A video experiment uncovers the potential pitfalls of relying on legal reforms alone to reduce female early marriage in low-income countries

The practice of early girl marriage (i.e. marriage before the age of 18) exists in most regions of the world and is particularly prevalent in developing countries. In the 47 least developed countries, the latest available figures show that 38% of women aged 20-24 years were married before the age of 18  (UNFPA 2019). Previous research confirms that the practice has myriad negative consequences, including lower educational investments for women, lower human capital investments in their children, adverse health effects from early child bearing, and weak social networks (Field and Ambrus 2008, Sekhri and Debnath 2014, Chari et al. 2017, Amin et al. 2018, Asadullah and Wahhaj 2019, Sunder 2019). While research highlights a variety of socioeconomic policy options that may tackle the problem ranging from cash transfers to women’s empowerment programming, there is limited evidence regarding the efficacy of legal solutions. 

Using the law to combat early marriage

In an effort to combat female child marriage, many governments in developing countries have introduced or reformed  laws to criminalise and penalise the practice (UNFPA 2020). However, it is unclear how effective the formal law can be for changing people’s attitudes and behaviour when the practice is in part governed by social norms (Ortner 1978; Dube 1997; Wahhaj 2018). This is particularly true in countries with weak legal enforcement, where the deterrence effect of legal punishment may be absent. For example, one study of a reform in Mexico which increased the legal marriage age showed that the legal change only served to convert formal marriages into informal unions, effectively driving marriages underground (Bellés-Obrero and Lombardi 2020). The authors conclude that “increasing the minimum age of marriage to 18 years old in Mexico reduced formal marriage but did not prevent school dropout or early motherhood”.

The expressive effect of the law

Nevertheless, some scholars, including legal theorists, have argued that the law may be effective in impacting behaviour. Even in the absence of strict enforcement, the law may have an “expressive effect,” influencing attitudes, norms, and behaviour by “sending a message about society’s values” (Sunstein 1996, McAdams 2000, Benabou and Tirole 2012). For example, Chen and Yeh (2014) provide evidence from the United States that simply providing information to individuals about the formal law may be sufficient to shift their attitudes regarding certain social norms. 

A video experiment to study the potential for laws to shift child marriage attitudes and behaviour

Motivated by these questions and hypotheses, in Amirapu, Asadullah, and Wahhaj (2020), we study  whether a change in the formal law regarding child marriage could impact the attitudes and behaviour of individuals. 

A new law to combat child marriage was approved by the national parliament of Bangladesh in March 2017. We randomly administered a video-based information intervention in June 2018 aimed at accelerating knowledge transmission in rural areas about the new law.

The video took the form of a short fictional drama involving the early marriage of an adolescent girl. The study respondents viewed one of three versions of the video on a handheld electronic device: 

  • A control group watched a version of the drama that referenced only the pre-2017 child marriage law. 
  • A treatment group watched a version of the video that referenced the new child marriage law, specifically the introduction of harsher punishments for facilitating child marriage. 
  • A second treatment group watched an alternative version of the video that referenced both the harsher punishments in the new law as well as a special clause in the law that permits child marriage in cases where the court gives its approval. 

Apart from these informational differences, the versions of the video were, shot-by-shot, nearly identical. 

Immediately following the information intervention, we measured a range of outcomes, including study participants’ views on appropriate marriage customs and their beliefs about attitudes towards early marriage in their own community. At the end of each individual interview, the study participants were given the opportunity to contribute part of their remuneration for participation to a prominent charity in Bangladesh that works on child marriage prevention. We conducted follow-up interviews with the study participants five and ten months later to collect information on marriage outcomes for their daughters who were unmarried at the time of the intervention.


Our study found that simply providing individuals with information about a new law was sufficient to change individuals’ stated attitudes and preferences regarding child marriage in some subtle ways. Individuals informed only about the harsher punishments for child marriage stipulated in the new law were subsequently more likely to contribute – and to contribute a larger amount, by 6 Bangladeshi taka on average – relative to a control group, to a charity whose aim is to combat child marriage. Individuals informed about both the harsher punishments and the exception clause were no more likely to contribute to the charity than the control group. Moreover, such individuals were more likely to report a lower appropriate marriage age (by 0.2 years on average). We found little effect from either treatment on participants’ beliefs about attitudes within their own community towards early marriage.

Five months after the intervention, we found that adolescent girls living in households where one or more adults had been informed about the harsher punishment in the new law (but not the exception clause) were more likely to be married (by 7.2 percentage points) than those in the control group. These effects remained at 10 months after the intervention. (See Figures 1 and 2 for a comparison of means in these outcomes across treatment/control groups.) 

Figure 1

Figure 2

These effects are absent in households in which only the mother of the adolescent girl viewed the treatment video, but large and statistically significant when the video is viewed by both the mother and (separately) by other members of the extended family – either the father or a family elder. The effects are also smaller in the case of households that received information about both the harsher punishment and the exceptions clause.


Our results have a number of implications for understanding whether and how the law can influence traditional marriage practices in a setting with weak legal enforcement. First, our evidence suggests that merely providing information about a new law is enough to change attitudes, behaviour, and marriage-related outcomes. However, the findings also highlight the possibility of a ‘backlash’ effect against a new law, as the intervention led to an acceleration of marriages for adolescent girls, the very behaviour that the law was meant to discourage. The 'backlash' effect we observe may be due to a perception of an increase in future enforcement of the law and/or state support for agency among adolescent girls. Finally, the results show that the strategic interests of mothers of adolescent girls are not aligned with those of the father or family elders. Consequently, it matters who within the extended family is informed about a legal reform. 

These findings highlight the challenges and potential pitfalls of relying on legal reforms alone to stem the practice of early marriage in low-income countries. 


Amin, S, J Saha and J Ahmed (2018), “Skills-building programs to reduce child marriage in Bangladesh: A randomized controlled trial,” Journal of Adolescent Health 63(3): 293-300.

Amirapu, A, M N Asadullah, Z Wahhaj (2020), “Can child marriage laws change attitudes and behavior? Experimental evidence from an information intervention in Bangladesh”, Working Paper

Asadullah, M N and Z Wahhaj. (2019), "Early marriage, social networks and the transmission of norms", Economica 86(344): 801-831.

Bellés-Obrero, C and M Lombardi (2020), "Will you marry me, later? Age-of-marriage laws and child marriage in Mexico", Journal of Human Resources: 1219-10621R2.

Benabou, R and J Tirole (2012), “Laws and Norms”, IZA Discussion Paper No. 6290.

Chari, A V, R Heath, A Maertens, F Fatima (2017), "The Causal Effect of Maternal Age at Marriage on Child Wellbeing: Evidence from India", Journal of Development Economics 127: 42-55.

Chen, DL and S Yeh (2014), “The construction of morals”, Journal of Economic Behavior and Organization 104: 84-105.

Dube, L. (1997), Women and Kinship: Comparative Perspectives on Gender in South and South-East Asia, United Nations University Press Ltd. Tokyo. 

Field, E and A Ambrus. (2008), "Early Marriage, Age of Menarche and Female Schooling Attainment in Bangladesh", Journal of Political Economy 116(5): 881-930.

McAdams, R H, (2000), “A focal point theory of expressive law”, Virginia Law Review, 1649-1729.

Ortner, S.B. (1978) "The Virgin and the State", Feminist Studies 4(3): 19-35.

Sekhri, S and S Debnath (2014), "Intergenerational Consequences of Early Age Marriages of Girls: Effect on Children's Human Capital," Journal of Development Studies 50(12): 1670-1686.

Sunder, N (2019), "Marriage age, social status, and intergenerational effects in Uganda", Demography 56(6): 2123-2146.

Sunstein, C R (1996), “On the expressive function of law”, University of Pennsylvania Law Review 144(5): 2021–2053.

UNFPA (2019), State of World Population 2019: Unfinished Business, United Nations Population Fund, New York.

UNFPA (2020), State of World Population 2020: Against My Will, United Nations Population Fund, New York.

Wahhaj, Z (2018), “An Economic Model of Early Marriage”, Journal of Economic Behavior and Organization 152: 147-176.

Behaviour Laws Child marriage Child marriage; Dowry; Gender; Financial incentives; Bangladesh