Can Media Campaigns Change Attitudes and Spark Actions to Reduce Early Child Marriage? Impact Evaluation of Breakthrough's Early Marriage Campaign in Jharkhand and Bihar States of India

"Breakthrough recognized that early marriage is a deep-rooted social norm that can only be countered if norms which perpetuate early marriage for girls are challenged."

In India, home to one in three of the world's child brides, there have been multiple interventions to reduce incidence of early marriage. The assessments and evaluations of these interventions point towards the importance of media, community mobilisation, and youth empowerment. Breakthrough is a global human rights organisation that uses pop culture and education-based interventions to inspire leadership for social change. This report presents the endline analysis of the impact of Breakthrough's early marriage campaign on various indicators.

Section 2 describes the intervention, theory of change, and research hypothesis. Based on formative research and experiences of Breakthrough, four areas were identified for intervention: perception about the safety and security and harassment of the girls if they are not married early; stigma at the household and Gram Panchayat (GP) levels against unmarried girls; lack of decision-making powers and platforms for youth to say no to early marriage; and the deep-rooted perception that "girls are a bad investment". These findings and issues of interest guided the design of messages and engagement modules to shape and locate Breakthrough's intervention within a human rights framework, with the goal of reducing the incidence of early marriage and increasing the average age at marriage in three districts of the Indian states of Bihar and Jharkhand.

Breakthrough knew that any successful intervention in early marriage would need to examine the individual-, familial-, and community-level norms and practices and contextualise strategies appropriate to each level. The normative underpinning (explained in greater detail through the qualitative analytic framework in Section 5.2 of the report) warranted a response steeped in social norms and behaviour change theories. Thus, the intervention sought to replace prevailing notions and beliefs about the value of the girl child, as well as the desire to restrict her mobility and decision-making power, with notions that perceive women and girls as equal contributors to the household and need to be treated with respect and afforded equal rights. Its focus on gender and social norms served as a base for intervention strategies, which included:

• Mass media campaign: Messages were aired on radio and television, in newspapers, and at cinemas. (See one example by clicking on the video, below.) The messages were positive and urged young fathers and boys to act against early marriage. The campaign encouraged young fathers to see their daughters as productive members of the household. It operated in all 280 GPs of intervention districts.
• Training: At the block level, Breakthrough conducted programmes to sensitise, train, and mentor representative of youth groups, self-help groups, educational institutes, etc. and key field staff in the government (departments of education, health, women, and children) on modules around the themes of gender rights, sexuality, and sexual harassment. It was expected that the representatives from community groups and the government field staff reached at the block level would carry messages and share the same at the GP level and below in their interactions. In addition, school-based trainings involved Breakthrough's staff working with adolescents and their teachers, and GP-based trainings engaged community influencers such as Anganwadi Workers and Panchayat members. The basic premise of these trainings was to combine the internal capacity within these groups with information and awareness regarding early marriage, towards creating agents of change within the GP. The training package was implemented in 21 blocks.
• Community mobilisation and interactive media: Breakthrough organised GP-level events such as Kishori melas (youth festivals), puppet shows, and video vans and used ambient media such as posters and wall paintings - all geared towards combating norms around early marriage. These components were implemented in 40 GPs.
• The full package (FP): The FP, which included all three components of mass-media, training, and community mobilisation, was implemented in 40 GPs.

Section 3 details the context of the of the study, including the information regarding study area and survey sample. The time of the intervention and study is briefly outlined in section 4. Section 5 describes the study design, the treatment-assignment and identification strategy, and measures to ensure data quality. In brief, the evaluation employed a cluster randomised control design employing a mixed-method procedure.

• The quantitative study sought to examine the changes in impact-level indicators such as age at marriage and incidence of early marriage across each treatment package. It sought to assess the extent, if any, of attributability of these changes to the programme through a panel study with baseline (in 2012) and endline (in 2019) with a midline verification (in 2014). The final analysis sample constituted 2,542 households.
• Qualitative tools - i.e., a polling booth, focus group discussions (FGDs), key informant interviews, and case studies - were employed for elaboration, enhancement, illustration, clarification, and triangulation of the results from the quantitative method. The qualitative study also examined any spillover of the programme components between treatment and control areas. The qualitative framework was based on social norms theory, which underpinned the programme's theory of change.

Section 6 provides information pertaining to how the intervention was implemented, along with monitoring information. The results are described in section 7. Brief highlights:

• As compared to the control GPs, the FP intervention increased the age at marriage for girls aged 13-25 on an average by 6.5 months and their education by 9 months. The FP also increased the percentage of girls enrolled in school by an additional 9% over the level that the control group achieved (92% in FP vs. 83% in the control group). The programme impacts were similar after controlling for stratification and other household and GP-level characteristics.
• GPs with the FP had a 0.407 standard deviation (equivalent to approximately 16%) higher gender attitude index than control GPs (p < 0.01). The measure of attitudes is an index of aggregated indices on gender equality, education, marriage, mobility, and knowledge. Attitude change was larger for education and mobility indices. For example, the households in the FP had 0.26 and 0.33 standard deviations more positive attitude towards girls' education and mobility than control GPs.
• One outcome of the knowledge about the legal age of marriage, laws, and negative consequences for early marriage was attempts by community leaders and members to stop incidents of early marriage. Evaluators learned of such incidents from one GP receiving the FP and one the community mobilisation package. From the latter GP, one adolescent girl said, "Breakthrough has had a lot of impact. People have become afraid of the consequences of early marriage. Girls tell their parents that they will approach Breakthrough if they are pressured into early marriage."
• There was no significant effect on attitudes pertaining to marriage and girls' work and responsibility. And there was little change in norms related to gender roles within the home; education is still mostly perceived as a vehicle for better management of the home, rather than for empowering women to work and be independent.

Section 8 discusses and interprets the results. One insight from the analysis is that the intervention led to a sizeable reduction in gender-based discrimination and gender-stereotyped behaviour. However, gender-egalitarian attitudes were driven mostly through the treatment effect on education norms, mobility norms, and knowledge norms, and not by attitudes related to women's workforce participation or to early marriage. The researchers note that the study area has particularly strong gender discrimination and amongst the highest rate of early marriage. The impacts of a similar intervention might be either larger or smaller in places with less gender-biased attitudes.

One direction for future research is to observe if the changes in the gender norms have materialised into gender-equitable intrahousehold allocation of resources. While the data show significant positive effect of the mass media coupled with community mobilisation and training programmes on education gender norms and level of schooling for adolescents, gender-differential education resource allocations are not explored here. Another question is whether the effects of this programme are sustained in the long run, leading to improvement in women's empowerment, labour force participation and occupational choice, and reproductive health and fertility years after the intervention has ended.

Section 9 provides specific insights for policy and practice. For example, in the geographies where Breakthrough works, there are already several civil society organisations running interventions at the grassroots level. Most programmes have mobilised communities and facilitated forming institutions via local resource persons. Collaborating with these agencies and incorporating appropriate components from the FP may be a way to reach scale. Along these lines, Breakthrough could develop detailed documentation and communication materials and codification of its packages for uptake by organisations interested in working on reducing early marriage or addressing gender norms.

In conclusion: "Despite being resource intensive, a multi-pronged, multi-stakeholder focused and comprehensive intervention that reinforces messages at different segments of the society...has potential to bring change in areas influenced by sticky social norms like early marriage. The changes in attitudes that are emerging among key constituencies have to be sustained to bring a large social change....Change agents from key constituencies are emerging. At the same time, it is also important to scale up to saturate a block and a district with the intervention towards reaching a tipping point from where the social change advances untethered."