Honduras and El Salvador have some of the highest rates of violence against women in the world. Six out of 10 women in these countries have experienced violence at some point in their lives. More than 40% don’t seek help, and when they do, less than 10% make use of legal or medical services. Sadly, 1,616 women were murdered in the region in 2019—double the global average.
The COVID-19 pandemic has made the problem worse. Rates of violence increased with lockdown restrictions. In 2020, Honduras’s emergency hotline received over 100,000 domestic violence calls and in El Salvador, sexual violence was perpetrated once every four hours.
Since 2018, BIT has been partnering with the Inter-American Development Bank (IDB) to explore how behavioral insights can encourage women in these countries to seek help. The results so far are encouraging. Amid pandemic lockdown restrictions, our two large-scale trials have directed more women to life-saving resources online.
Can behavioral insights address barriers to seeking help?
Seeking help isn’t easy, especially for women experiencing violence. It’s a complex journey—from realizing that one is in a violent situation to finding help and building the conditions to leave. The slightest barrier in the process can determine if someone finds safety or not.
A behavioral insights approach can help break down the complex process of getting help into specific targeted behaviors and identity challenges women face at every step.
We collaborated with the Salvadoran Institute for the Advancement of Women, the National Women’s Institute (INAM) in Honduras and Ciudad Mujer in both El Salvador and Honduras. Ciudad Mujer is a government program offering support services for women, including violence response.
We began by reviewing behavioral science literature and conducting qualitative research with survivors. We then conceptualized the help-seeking process into three stages, illustrated below.
Some of the behavioral barriers women face in this process are:
- Uncertainty aversion: Our tendency to favour known risks (e.g., their intimate partner) over unknown factors or benefits (e.g., accessing support services but also smaller factors such as ‘what will they ask me when I call the helpline?’)
- Sunk costs fallacy: When we continue a behavior or endeavor (e.g., a relationship with an violent partner) as a result of previously invested resources that cannot be recovered (e.g., time or effort), even though it diminishes our overall wellbeing.
Testing these ideas: Two online trials
We initially planned to run in-person field trials with the IDB. But the COVID-19 pandemic and resulting lockdowns presented a new challenge: reaching women who were isolated with their aggressors quickly, through safe channels. Physical interventions like posters and flyers were no longer an option.
Collaborating with our partners in both Honduras and El Salvador, our team developed two online randomized controlled trials (RCTs) testing Facebook advertisements that encouraged survivors to find support.
Advertisements addressing behavioral barriers
In Honduras, we designed a five-arm RCT. The intervention took the form of four different Facebook advertisements, each addressing a behavioral barrier we identified during our initial research: uncertainty aversion, availability bias, sunk costs, and lack of safety plans.
This ad addresses sunk cost fallacy and guides users to Ciudad Mujer’s website
The ads encouraged women to visit the Ciudad Mujer webpage, which has resources for survivors of gender-based violence. Our control consisted of an image that simply provided information. We measured the effects of the intervention by the number of users who clicked through to Ciudad Mujer.
Our sample included 829,445 women. We randomly assigned people to groups of about 165,000 each to view either an advertisement intervention or the control image on their Facebook feeds.
Women who saw the ad dispelling uncertainty aversion were 19.4% more likely to visit Ciudad Mujer’s website than those in the control group, followed by those who saw the ad addressing sunk costs, who were 12.9% more likely to go to the website.
If scaled to our entire sample, the uncertainty aversion ad would have prompted about 6,635 more women to visit Ciudad Mujer’s website and explore life-saving services.
Infographics to help expose violence
In El Salvador, we ran a three-arm RCT to test the effects of the Violentómetro infographic (below), which explains what different types of violence can look like.
The Violentómetro explains what is considered violence and when you should seek help
Because this thermometer visual is already used across five countries in Latin America, we also tested a new infographic that was informed by behavioral science—Rueda de la Violencia. This image depicted violence as a cyclical process (instead of linear) that can escalate at any point in time. It also showed all forms of violence (e.g., physical, psychological, sexual) as equally severe, instead of classifying some types as less harmful than others.
Our sample included 716,279 women Facebook users. We measured their likelihood to view the content on Ciudad Mujer’s web page based on the image they saw in their feeds.
Both intervention images significantly increased help-seeking behavior. Women who saw Violentómetro were 137% more likely to access resources on the Ciudad Mujer website than those in the control group. Those exposed to the Rueda de la Violencia were 56% more likely to visit Ciudad Mujer online.
Improving communications to increase help-seeking
The results of both of our trials are promising. But this is just the beginning. We recommend scaling these studies more widely and testing the infographics in more timely and relevant environments, both online and in-person. If you’d like to learn more about BIT’s work in Central America or partner with us, please contact us here. Both BIT and the IDB are committed to using evidence and behavioral science to prevent all forms of violence against women and girls.