.2
No/not my concern	12.4				16.8
My community expects the kazis (marriage registrars) to stop child marriage	Father	Mother	Boy	Girl	P-value^†,*
Does not expect so	30.0	26.6	34.1	30.4	.002
Expects so	70.1	73.4	65.9	69.6

*Statistically significant (P-value < .05).

^† Chi-squared test of independence was performed to assess the possible relationship between type of respondent and perception.

Very similar findings emerged from the qualitative study. An additional insight was that early marriage for boys was a driver for early marriage for girls.

Financial security

While pointing out poverty as a reason for child marriage, the respondents across the gender and geographic areas stated that marriage is a better solution for financial stability. Both parents and adolescent girls reported that financial hardship was the prominent factor for early marriage. Parents noted that when there was a prospective groom who could provide financial security, they wanted their daughter to be married right away. Another reason was that the dowry increased with the increase in the age of the prospective bride.

Early marriage for boys

Qualitative findings demonstrated that one of the reasons for child marriage was the early age of marriage for boys; according to prevalent social norms, the bride had to be even younger than the groom. Fathers stated that boys who were from well-off families and pursued their studies usually got married in their early twenties. Sometimes a few non-poor families with young sons looked for younger brides, often without any dowry, to avoid their son getting engaged in any romantic relationship.

If the groom’s family was poor, and there was a possibility of receiving a dowry, some parents married their children off at an early age.

“The sons from low-income families get married at an early age because, through marriage, the family receives dowry, often cash. They use that money for business purpose or invest in getting a job for their sons.” (FGD Father, Kumarkhali, Kustia)

The practice of dowry was an undercurrent for both the above reasons.

The practice of dowry

Around two-thirds of the respondents thought that their communities expected a dowry from the bride. More than 60% of the mothers and adolescent girls thought that the practice of dowry had gone up in the past five years, whereas a much smaller proportion (a little more than 40%) of the fathers and the boys thought the same (Table 7). The majority of the respondents were well aware that families demanding a dowry could be punished with imprisonment and a monetary penalty, but this did not make a difference to the practice of dowry.

Table 7. Perception on exchanging dowry (respondents’ perception in percentage)

	Father	Mother	Girl	Boy
Community expects dowry from bride					P-value^†
	65.8	68.0	66.6	65.3	.558
Perception of dowry practices in last 5 years in their community					P-value^†*
Increased	46.7	65.0	60.5	43.6	.000
Decreased	29.8	23.0	22.6	30.5
Remained same	23.2	11.5	13.6	21.2
No idea	0.4	0.5	3.4	4.8
Perception on punishments for exchanging dowry					P-value^‡
No punishment	4.1	7.1	3.9	4.4	.001*
Considered shameful to society	12.8	5.6	5.5	10.4	.000*
Imprisonment	83.6	81.8	83.1	79.9	.101
Monetary penalty	65.8	53.0	51.2	56.5	.000*
Community arbitration	10.7	8.0	6.3	10.0	.001*
Formal warning	15.7	7.2	5.8	11.6	.000*

*Statistically significant (P-value < .05).

^† Chi-squared test of independence was performed to assess the possible relationship between type of respondent and perception.

^‡ Chi-squared test of homogeneity was performed to assess the possible variation among different respondent categories by each response.

In addition to legal consequences, respondents were aware of the harmful health consequences of dowry practices. The bride’s family was expected to send the dowry to the groom’s family before or during the wedding ceremony. Therefore, there was pressure on the bride’s family to give a sufficient amount of money and high-quality gifts. Participants in this study explained that young girls not only suffered from psychological trauma because of inadequate gifts but often were physically assaulted by their in-laws. In an FGD of the fathers’ group, one person stated:

“One has to provide everything according to the demand of the family. If someone fails to provide half of the things, then the husband and his family torture the girl. Sometimes the bride commits suicide because of this. In many cases, the bride is sent back to her family or ends up in divorce. These monster families often torture the bride only for the dowry.” (FGD Father, Dimla, Nilphamari)

If the bride was not asked explicitly for a dowry, she often had to deal with the repercussions later. The bride and her family were reportedly expected to give presents every year after the marriage, which was often accentuated by gift-giving practices from other brides in the community. This can be stressful as well as problematic, as one mother said:

“Now, if the groom’s family does not ask for anything, we have to give them more. As they do not ask, that is why we try to give them more. If they do not ask, we will give, and if they ask, we must give. Now it has become a system in society. If you do not give your girls anything, it is treated as if you do not love your girls. It has become a mandatory social rule.” (FGD Mother, Sador, Tangail)

We found that boys were also affected by the dowry practice. According to the father and adolescent boy respondents, boys who belonged to the higher-income families were more likely to marry after finishing their studies. The respondents from the parents’ groups claimed that parents of boys want dowries because it relieves them of the financial pressures of starting their son’s business. One respondent explained:

“If our son gets married to a well-off household, then he will get a good amount of dowry and can start a business of his own.” (FGD Father, Domar, Nilphamari)

Both the adolescent groups stated that due to the family’s financial instability, their parents wanted boys to marry early, which would bind them to run a family without being prepared for married life. A respondent mentioned,

“One of my known brothers aged 15 years didn’t want to get married at that time, but his parents forced him to get married without letting him sit for J.S.C examination in class eight. They hoped that they would get a good amount of money (dowry).” (FGD Girl, Domar, Nilphamari)

Sexual harassment: a perceived reason for child marriage

In all three districts, most participants identified sexual harassment as a significant cause of child marriage. Quite often, girls dropped out of school due to sexual harassment. According to a respondent in a Mothers’ FGD,

“Most of the times the girl dropped-out, from study and the parents, are reluctant to let the girl study; they fear if they allow their daughters going to the school, what if something bad happens?” (FGD Mother, Kumarkhali, Kushtia)

Parents were worried that harassment could damage the girl’s reputation and negatively impact her prospects of a good marriage. One adolescent boy described his situation:

“My elder sister used to go to study a bit far from home. Every day while passing through the street, a boy used to disturb her and kept proposing to her continuously! When my sister disagreed, he started spreading aspersions that she had a relationship with him and committed heinous acts. My family thought that what if something more dangerous happens next time? So, they got her married. She was studying in intermediate 2nd year (12th grade) that time and was 16–17 years old.” (FGD Boy, Domar, Nilphamari).

According to adolescent girls, sexual harassment made victims feel extremely helpless. Adolescent girls reported that they did not feel comfortable sharing their sentiments about sexual harassment with their parents, as they often get blamed if they talk. One adolescent girl explained:

“Most of the time, people blame the girl. Suppose a boy is following a girl on the road, but the girl hadn’t noticed him, then everyone rebukes her saying you must have done something otherwise he wouldn’t have followed you! Mothers thought her daughter must be done some sort of activity for being harassed otherwise why other girls are not facing this.” (FGD Girl, Domar, Nilphamari)

Sometimes it was believed that the community approved the sexual harassment of adolescent girls. Respondents said that the reason sexual harassment continued to happen was that social norms made the punishment of perpetrators difficult. According to adolescent girl respondents, the community rarely supported the girls and instead accused the girls of enticing the boys. Two fathers condemned societal discriminatory gender outlook as the cause of harassment:

“Our social practice and regulations! If there were physical punishments for sexual harassment, nobody would dare to do this. There are laws in the country, and everybody knows that, but there is no implementation. If a boy sexually harassed someone, then the area’s Member would inform the local chairman; still, there would be no punishment. If there was castigation no boy would ever dare to do this kind of act. We want the punishments to be implemented and the barriers from society to disappear.” (FGD Father, Mirpur, Kushtia)

A majority of the respondents agreed when asked whether they believed that sexual harassment was a punishable offence. Only around 4% disagreed with this statement. There was consensus across the respondents’ groups on the definition of a “real man” or what a “real man” ought to be. Paradoxically again, this study found that while most respondents believed a “real man” stands up against sexual harassment, in practice societal attitudes tend to blame the girl instead. The fathers’ group also stated that economic and social hierarchy permitted injustice to the victims of sexual harassment; usually boys from wealthy or powerful families never received any punishment.

“For those who are rich and powerful in society, no one takes any legal actions against them. If we go to the Chairman, he would say these are nothing. But if a poor boy does this then they will arrest him and will make him leave his house or will make him give fifty thousand takas or they make them get married.” (FGD Father, Domar, Nilphamari)

The arrival of mobile phones and the internet has made it easier for young adolescents to start “affairs” or “love relationships”, and this has become a matter of distress among the parents. The parents stated that sometimes children are in a relationship despite their willingness to continue their education, which results in early marriage.

Loopholes in the implementation of the law against child marriage

Most respondents identified several loopholes in the implementation of the law against child marriage. A majority (around 70%) of the parents (both fathers and mothers) believed that their community expected the Kazi, or marriage registrar, to take an active part against the practice of marriage at an early age (Table 6). Although the government is currently trying to stop child marriage by promoting and issuing birth certificates through the local government, ambiguities exist in the system. Several respondents reported that the parents could manipulate their children’s date of birth by bribing the officials. Participants described events in which parents escaped with the bride and groom and hosted the marriage elsewhere. The local police did not trace them until the couple had already married. The community expressed concern that they were unsure whether the local authority could not identify unlawful marriages or avoided prosecuting due to societal pressure. Even if the law enforcement officials tried to intervene, they would have been unable to act when the wedding took place in another area. In these ways, loopholes in the system make it difficult to prevent child marriage within the community.

Discussion

Despite knowledge of the child marriage law among community members, the practice still exists in Bangladesh. Study data showed that the community practises child marriage mostly because of the dowry, sexual harassment, and poverty or financial hardship. Some other reasons for child marriage have also come out from the qualitative findings, such as getting involved in a romantic relationship, good proposals, and school dropout. Not only did most of the respondents across all categories know the child marriage law and its implications, but they also disapproved of child marriage. Previous research has suggested that families suffering from economic adversity were more prone to marrying their daughters off at a young age.^37,38 The existing perception was that marriage could be a better solution for the families’ inability to provide economic support or educate their girls.

Dowry remains one of the most common practices linked to child marriage, yet most respondents wished to eradicate this practice because of its perpetuating force on child marriage. The widely practised tradition is that goods and money are transferred from the bride’s to the groom’s family, although it is illegal under the Bangladesh Dowry Prohibition Act of 1980.²⁸ According to the FGD respondents, dowry requirements increase with the girl’s age and education. Hence, parents were more willing to marry their daughters early; this was also revealed by previous research, which found similar dowry practices.³⁹ Another study conducted in 2010 showed that the demand for dowry has escalated in past years, which was in line with the present study findings that dowry exchange has increased extensively in the past five years despite people being well aware of the punishments for it.⁴⁰ Previous research has found that in Bangladeshi rural culture, cheaper dowry expectations encouraged parents to marry daughters early, which is a direct cause of early school dropout for girls.^31,41

Child marriage places family responsibilities on boys and girls before they become mature enough, which has significant consequences for their wellbeing and violates many human rights.²⁴ As they leave school early, they often lack the skills and education to prepare for meaningful employment, which hinders them taking family responsibility. This study documented that boys were also the victims of child marriage, which is indirectly linked with financial constraints. In some cases, poor parents opted to get their son married early to get a dowry, which would help their son to get established financially. Findings from other research have shown that it is evident that girls and boys mostly do not have much choice as to whom or when they marry.²⁴ Another study revealed that adolescent boys and young men were consistently referred to as irresponsible, good-for-nothing, and only wanting to have fun; marriage was considered as a way to make them “responsible men” who can take responsibility for a family.⁴² As dowry has been a social practice and a norm for decades, it will take time to be eliminated.⁴³ New social practices can be encouraged, such as investing the dowry fund in initiating entrepreneurship for the daughters. Social awareness programmes need to align to change community perceptions against early marriage and dowry practices to protect children’s rights and wellbeing.

This study found that perceptions related to sexual harassment led to child marriage. Parents are interested in marrying off their daughters early to avoid social criticism and protect them from being sexually harassed. Findings from other studies showed that many parents believed that their daughters would be better off under the legal control of a socially acceptable male guardian.^44,45 The present study findings shed light on the fact that it was not uncommon for a girl to be identified as a victim of sexual harassment. Still, the community and family members shamed her for the incident, similar to another study conducted in Bangladesh in 2013.⁴⁶ Sexual harassment is a severe human rights violation and is a primary reason that girls drop out of school in Bangladesh.^47,48 As the frequency of sexual harassment in Bangladesh increases, the parents’ response is to marry their daughters off as soon as possible.⁴⁹ Many perceived early marriage as ensuring girls’ safety, especially when they are at high risk of physical or sexual assault.^1,50

Findings from this study also illuminated that punishment for sexual harassment in the community was often not executed. Laws against sexual harassment should be strengthened and enforced effectively to protect girls’ sexual and reproductive health and rights. Other studies highlighted that patriarchy, gender norms and the notions of honour are responsible for discrimination against girls, and as underlying causes of sexual harassment.^46,51 Unfortunately, there is no explicitly defining law against sexual harassment in Bangladesh.^9,13 The gaps and opportunities for interpretation and the laws’ applications often develop loopholes in the system to escape appropriate punishment from the law.⁵² New laws defining sexual harassment forms and prescribing harsher penalties need to be enacted to ensure girls’ reproductive health and rights. Legal knowledge and the supporting law and policy are being overshadowed by harmful social norms that lead to child marriage. Hence, the local influential persons, community leaders, NGO workers, and school committees can play a stewardship role. Both adolescent boys and girls can be included in programmes to stop sexual harassment. Plans should be introduced to allow for open discussions with communities to change their misconceptions that encourage sexual harassment. This study exposes the ambivalence in individuals about the practice and attitudes of sexual harassment that push girls toward marriage at an early age. Implementing and enforcing strict laws against sexual harassment for girls can help eliminate child marriages.

The strength of this study was the rigorous methodology that the study followed. Study participants were selected in a way that represents and covers an equal number from all the four study groups among three districts. Homogeneity was highly maintained while selecting the respondents for the FGDs. The limitation of this study was relying on a face-to-face interview to explore perception and attitude. In this type of interview, there is a potential bias towards receiving socially desired answers, particularly on socially sanctioned topics like child marriage. This study did not find new drivers (i.e. climate change, precarious livelihood) as its focus was to understand the perception of child marriage practices in Bangladesh relating to social norms rather than exploring the key drivers of child marriage.

Conclusion

Our research shows that the community, though aware of the child marriage law, is yet to enforce it. Therefore, pursuing community knowledge of the laws and the detrimental effects of these harmful practices may not be as effective as one might assume. The respondents have identified significant loopholes in the legal system, noting that reform of the system is necessary. Monitoring the marriage registrar, distribution of birth certificates and implementing checks and balances for law enforcement are important steps in mending the system. Existing and entrenched social norms generate a robust process for the inter-generational perpetuation of child marriage. However, depending only on changing social norms and enforcing the law to end child marriage practices will not be adequate; expanding efforts to other structural factors (i.e. educating girls, economic development, expanding employment opportunities, and policy work) can complement the initiatives to end child marriage. Effective context-specific interventions to prevent child marriage are desperately needed. Civil society and policy-makers need to be engaged to trigger structural and cultural changes to compensate for the harmful social norms and practices; such a holistic and comprehensive approach could result in ending child marriages in the long run.

Acknowledgements

This study was made possible through the help and support of numerous individuals at different stages of the research cycle. The study was conducted by BRAC James P Grant School of Public Health, BRAC University. We would like to acknowledge UNICEF Bangladesh research, programme and operational teams in supporting the study. The research team is incredibly thankful for the valuable time and experiences shared by all participants in the study. We are grateful to Gordon Shelby for her critical comments and Ayesha Ahmed for editing the manuscript. NK and TS planned the study. MS, SA, TS, NK, JE contributed to the conceptualisation and design. SA, AT, NI contributed to the development of the tool. SA, AT, NI involved with content development, as well as interpreting results. SA, CW drafted the initial version of the manuscript. MS, CW, SA reviewed parts of the analysis and reviewed draft articles. All authors edited the manuscript for important content, read and approved the final manuscript. The views expressed in this manuscript are those of the authors and do not necessarily represent the official position of Government of Bangladesh, UNICEF, BRAC JPGSPH and other partners and donors.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The anonymised dataset used during the study is available from the corresponding author on a reasonable request.

Additional information

Funding

This study was commissioned by UNICEF through funding received from the Governments of Canada and Netherlands, and the Swedish Committee for UNICEF.