Child Marriage Acceptability Index (CMAI): South and Central Sulawesi, Indonesia

This study aims to assess child marriage acceptability in the two locations in Indonesia by gender inequality, nancial security, education rates, legal frameworks, dowry, sexual and gender-based violence (SGBV). Methods This study used a quantitative approach with a cross sectional study design. A total of 1,000 respondents consisting of 500 households in Bone District, South Sulawesi and 500 households in Palu, Sigi, Donggala District in Central Sulawesi participated in the study. Data analyses were conducted based on the Acceptability Child Marriage Index (ACMI) by using the bivariate correlation, ANOVA (analysis of variance) and logistic regression. Results This study found several signicant factors contributed to child marriage acceptance in Central and South Sulawesi: household nancial security (p=0.016), dowry (p=0.000) and legal frameworks (p=0.017) based on ANOVA analysis. After conducting bivariate correlation, dowry (p=0.000) and sexual and gender-based violence (p=0.000) remain signicant factors. Dowry (p=0.000, with expected B=0.122), and sexual and gender-based violence (p=0.001, with expected B=0.064) remains signicant after the linear regression analysis. collected and organized data. Ratnaningsih analyzed and interpreted the results. analyzed validity and reliability test.


Introduction
Child marriage remains an important social issue worldwide and it is a violation of human rights. Girls are more vulnerable for child marriage compared to boys. UNICEF reported that there has been a decline in the global prevalence of child marriage in the last decade. However, there are still around 12 million girls who experience child marriage per year. The highest prevalence of child marriage is in sub-Saharan Africa and South Asia, where an estimated 38% and 30% of girls marry before 18 years, respectively [1]. Countries and societies with high gender inequality (e.g. laws and customs that exclude girls from decision-making or economic and political rights) are more likely to feature a high prevalence of child marriage.
Child marriage remains an important social issue in Indonesia. UNICEF data ranks Indonesia eight in the world and second highest in Southeast Asia in terms of the percentage of marriages in which at least one of the spouses is under 18 years [2]. Child marriage in Indonesia has decreased slowly in Indonesia, from 14.67% in 2008 to 10.82% in 2019. However, 22 out 34 provinces in Indonesia still had high child marriage rates above the national average in 2019. South Kalimantan had the highest rate (21.2%) followed by Central Kalimantan (20.2%) and West Sulawesi (19.2%). On the other hand, the Indonesia government has a target to decrease the child marriage prevalence to 8.74% in 2024 as part of its national development strategy [3].
The minimum number of participants was calculated based on Hulley et al (2013) (p.57) method, where N (sample) = 16: (standardized size effect) 2 . The standardized size effect to be used is 0.3, with a con dence interval of 95%, and power (β) of 80%. From the number of samples that have been calculated, researchers add an additional sample up to 10% of the total including adding samples in the control area. So that the number of samples becomes 500 for each location. The total sample that participated in this study was 1000 respondents (500 respondents from Bone, South Sulawesi and 500 respondents from Palu, Sigi, Donggala, Central Sulawesi).

Data collection
Respondents aged 30 years and over from 8 villages in Bone, South Sulawesi and 19 villages in Palu, Sigi and Donggala in Central Sulawesi who have children aged 13-15 years were recruited. With the help of the village leaders, we mapped the houses and determined which respondents were visited by using random techniques, for example with the distance of every 10 houses, and if there were no families with children aged 13-15 years, we visited the house next door. The data collection used the mWater Surveyor App on tablets and smart handphones. The survey took approximately 50-60 minutes to complete. The data that has been collected is stored by Tulodo Indonesia. Only UNICEF Indonesia and Tulodo Indonesia have full access to the set data. If the set data want accessed by an outside party, they must ask UNICEF Indonesia for approval to use the data. This research has received ethical approval from the Ethics Commission for Research and Community Health Service, Faculty of Public Health, University of Indonesia (No.256 / UN2.F10 / PPM.00.02 / 2019) from the Provincial Government of South and Central Sulawesi. Respondents in this study signed an informed consent which contained information about the length of the interview, the con dentiality of the data provided, voluntary participation of the respondents, no potential hazards generated after the research, and the research results would not be used by anyone other than researchers and interested parties.

Data analysis
To analyse the quantitative data, the SPSS 25.0 for PC was used. Descriptive statistics were used to analyse demographic data, such as gender, age, income, and education status. To answer the research question, inferential statistics were used. The questionnaires were developed based on the Child Marriage Acceptability Index (CMAI) developed by Asia Child Marriage Initiative (ACMI) from Plan International.
The CMAI is comprised of a set of indicators (and associated variables) that explain the presence of environmental factors associated with the acceptability of child marriage. The main indicators for the CMAI included household nancial security, education level, legal framework, dowry and sexual and gender-based violence. This study used 21 questions on attitudes and perceptions that relate to various elements of acceptability of child marriage based on the CMAI questionnaire developed by Plan International and ACMI. Each question has its own score, for example the ideal age for marriage, disparity in ideal age for marriage for girls compared to boys, and so on. For the perception of child marriage, the responses were arranged along a 7-level Likert scale consisting of: strongly disagree, disagree, slightly disagree, neither disagree nor agree, slightly agree, agree, and strongly agree. For testing of each question from CMAI, there are two types of statistical tests used: the validity test uses the Corrected Item-Total Correlation, while the reliability test uses Cronbach's Alpha (0.693). For the household nancial security indicator, the household income was categorized into six groups: those whose income is in top 10% of national average; 4th quintile (top 60-80%), 3rd quintile (40-60%), 2nd quintile (bottom 20-40%) and 1st quintile (bottom 20%) and those living below poverty line). It also measured whether households had su cient food, clothes, medicines and school items; then the responses were classi ed into always, most of the time, sometimes and never. The education indicator was measured by classifying households' education into having formal/mainstream education, basic education/madrassa, and no education). The legal framework was measured by asking whether households have correct knowledge of the marriage law including whether they register their marriage and own identity documents. The dowry variable was measured by exploring whether marriage involves dowry practices and whether younger brides or grooms require lower dowry price. The last indicator, the sexual and gender-based violence (SGBV) measures the acceptance of sexual violence against women/girls and male control over marriage, and the belief that child marriage prevents sexual harassment.
After calculating the acceptability scores of each variable, a series of statistical tests, including bivariate correlation, ANOVA and regression analysis were conducted to test the associations between the independent variable and the child marriage acceptability scores. (202 people) had income less than IDR 500,000; income between IDR 2,000,000-3,000,000 (13.0%, 130 people), income between 1,000,001-2,000,000 (6.0%, 60 people).   In total, there are twenty questions comprising the Child Marriage Acceptability Index (CMAI). and above 40 years (11.2%, 112 people).    Central and South Sulawesi. A number of studies showed that nancial transactions around marriage may also contribute to the practice of child marriage, especially in contexts of poverty and vulnerability. In communities where the groom or his family pays a bride price at the time of marriage, which is often the case in parts of Africa, parents may bene t from marrying their daughters early if waiting increases bride prices. By contrast, in communities where the bride brings resources at the time of marriage (dowry, which is more prevalent in South Asia), the required dowry to be paid by parents may be lower if the bride is younger. Marrying a daughter at a younger age also reduces the investments that a family has to make in her education, without necessarily curtailing future returns to those investments if those returns bene t mostly the groom's family. This may lead parents to reap immediate bene ts from an early marriage even if this is not in the long term interest of the girl marrying early [17].

Results
Nonetheless as a result of the mounting pressure to control girls' behavior and marry them early, the practice of dowry is widespread and demands are increasing, a view supported by participants in this research. As one older participant in research notes: "When I got married, there was no thought of the groom getting a single dime. Rather the groom had to pay 100 Bangladesh taka (USD1.18) to bring the wife along with him. And that money had to be paid to the father of the bride. The groom had to pay the cost of the ceremony. Now, it's the other way around. If you go to any girl or the girl's father, they ask for money or things. (Older man, Gaibandha)" and "Now it is quite impossible to marry without dowry. The amount is 10,000 taka (USD 118). 20,000 taka (USD 236) and 30.000 taka (USD 354) and more… now it is quite impossible. Most are not married without dowry. There is a great demand [on the family for] money, though they forget to know the name of the girl but not forget about dowry. It is very funny, the cow that is given as dowry, rstly they wanted to see it. Their curiosity is about how nice the cow is. It is like a market of dowry, they take [marry] a little girl but do not take her without dowry. (Older man Gaibandha)" [18].
In India where dowry payments are common, shocks may reduce the probability of child marriage, possibly, because a girl's or a boy's family is unable to meet the dowry requirements. Cultural norms also heavily infuence child marriage. In societies where bride price payment is practiced (i.e. the groom's family provides assets to the bride's family in exchange for marriage), the bride's family may reap immediate nancial bene ts from marrying their daughter. In other context a younger bride is more desirable as she has more time to commit to her new family and bear more children. Thus, where a bride price is paid, the bride's family may gain greater bene t the younger their daughter is, which may motivate parents to marry their daughter early. Similarly, in circumstances where dowry payment is practiced, a smaller dowry may be required for a younger bride, so parents may be incentivised to marry their daughter at a young age to avoid the increasing cost [19].
In Bugis culture, the most dominant ethnic in South Sulawesi and some in Central Sulawesi, there is a form of a bridewealth practice named uang panai. A study from South Sulawesi reported that bridewealth practice (uang panai) was determined by the women's social status, age, education level, and pride of the bride and her family. Uang panai is paid by the groom to the bride and her family to conduct the wedding ceremony [20]. As uang panai has social value in Bugis culture, it may explain our ndings that bridewealth/dowry practices were widely accepted by the community.
A study in Donggala in Central Sulawesi province reported that virginity was also considered a main factor in determining the amount of dowry besides level of education, employment status, religion, and behavior and beauty. It can be measured whether the girls are already pregnant (kawin kecelakaan) or still virgin (kawin adat). Those that are kawin adat will have a higher dowry amount compared to those who experienced kawin kecelakaan (kawin lari, getting pregnant, and marriage because of being caught red-handed by a traditional leader). The amount of money and dowry brought by the prospective groom is determined by the results of the family deliberations of the prospective husband and wife. Study shows the level of socio-economic status of the woman's family whom she intends to marry also greatly determines the size of the marriage fee. This status is usually measured by virginity, level of education, employment status, religion, and behavior and beauty. A community leader in Tanjung Batu stated the following: "If a woman who wants to get married has a high social status (the woman is still a virgin, works as a civil servant, bachelor, religion, physical appearance), then the dowry for this is at least 30 million rupiahs (USD 2,100). But if they just graduated from high school…., only 10-20 million (USD700-1,400)." [21] 4

.2. Household nancial security
This study also found that nancial security including income also contributed to the acceptability of child marriage in Central and South Sulawesi. High prevalence of child marriage in Indonesia is related to poverty. A study in Indonesia reported that child marriage especially for girls has a tendency limit their income generation capacity so they can have a higher standard of living. In this study, women's economic conditions by age group were classi ed into ve categories: lowest 20%, lowest 20%-40%, lowest 40%-60%, lowest 60%-80%, and highest 20%. The lowest 20% and lowest 20-40% welfare status re ects poor economic conditions. Using logistic regression, it was found that there is a negative correlation between child marriage and income per capita [22]. The ndings support our study as there was also a negative correlation between household nancial security or income with child marriage acceptability.
A number of studies also have explored that poverty and the lack of viable income-generating options for girls and young women are important factors contributing to high child marriage rates. Skills and nancial incentives are sometimes linked to investing in education in girls and/or on the condition that they do not marry until the age of 18 [23]. Child marriage leads women to have children earlier and more children over their lifetime than if they had married later. It affects girls' educational attainment and literacy negatively, thereby curtailing future opportunities for them to compete for well-paying jobs [17].
A lack of their own income and nancial planning skills has the potential to reduce bargaining power for women within the household as well as investment in their children, affecting future generations and contributing to girls getting married at an early age. Similarly, early marriage is also likely to limit the earning capacity of women by reducing their education, work experience before marriage, and ability to work outside the home while married. Even if women who experience child marriage, later contribute more to household income, the improvement in income alone may not always ensure good health outcomes for them. They and their family members (e.g., their babies) will still be more likely to experience stunting, physical disabilities, and risk of degenerative diseases in the future [23].
Child marriage is often considered a way out to lessen the economic burden on the family. Supporting household economic security may provide an acceptable alternative to marriage and increase the value and contribution of the children, particularly girls to her family, for example, having a higher income. The intervention can be conducted by providing scholarships or support for children to access education and conducting livelihood programs.

Knowledge of legal framework
This study found that the knowledge of the legal framework is a signi cant contributing factor to the acceptability of child marriage in Central and South Sulawesi by using ANOVA, even though it did not remain signi cant after the regression linear analysis. The legal framework was measured by asking whether households have to correct knowledge of the marriage law including whether they register their marriage and own identity documents.
Knowledge of the legal framework of marriage in Indonesia may contribute to child marriage prevention. Indonesia has made progress by increasing the legal minimum of marriage for girls. Previously, according to Marriage Law No.1/1974, Indonesia allowed girls aged 16 years and below to marry. The Marriage Law was amended in September 2019 and it increased the age that girls can be married from 16 to 19 years, the same age as boys. The legal framework is an enabling environment factor for child marriage prevention, even though some studies reported that it is not su cient to create lasting change [17].
Field data from West Bengal reveal that the PCMA (types of legal law codes in India) had success in reducing the number of such marriages or punishing its practitioners. Thus, the state had only six registered police cases of child marriage in 2008. Data for the year 2009 are not yet available. The district of Malda did not register a single case of child marriage until September 2010. The district has state-imposed any injunction to prohibit child marriage in recent times. This reveals that the new law has motivated the relevant persons to implement the law. Conversely, the existing DSWOs and CMPOs (types of legal aid agencies in India) need to wait for changes in social attitudes to enable them to initiate any real action against the violators of the law [24].

Sexual and gender-based violence
This study found that SGBV (sexual and gender-based violence) was a signi cant contributing factor to the acceptability of child marriage in Central and South Sulawesi using ANOVA analysis and linear regression analysis. The sexual and genderbased violence (SGBV) measure explored the acceptance of sexual violence against women/girls and male control over marriage as well as the belief that child marriage prevents sexual harassment.
Child marriage is a manifestation of gender-based violence and a violation of the fundamental human rights of women as many girls are forced into marriage against their freewill and consent. It is harmful to children as it robs them of their childhood innocence and turns them into "adults" prematurely. For girls, the age in which they are married renders them unable to negotiate safe sex and are therefore vulnerable to sexually transmitted diseases and domestic violence. The child brides are forced into sexual intercourse with their spouses as soon as they are married resulting in very early pregnancy and resulting in vesico vaginal stula (VVF) obstetric stula, when such girls give birth eventually at very tender ages [25].
In societies where gender norms devalue girls' position, girls are more likely to experience violence within marriages. They are also more likely to experience physical and sexual abuse than those who marry later. Girls not only experience abuse from their husband, but also from other family members. As girls enter into marriage, they may be forced to have a marital sex earlier than they are ready. Furthermore, it may lead to adolescent pregnancy withave high risk of complications during birth and negative impact on the young mother and the baby. Gender-based violence can also harm children's mental and physical health, and increase the perception of violence as acceptable. Children who witness violence are also more likely to perpetrate violence as adults. Various studies reveal that there was a strong likelihood that violent or child abuse will become a continuing cycle of violence. The rates of abuse are higher among women whose husbands were abused as children or who saw their mothers being abused [26].
Gender-based-power relations between young wives and husbands, parents and in-laws are not equal. The unequal genderbased power relations in female child marriage practices in poor families are related to limitation of knowledge and reproduction of power [27]. Unequal gender relations are continuously reproduced through the imposition of negative social labeling on girls. The prevention of female child marriage requires a comprehensive approach by addressing the social and cultural values, especially promoting equal gender relations. One of the solutions is empowerment based on equal gender perspective [28].

Limitations
As the study used the cross-sectional study design, it assessed the independent and dependent simultaneously (capturing in one single time). The limitation of the study is that we cannot draw the causal relationship between the independent and dependent variable. The association between the variables are predictive in nature as we measured the respondents' knowledge, attitudes and perceptions toward child marriage issues. The further study will be needed to analyse the relationship between these two variables, for example, to analyse the determinant of child marriage with focusing more on the sample from populations who experienced child marriage.

Conclusions
Overall this study concluded that based on CMAI, there were some respondents that have positive perceptions toward child marriage. This study also found that in areas with dowry practices, child marriage is more likely to be accepted by the society as the amount of dowry is also determined by factors such as age, education level, girls' status, and so on. Dowry practices will continue as it is part of the culture in South Sulawesi, however there is a need to change the community perception towards dowry practices in order to address child marriage and gender inequality. For examples whether younger girls will have lower "uang panai" compared to those aged more than 18 years. Sexual and gender-based violence was also a signi cant contributing factor to the acceptability of child marriage. In societies where girls are not valued the same as boys, violence and abuse is more likely. On the other hand, SGBV is also a negative impact of child marriage. Women who were married as children were more likely to experience violence and abuse compared to those married as adults.
To protect children' rights particularly girls, there is a need to combat child marriage by developing interventions and policies to prevent children from marrying and protect those who do marry early from violence and abuse. There is also a need to ensure that all women have the resources and support to leave abusive relationships. Sexual and reproductive health education is also needed for parents and adolescents to increase their knowledge on the issues and improve their behaviors. As Indonesia has increased the minimum age of marriage for women, it could support the child marriage prevention program in national and subnational level. Declarations