In parts of India, female domestic workers from the Dalit face caste-based discrimination that results in social exclusions such as segregated facilities and a lack of labour protections. For these women, gender, caste and socio-economic status are some of the factors that overlap to reinforce their marginalization. 

Within data for development, taking an intersectional approach sheds light on who is being left behind. It enables us to examine systems that reinforce inequalities, identify the voices of those at greatest risk of marginalization and take action to address these issues.

The Inclusive Data Charter (IDC), Champions and partners recently launched a series of knowledge products for practitioners unpacking intersectional approaches to data. This research drew on the combined experiences of our partners to distill the key strategies to implement intersectional approaches to data. 

Here are five tips for promoting data inclusivity, drawing on the work of our partners across the globe. 

1. Center marginalized voices in research. When the Centre for Internet & Society (CIS) undertook a project to build digital platforms in the domestic and care work sector in India, researchers initially planned to ask direct questions about how caste discrimination impacted women from Dalit and indigenous communities. But members of the Domestic Workers Union who were included as co-researchers on the project cautioned against this approach based on their personal experiences of domestic work because of the sensitivity of the subject. CIS researchers adjusted their questions in response. Engaging union members in research design brought the realities of domestic workers' experiences to the forefront, enabling more robust data collection and project design.

Centering people’s voices requires considering the range of factors that impact personal experiences of marginalization. Age, gender, ethnicity and disability status are just the beginning. Such a person-focused and inclusive approach leads to better data and research design. 

2. Promote equity across the data value chain. Inclusive data practices start with asking critical questions at every step from data collection through dissemination and publication: Who is being excluded? What systems enable these exclusions? And what actions can be taken to reduce inequalities? 

This enabled the UK’s Office for National Statistics Centre for Equalities and Inclusion to produce COVID-19 data that reflected diverse and disparate impacts. Researchers conducted an audit among a wide range of stakeholders to understand data gaps and opportunities. They used the findings to produce data disaggregated by social group and religion and to measure COVID-19’s impacts on people with disabilities. 

Development Initiatives (DI) uses the P20 method as a simple, practical test to target aid, track progress and measure inequality. P20 involves selecting an indicator and focusing on people in the lowest 20th percentile. DI has developed two other methods to explore data systems from an intersectional perspective and centre the views and perspectives of those at greatest risk of marginalization. DI incorporates these three intersectional approaches into all its projects.

3. Use data to increase contextual awareness. Practitioners must understand local historical, political, economic and social trends to effectively identify and address inequalities. In 2019, the Internal Displacement Monitoring Centre (IDMC) used context analysis in a study of the impacts of displacement on children’s education in Somalia. Using an intersectional lense enabled IDMC to design data collection in ways that captured particular vulnerabilities that differed among groups and uncover gender disparities in school enrollment.

Developing processes to deepen understanding of the local contexts and challenges faced by marginalized groups should be ongoing and continually informed by data. Sightsavers used data to develop and improve rapid assessments for avoidable blindness in Kogi, Nigeria. Sightsavers staff took users’ needs into account to make content accessible to people regardless of the type of disability they experienced.

4. Make institutional data systems inclusive and safe. Creating inclusive and safe data systems requires deliberate actions to drive systemic change. In one example, Kenya’s National Bureau of Statistics (KNBS) is developing inclusive data processes under a national data strategy which calls for strengthening frameworks with regards to legal and confidentiality concerns, infrastructure and quality assurance. KNBS organized technical working committees around themes such as disability and gender to harmonize data sources and engage stakeholders in understanding and reflecting the needs of marginalized groups in data processes. 

To create inclusive data systems, development practitioners should start by taking a critical look at the tools, processes and policies that influence how institutions carry out their work with data. Revisions to these should ensure people who face marginalization are included and protected.

5. Build more inclusive institutions. Foster inclusive data practices by reflecting on identity at the institutional level and by focusing on how intersectionality within an institution can shape research and data systems. Governmental statistics offices in Colombia and the UK are doing this by taking active steps to reduce inequalities and foster inclusion in the workforce. 

In the UK, the statistics authority’s inclusive data task force appraises intersectional issues across institutions in the national system. The agency collects public views and ways to improve inclusivity in data and statistics. Similarly, in Colombia, the National Statistics Office’s (DANE) develops explicit approaches that meet the needs of local groups. As a result, DANE has published guidelines and standards (available in full in Spanish with an executive summary in English) to mainstream diversity and inclusion across Colombia’s national statistical system.  

Each of these strategies prompts us to critically examine data practices and dig deeper. An intersectional lens asks important questions such as: Whose voices are hidden, excluded from or discriminated against in data? And how do we center those voices to address inequalities? Intersectionality demands that we reflect not just on data but on the people, systems and structures that create and shape the data we use. 

Only by addressing existing power imbalances can we build more inclusive institutions, make better decisions and create the data needed for policies and programs to improve lives.

Tichafara Chisaka is Program Manager of the Inclusive Data Charter. GPSDD Policy Officer Karen Bett contributed to this post. Find out more about the examples highlighted in this blog in the IDC's intersectionality products series