Gender Data: Why We’re Stuck in the What, and How to Move On

Eleanor Carey May 30, 2019

Anyone coming to a new subject will be familiar with the feeling that everyone around you is speaking a different language.

Yet there’s something particular, I think, about the words and abstract concepts surrounding the quantitative disciplines that shuts people out. It can take so long to grapple with this language and the exactitudes of the discipline, that it’s easy to lose sight of what all the effort is for. I came to statistics with no mathematics background to speak of, and in those first few months I wasn’t on a learning curve, I was on a vertical scale up a cliff face!

For me, the turning point came when I was able to move out of the abstract and bring to the fore all of my questions about how societies function, particularly for women and girls. But the data to answer these questions was largely absent; I had to change my research question many times until it became something the data could answer. This experience—by no means unusual—raises two issues. The first is the fundamental importance of questions to uncover gaps in our knowledge and relatedly, gaps in gender data about women and girls. The second is that the available data can curtail the types of questions we can ask, stalling the search for the answers we need to have real world impact.

To avoid this stalling, I find it helpful to consider Sociology professor Norman Blaikie’s typology of questions to determine where the focus on gender data currently lies. Blaikie identifies 3 types of questions about the world: what, why, and how. “What” questions give us a description of the way things are: What number of women participate in the labor force? “Why” questions tell us about relationships between events or groups: Why do mothers have lower labor force participation rates than non-mothers? And “how” questions interrogate the exact ways in which we can bring about change: How can we increase mothers’ labor force participation rates? These three question types are iterative and imply increasing complexity. “What” questions form a basic foundation, the answers to which lead to “why” questions, which in turn beget questions centered on “how” situations change.

Different types of data and different approaches to analysis are needed to answer each of these types of questions. But right now, the balance is off. Particularly when it comes to global data on women and girls, we’re stuck in the “what”.

This is understandable. The gaps in our knowledge about women and girls are so rudimentary that we have been focused on getting the basics right, making sure that women and girls are visible and included. And since the advent of the SDGs, much of our data system has been geared towards providing data for indicators, most of which are proportions requiring basic counts. For example, we tend to look to cross-sectional surveys, censuses, and administrative data to answer our questions and build our foundational knowledge, but even with these sources there is much more work to be done, especially given the current focus on pursuing multiple disaggregations.

In order to move towards answering “why” and “how” questions, we need to both further exploit the data we have, as well as invest in new types of data at scale. Longitudinal data, for example, which tracks the same individuals over time, can help to uncover dynamic patterns of change and make connections between events and later outcomes—e.g. between becoming a mother and patterns of labor force participation. Further investment in administrative data and linking of datasets could also help to track the drivers of progress at an individual and population level over time.

In the age of the data revolution, it can sometimes feel that we are racing towards improving data supply without deep thinking on the types of questions we are producing data to answer. This supply-side approach results in wasted resources; but there’s a more profound risk too. If we don’t clearly articulate the questions, we might be producing more of the same type of data, inadvertently narrowing our focus on the “what” questions, without providing the raw material to answer the thornier “why” and “how” questions that demand attention to resolve the most pressing gender issues.


Stay in touch. Sign up for gender data updates.