The case for collecting disaggregated data on women and girls is well-established, yet with only 13% of countries currently possessing dedicated budgets for collecting and analyzing disaggregated data on women and girls, we have a long way to go in fulfilling our commitments on gender SDG targets.
However, this should not discourage us from trying to put what we already have to better use. The recently published ‘Ready to Measure: Phase II Indicators Available to Monitor SDG Gender Targets’ by the Data 2X partnership at the UN Foundation has identified “20 indicators (16 identical to or closely related to SDG indicators and four additional, complementary indicators) that are publicly available, albeit with varying levels of quality and coverage1.
The report is a must-read for data enthusiasts and those invested in moving the needle on ‘gender equality’.
There are three strands of furthering the goals on data focused on women and girls that may be useful for us to consider. Much of this thinking is influenced by our work in developing data systems for SDG reporting and use in Asia and Africa.
Good Data: Interoperability and Use
Data interoperability is the ability to process data assets from multiple sources and in multiple formats, and to integrate them into coherent information products or services (e.g. for mapping, visualization or other forms of analysis)2. Interoperability is key to effective data use.
Understanding how existing data sets interact with one another and evaluating the extent of ‘interoperability’ between data indicators on women and girls, where we are confident of data quality, can be critical to establishing how much we can learn from the current data.
Mapping the most critical data points relevant for women and girls, and then finding ways to integrate data sets across sectors and departments on the status of their well-being. For example, owing to relatively strong coverage on demographic data and civil registration, we have a reasonable database on the incidence of early marriages and adolescent child birth. If this data is designed to be interoperable with the content on the health information gathered on young mothers, it will substantially enhance the quality of intervention design and targeting.
There are many approaches one can use in evaluating the extent of interoperability or enabling greater interaction between existing data sets. For example, the Resource Description Framework (RDF) allows us to integrate disparate data streams based on geo-spatial relationships. The goal of such a framework is to provide a first-step towards enabling more data-driven analysis and management of existing data systems.
Prospectively, knowing what the programmatic priorities are and having the right frameworks and data sharing agreements in place, while thinking about interoperability, procuring vendors and setting up data systems, is critical to ensuring the data generated is put to effective use. Here, one way to start is by identifying all departments which have programmes related to these 20 indicators, and ensuring that the intersection indicators are measured as a part of the program log frames and reporting requirements. This will ensure that at least these 20 indicators are reported consistently, across geographies and by multiple agencies (enabling triangulation).
Bad Data: Data Standards and Levers for Change
Data reliability and representativeness are critical attributes to determining quality. The presence of data standards and collection protocols is a critical first step in this direction. Once created, these protocols must be accompanied by a system of incentives (pull factors) and compliance drivers (push factors).
Traditionally, good data on women and girls has been collected in those countries where politically, there is a prominent level of interest in the welfare of its women and girls (for ex: women headed countries or countries like Rwanda, Bolivia and Dominican Republic with high representation of women in government). Replicable incentives for driving ‘good data’ on women and girls can come in the form of additional grants supported by third party audits of data quality on women and girls. Gender data audits, where the government’s ability to build systems for generating good data that is sex-disaggregated is evaluated, can also double up as a compliance lever for accessing donor funding for initiatives.
Building the capacities of statistical bodies in countries and creating a body of work which can establish the relevance of models which are sensitive to women and girls in influencing non-gender outcomes, such as program cost effectiveness, is also critical to empowering and encouraging governments to include gender lensing into their reporting and generating more credible data on their women and girls.
No Data: Hierarchy of Choice for New Data Generation
Where no data exists, an assessment of the most cost-effective approaches to generate data sustainably must be evaluated. The hierarchy of choices that support new data production include: technology (use of machine-based/automatic systems of data generation), administrative reporting processes, government data collection initiatives (census and sample surveys) and private data collection processes in that order.
For example, use of SMS based inputs crowdsourced from women and girls on ‘safety perception’, along transportation routes, is a smart and simple way of gathering data on an ongoing basis.
Knowledge of what we have and don’t is a great starting point. Thanks to Data 2X and their efforts, we already have a useful view of this. It is time to put this knowledge to ‘use’ and drive specific actions to move the needle on data focused on women and girls, one megabyte at a time.