BANGKOK — Collecting, analyzing, and communicating gender-disaggregated data are crucial steps in unlocking opportunities for women and girls, and in tracking progress toward the Sustainable Development Goals.
A new report from UN Women is pessimistic about progress towards the Sustainable Development Goals on women and girls, and points to the need for better data to drive change.
“We have less than one-third of the data we need to monitor the gender-related targets in the SDGs,” Papa Seck, chief statistician for UN Women, told Devex during a recent webinar. “This has as much to do with lack of data production as it does with poor data dissemination strategies by national statistics systems.”
Yet for all actors involved, closing these data gaps is intensely complicated — often fraught not just with technical challenges but political and societal barriers, too. Gender data’s many implications makes it perplexing from a strategic planning and communications standpoint, and utilizing it for advocacy purposes comes with its own set of questions.
Devex spoke with Seck, along with gender data experts from Equal Measures 2030 and the Bill & Melinda Gates Foundation, to break down several best practices in collecting, analyzing, and communicating gender data.
1. Do consider how you are framing an issue first.
Holes in gender data stem from a lack of measurement for many events in women’s lives, including unpaid work and labor migration. It’s important to foster a deep understanding of the local context and to keep in mind that data — or its absence — is political, according to Yamini Atmavilas, India lead on gender equality for the Gates Foundation’s India country office.
Globally, less than one-third of countries disaggregate statistics by sex on informal employment, entrepreneurship, and unpaid work, or collect data about violence against women, for example, according to Data2X. But advocates advise that this and more is required to contribute to policy dialogue and move the needle on women and girls’ empowerment.
Data collection should start with a theoretical understanding of what you’re trying to measure and why, Atmavilas said.
“Even today, we often neglect to understand that formal and informal work are along a continuum for many women,” said Atmavilas, who explained as an example that a woman cooking food for her family might also be selling some of it at a market — a common occurrence that can go unrecognized in data collection.
The nature of women’s migration is another area that can suffer from assumptions, namely that it is only related to a change of residence upon marriage. Women’s labor migration, on the other hand, is “often left out of a lot of our conception framings and it ends up being an invisible part of our work when we try to measure,” she said.
“Easy-to-use data is available, accessible, and actionable.”— Albert Motivans, head of data and insights for Equal Measures 2030
Framing can only come from those who know the issues well, and statisticians are not always best placed to do that, according to UN Women’s Seck. Deeper policy analysis should come from people who understand gender concerns, which is why UN Women is focusing on improving a dialogue between data users and producers.
2. Don’t let the absence of baseline data stop you.
When designing and setting up evaluations for government programs, Atmavilas has often found herself scratching her head over the absence of a baseline, a set of data collected at the beginning of a study or before an intervention has occurred.
During one evaluation of a program that used bicycles as an incentive for adolescent girls to stay in school, “one of the biggest challenges I faced was absence of any data other than how many girls received the bicycles,” she said.
Atmavilas took on the difficult task of constructing retrospective assessments, going back into the communities and tracing some of the girls’ experiences four or five years prior to the program in order to find out what affect the bicycles may have had on school retention, attendance, learning outcomes, and the girls’ own agency.
There is also a “classic challenge,” she added, when an established baseline that examines mobility or decision-making may no longer be telling you the most interesting things about what’s changing in women’s lives based on certain interventions.
In these cases, it’s time to turn to innovation in sampling or levels at which data is collected: “Perhaps at higher, more aggregated levels you hold on to certain measurements for the sake of compatibility, but at the granular level you can build in the newer, more updated ways of measuring so that they become more usable.”
3. Don’t be afraid to diversify your data sources.
Easy-to-use data is available, accessible, and actionable, according to Albert Motivans, head of data and insights for Equal Measures 2030.
Equal Measures has charged itself with putting easy-to-use data in the hands of girls and women’s movements and is currently constructing a global index around gender-related SDG targets to provide a comprehensive picture of progress across gender equality issues.
Motivans referenced several other useful places to find a wide-ranging look at the lives of women and girls, such as the World Economic Forum’s Global Gender Gap Report, United Nations Development Programme’s Gender Inequality Index, and Georgetown’s Women, Peace and Security Index. UN Women, meanwhile, is set to launch a portal during the World Data Forum in October to provide a comprehensive look at gender-related SDG indicators.
But it’s also important to complement indexes or other official indicators with other measures, like perceptions around social norms, and with data generated by girls and women, Motivans said. It’s “almost a forensic exercise,” he said, of piecing together different clues that can better tell a story about the lived realities of girls and women.
UN Women, which is working with statistics offices to increase demand for gender data at the national level, has found that “there is a lot of data that is being produced but that is outside of the national, official statistical systems,” Seck said, such as data that comes from programming or from other government ministries. The agency is now working with many offices to enable them to use the data they generate through their own programming, he added.
Atmavilas recommends being open in terms of both qualitative and quantitative metrics and recognizing that there are many interesting new ways of capturing information and data aside from simple surveys or “hundreds of multiples choices.”
4. Do be an informed data user.
One thing Seck has learned through his work with policymakers and advocates is that they often don’t know where to look for the information they need. Other times, they continuously have to chase data producers, yet receive little to no support to understand the data being given to them.
“The mantra of ‘know your data’ is not only important for researchers but really important for advocates and really important for policymakers,” said Equal Measures’ Motivans.
At the 2018 Australasian Aid Conference, a new report from the Research for Development Impact Network outlined the impacts the research community has had on aid policy. Speaking with Devex, RDI Network Project Steering Group member Joanne Crawford discussed the findings and recommendations for both researchers and donors to create an aid program that better utilizes evidence-based policy.
This starts with demystifying technical jargon. Even simple and common indicators like a rate or ratio can seem complicated if the data isn’t explained well, Motivans said. Other measurements, like stocks and flows or prevalence and incidence are vital for advocacy and for making policy decisions — which means they’re even more vital to explain and understand in the first place.
UN Women is developing data literacy programs to ensure that the basics of gender statistics are available to national statistics offices, for example.
5. Do consider ways to make data collection participatory.
At the end of the day, Atmavilas would like to see less “data extraction” and more of an effort to go back to the communities where data was collected, armed with useable data that population could use for their own planning and advocacy.
There are ways to make collection more participatory, she said, and to ensure that communities you are doing research with are “part of the generation of hypotheses, of the tools, and are partners in the research as well as in use of that data for their own communities.”
Unfortunately, this still happens too rarely, she said, and finding ways to build data systems that are accessible, transparent and grounded with communities should be a bigger part of the data conversation moving forward.