Financial Incentives and an Adolescent Empowerment Program to Reduce Child Marriage in Rural Bangladesh

Nina Buchmann
Shahana Nazneen
Xiao Yu Wang
Barisal, Bhola, Patuakhali, Bangladesh
15,576 girls in 460 communities
2007 - 2017
Target group:
  • Children
  • Women and girls
  • Youth
Outcome of interest:
  • Dropout and graduation
  • Enrollment and attendance
  • Age of marriage
  • Women’s/girls’ decision-making
  • Gender attitudes and norms
  • Age of childbearing
  • Fertility/pregnancy
Intervention type:
  • Financial literacy
  • Incentives
  • Soft skills
  • Empowerment training
  • Norms change
  • In-kind transfers
  • Non-monetary incentives
  • Parental engagement
AEA RCT registration number:

Despite laws in many countries prohibiting child marriage, it remains prevalent and may lead to negative health and education outcomes for young women and their children. Researchers evaluated the impacts of an incentive program and an adolescent empowerment program on child marriage, teenage childbearing, and education in rural Bangladesh. Financial incentives conditional on delayed marriage reduced child marriage and teenage childbearing, and increased education. The empowerment program increased education but did not reduce child marriage or teenage childbearing.

Policy issue

Child marriage, a marriage occurring before the age of 18, is correlated with negative outcomes for women and their children. Women who marry early complete less schooling, face more difficulties in marriage, and are less likely to access prenatal and antenatal care. Children of child brides also tend to have worse educational and health outcomes. While many countries have instituted laws to prohibit child marriage, the practice remains the norm in many countries despite large gains in female education and employment, affecting 650 million women and girls alive today who married before they were 18.1

One reason why early marriage remains pervasive could be reputational concerns. Since less socially conservative girls are more likely to work, they tend to benefit more from additional years of education and from marrying later. As a result, delaying marriage may be seen as a signal of low adherence to traditional gender norms, which is often undesirable to grooms. To avoid these perceptions in the marriage market and signal that they are willing to adhere to desired conservative norms, parents may push their daughters to marry earlier than they would otherwise.

Could a small transfer conditional on parents postponing their daughters’ marriages induce even the most conservative girls to marry later, thereby weakening the signaling dynamics and decreasing early marriage for all girls? Relatedly, could empowerment programs that reduce girls’ adherence to traditional norms unintentionally increase early marriage by strengthening grooms’ beliefs that a woman delaying marriage is not conservative? That is, by making it costlier for women to delay marriage for fear of being seen as nontraditional and less desirable?

Context of the evaluation

As of 2020, Bangladesh had the 8th highest level of child marriage in the world. Fifty-one percent of women aged 20-24 had been married before age 18. Early childbearing was also common for child brides—almost 5 in 10 of them had given birth before age 18—and they were over four times more likely to be out of school than unmarried girls.2 Parents were highly involved in the matchmaking process, and dowries, though illegal, were common.

This study took place in communities in six rural sub-districts of south-central Bangladesh where Save the Children was managing a food security program for pregnant and lactating mothers. In the same locations, Save the Children also designed and implemented an adolescent empowerment program called Kishoree Kontha or “Adolescent Girl’s Voice”. The program curriculum focused on educational support to enhance literacy, numeracy, and oral communication, as well as training on life skills, negotiation, legal rights of women, and nutritional and reproductive health—similar to other empowerment programs being implemented worldwide.

For this study, eligible participants included all unmarried girls in the study communities who were aged 15-17 at the start of the program. At baseline, participants were 15 years old on average and 65 percent were still in school.

A group of girls with their ration cards in one of the study villages.
The conditional incentive program delivered cooking oil to families with unmarried girls aged fifteen through seventeen. Photo: Libby Abbott | J-PAL
Libby Abbott

Details of the intervention

In partnership with the Bangladesh Development Society and Save the Children, researchers conducted a randomized evaluation to test the impact of an incentives program for delayed marriage and an adolescent girls’ empowerment program on child marriage, teenage childbearing, and education.

Beginning in 2007, researchers randomly assigned 460 rural and semi-rural communities to four different groups:

  1. Conditional incentive program (77 communities): All girls aged 15-17 and unmarried at the program start received ration cards indicating their eligibility to receive cooking oil every four months until they married or turned 18. Girls collected the oil by presenting their ration card, which was checked against a beneficiary list at oil distribution points. The value of the incentive was approximately US$16 per year. Cooking oil must be purchased regularly by every family in Bangladesh and thus is a close substitute to cash. Every four months, from April 2008 to August 2010, monitors conducted unannounced household visits to check that the girl was still residing in her parents’ home, and interviewed family members, neighbors, and community leaders about her marital status to verify continued eligibility. Monitors removed girls’ names from the eligibility list and took their ration cards away when they married or turned 18.
  2. Empowerment program (153 communities): In these communities, Save the Children invited all girls aged 10-19 to participate in one of four six-month cycles of their adolescent empowerment program, Kishoree Kontha. The program provided meeting places where girls could meet five to six days a week to socialize and receive educational support and social competency training on topics like negotiation. In half of these communities, the program also included financial literacy training. Program staff trained two to four girls per site to deliver the empowerment program to their peers, and staff checked in on sites every few weeks. The study only analyzed the sample of girls aged 15-17 at program start, the same age range observed in the conditional incentive group.
  3. Combined intervention (77 communities): All girls who met the eligibility criteria for the two interventions received both the conditional incentive and the empowerment program.
  4. Comparison group (153 communities): All girls received none of the interventions.

Researchers surveyed participants and their parents before the programs began and followed up with the parents one year and 4.5 years after program completion. While the first survey collected data on marital status, age, and education of all household members, the follow-up surveys asked parents about their daughters’ history of marriage, childbearing, and education. In total, researchers collected complete data on 15,576 women.

Results and policy lessons

Overall, researchers found that the conditional incentives reduced child marriage and teenage childbearing and increased educational attainment, without affecting dowry or compromising spousal quality. The empowerment program increased educational attainment but did not affect child marriage or teenage childbearing and in fact increased the dowry brides had to pay.

Timing of marriage and childbearing: Girls in communities that received the incentive program were 4.9 percentage points (17 percent) less likely to marry before the age of 18 relative to girls in the comparison group. Younger girls—aged 15 at program start—were exposed to the intervention for a longer period and became even less likely to marry early, with a 7.5 percentage points (19 percent) reduction in their likelihood of marrying before turning 18. These impacts extended to girls living in comparison communities that were close to the incentive communities, whom community members could mistake for incentive program participants. These girls became 2.9 percentage points (10 percent) less likely to marry before turning 18 compared to girls living in more distant comparison communities—an effect that was driven by less socially conservative girls.

Further, girls who received the conditional incentive were 1.6 percentage points (7 percent) less likely to have children during their teenage years. The empowerment program did not lead to any changes in the timing of marriage or childbearing, nor increase the effectiveness of the incentive program when combined.

Education: Girls who received the conditional incentive and were 15-17 years of age at the start of the program were 3 percentage points (6 percent) more likely to be in school one year later. This effect was larger for girls aged 15 (8.6 percentage points or 18 percent), who were also 5 percentage points (18 percent) more likely to remain in school 4.5 years after the program. Thus, the positive effects were more meaningful for girls who were younger at the incentive program’s outset and therefore enrolled for longer.

Girls who received the empowerment program were no more likely to be in school after the program, although they completed 2.1 months of additional schooling relative to those in the comparison group (a 1.6 percent increase). Combining the incentive and empowerment programs was no more effective than providing the programs individually.

The results indicate that the empowerment program encouraged girls who otherwise would have dropped out of school well before marriage to stay in school only up until the time of marriage, increasing their grade attainment but precluding a difference in enrollment at higher ages. In contrast, the incentive allowed more women, who would have otherwise dropped out of school to get married, to instead stay in school past the age of 18.

Dowry size and quality of match: Women in incentive communities did not pay different dowries for marrying later and saw no changes in indicators of the quality of the match, such as the husband’s education, age, or ability to generate income. Meanwhile, women eligible for the empowerment program paid 7 percent (US$39) more in dowry for the same quality of matches relative to women in the comparison group. This suggests that brides can be penalized for participating in the empowerment program or being empowered. As socially conservative brides become rarer in empowerment program communities, less conservative women must either pay more or marry earlier to compete for the same quality match.

Cost-effectiveness: The conditional incentive program was highly cost-effective; researchers estimated that every US$1,000 spent on the program led to 4.9 years of delayed marriage, 1.1 averted child marriages, and 3.7 additional years of schooling.

Taken together, these findings suggest that child marriage in Bangladesh is not a deeply-held preference. Rather, it stems from a desire to signal adherence to traditional gender norms in the marriage market. As such, providing a financial incentive conditional on remaining unmarried was an effective approach to delay marriage for girls in this context, as it allowed young women to delay marriage without revealing themselves to be of an undesired bride type. On the other hand, the effects of the empowerment program were limited by the signaling dynamics at play: girls pursued more education but were not able to delay marriage as the empowerment program made it even more important to signal being a socially conservative bride.

United Nations Children’s Fund. 2021. Towards Ending Child Marriage: Global Trends and Profiles of Progress, UNICEF, New York. Accessed April 21, 2022.
United Nations Children’s Fund. 2020. Ending Child Marriage: A profile of progress in Bangladesh. Accessed March 31, 2022.