The answers you seek aren’t always the answers you find. This happened to a team of FHI 360 and Vanderbilt University researchers when they found an unintended consequence while evaluating a girls empowerment intervention in Mozambique. The good news is that Rachel Lenzi et al. expanded their original evaluation to learn more about why and how this unintended consequence occurred, and even better, they took the time to publish their findings in a journal article! Here we summarize the article.
The unintended consequence
Based on this discovery, Lenzi et al. adapted their protocol for the second round of interviews. The authors asked specific questions to determine what respect meant to respondents and what made them believe the intervention positively influenced respectful behavior. They found that respondents generally believed Women First trained girls to behave respectfully, which they defined as: showing humility and deference to their parents and others; being productive and serving others within their household and community; and being sexually abstinent or chaste and behaving modestly.
Making sense of the unintended consequence
The authors suggest both positive and negative ramifications of the three dimensions of being a good girl on the four main objectives of the intervention: reducing girls’ HIV risk, increasing girls’ school attendance, reducing girls’ gender-based violence risk, and increasing girls’ empowerment. For example, on the one hand, being sexually abstinent reduces HIV risk; on the other, being deferent inhibits a girl’s ability to refuse unwanted sex. More obedience could reduce schooling for married girls, but sexual abstinence could reduce early marriage and thus increase schooling for unmarried girls. Being more productive and service oriented could hurt empowerment by reinforcing norms of unpaid work but could also help empowerment by increasing engagement with the larger community.
The authors present a table of the potential positive and negative relationships. One thing we notice when looking at that table is there are no positive consequences of the good girl norms when it comes to reducing gender-based violence risk.
Avoiding unintended consequences when designing a program
While you can’t anticipate all possible consequences (hence “unintended”), you can do your best to understand the possible effects of your intervention. Lenzi et al. lay out a few quick tips to help avoid unintended consequences in your program design process.
First, do careful formative research. In their endline analysis, Lenzi et al. find that gender norms about respect already existed pre-intervention and Women First unintentionally reinforced those deeply embedded norms. The authors acknowledge that careful formative research can uncover community and individual perceptions of gender norms which can then be leveraged to positively impact the intervention outcomes.
Second, include the community. Lenzi et al. acknowledge that women empowerment programs require institutional changes within the community, and therefore also require community-targeted components. Women First specifically targeted young girls but overlooked important connected populations that could affect (and be affected by) the intervention: community leaders, parents, teachers and family members. It’s critical to understand how the intervention will affect connected populations (in addition to the target population) when designing an intervention. This is where unintended consequences may be lurking, and where you can shine a light on them.
Third, collect and share appropriate program monitoring data. Lenzi et al. note that when you share data or disseminate such findings, others can see possible unintended consequences (that they might otherwise overlook) of a similar program, as well as the pathways by which those unintended consequences can occur. Collecting and analyzing data throughout the intervention – not only at the end – allows you to find potential unintended consequences earlier so that you can adapt the program accordingly.
When you do find unintended consequences, we hope you do what Lenzi et al. did – research them further and publish the findings. The answers you actually find might be the answers others seek.