The next decade is critical to make progress on the Sustainable Development Goals (SDGs), a set of shared goals to improve lives and protect the planet that were adopted by the United Nations in 2015. One of the goals, SDG 11, focuses on making cities and human settlements inclusive, safe, resilient, and sustainable. Every year, disasters caused by natural hazards displace millions of people from their homes. In 2018 alone, 17.2 million people were displaced by disasters. Among internally displaced people (IDPs), women and girls are disproportionately exposed to risk, increased loss of livelihoods, security, and even their lives. For example, women are more likely to be victims of violence during disasters, to be at greater risk of economic vulnerability, and to make sacrifices for the well-being of their families.
According to the U.N., one of the biggest barriers to progress on these goals is not having access to real-time and representative data that can help inform decisions. The same goes for disaster data, where data gaps continue to impede proper understanding of how disasters affect men and women differently. Displacement trends are seldom disaggregated by gender, meaning we do not have displacement data broken down for men and women. Facebook’s Project17 initiative aims to provide partners with additional data, tools, and insights to help make progress on the SDGs, specifically focusing on gender data.
Today, we are announcing the launch of Facebook’s Gender Disaggregated Displacement Maps as part of the Disaster Maps product suite. These maps will help our nonprofit and research partners understand how men and women are displaced differently after a disaster. Our Displacement Maps were developed with the Internal Displacement Monitoring Centre and produce daily estimates of the count of displaced people and where those people are. Partners with access to these maps will now be able to see what proportion of those displaced are men and what proportion are women, where women are relocating, and when both are able to return. All of these calculations use aggregated and de-identified data from people who are using Facebook on their devices and have opted in to location history. This launch is part of Facebook’s commitment to contribute to the SDGs by enabling international organizations, development experts, and other trusted partners to leverage Facebook’s data to bridge gender data gaps, answer research questions, and help drive progress on gender equality.
Typhoon Hagibis hit Japan in October 2019 and was the strongest storm to hit Japan in over 60 years. The severity of the storm led government officials to issue evacuation orders to a significant proportion of the population. With our new Gender Disaggregated Displacement Maps, we can now better understand how the typhoon affected women and men differently. As figure 1 shows, slightly more men (2.33%) were displaced than women (1.92%).
Figure 1: Percentage of people displaced, by gender
Figure 1: Percentage of people displaced, by gender, for Typhoon Hagibis
This new data also allows us to understand whether men and women are able to return home at a similar pace and timing. Figure 2 shows that men had a faster return than women. Women’s rate of return plateaued a few weeks after the disaster, while men’s rate of return continued with an ascendent trend. These insights correspond to our partners’ findings on the ground and could be critical to help them design targeted interventions:
“The analysis of gender-disaggregated data appears to confirm one of our hypotheses — that men return home more quickly than women to repair damaged property and resume their livelihoods, whereas women remain displaced in places where they can access services for themselves and their children. Having data to back this up is important because it has operational implications in terms of what is needed, where, and for how long — particularly for people who are not staying in official evacuation shelters.”
—Justin Ginnetti, Internal Displacement Monitoring Centre
Figure 2: Number of people returned, by gender
Figure 2: Number of people who returned, by gender, after Typhoon Hagibis
In late 2019 and early 2020, massive wildfires burned more than 27 million acres of land across Australia. Economists estimate the impact of the bushfires is about $2.4 billion, and nearly 3,000 homes were destroyed across the country. Using our gender disaggregated displacement data, we can assess how the wildfires affected men and women differently.
We looked at the Green Wattle Creek Fire across Eastern New South Wales and the Cudlee Creek Fire in Adelaide Hills in South Australia to see whether we could detect different trends among men and women.
Two weeks after the start of these fires, our displacement maps showed that 8.8% of the population had been displaced, and again we noticed a difference between men and women: a higher percentage of men (9.29%) than women (8.3%) were displaced.
Additionally, we saw different patterns of location during displacement across genders. Figure 3 shows that women are more likely to be displaced in the same city, while men are more likely to be displaced abroad. We can also see that men and women are equally likely to be displaced in another city within Australia. With this data, our partners can not only plan where to focus their operations when targeting the needs of women, but also understand how men and women may be affected differently by disasters.
Figure 3: Displacement distance, by gender
Figure 3: Displacement distance, by gender, for the Australian wildfires
In addition to what we saw in the data on timing and distance of displacement, our partners told us that there are many ways that women’s displacement experiences may differ from men’s. To learn more about these differences, we surveyed people on Facebook who had experienced at least one night of displacement as a result of Typhoon Hagibis. Two key ways in which people’s experiences differed across gender were in the areas of evacuation decision making and disruption to normal work.
When asked “Whose decision first led you to leaving your home?” women (42.5%) were more likely than men (24.2%) to say that it was a family member’s decision to evacuate (figure 4). In regression models adjusted for age and education, the odds of women reporting that they evacuated in response to a decision made by a family member (as opposed to being the decision-maker) were twice (OR = 2.1) that of men, who were more likely to say it was their own decision.
Figure 4: Whose decision first led you to leaving your home?
Figure 4: Proportion of survey responses by gender, for Typhoon Hagibis
Among people who work, women (52%) were more likely than men (45%) to say they did not work as much as they normally do while displaced as a result of the typhoon. In models adjusted for age and employment type, women had significantly higher odds (OR = 1.42) than men of reporting disruption to normal work, with model-adjusted rates of 64.7% for women not working as much as usual, compared with 35.7% for men (figure 5).
Figure 5: Did you work as much as you normally do while away from your home as a result of Typhoon Hagibis?
Figure 5: Distributions of survey-reported work disruption, by gender
There is much more to learn about how women and men experience and recover from displacement. Our Gender Disaggregated Displacement Maps are a new tool for understanding these differences, and survey data can be a useful tool to complement these trends. Here, we’ve reported that while men were slightly more likely to be displaced overall, displaced women were away from their homes longer, despite their being more frequently displaced in the same city as their homes. Women in Japan experienced evacuation as part of a family decision, and their work was more likely to be disrupted by displacement, perhaps in part due to experiencing a longer displacement duration. Accessing data sources like Facebook’s Displacement Maps, in conjunction with surveys on the platform, is a new approach to improve our understanding of these differences, which is a crucial step toward helping displaced people in ways that are specific to their particular needs.
Partners interested in using Gender Disaggregated Displacement Maps can reach out to firstname.lastname@example.org. To learn more about our Data for Good program, visit our website.