Elimination of health disparities requires recognition of the subjectivity of data and the power of data to dictate and reinforce narratives, accompanied by intentional reform of data practices.
Data are a social construct, made for and by people, and the ways that data are collected and used in public health practice is an act of power with profound consequences. Recently, with COVID-19 and other simultaneously occurring epidemics and pandemics, the need for high quality public health data has entered the public’s consciousness and acquired renewed urgency. We saw how damaging the dearth of data could be to communities of color in the early days of COVID-19. In 2020, with few states sharing—or even collecting—data based on race and other important identity factors, fewer resources were allocated where they were most needed. The result was much higher rates of death for Black and Latino populations. This is compounded when valuable data are withheld or hidden, hastening the spread of COVID-19 and other infectious diseases among the most vulnerable and marginalized like people who are incarcerated or in nursing homes.
Even the best “data-driven” decisions can perpetuate harm if equity, specifically anti-racism and intersectionality, has not been considered in all aspects of the data collection, analysis, interpretation, and dissemination. In our recent paper, we describe the process the NYC Department of Health and Mental Hygiene undertook to create Data for Equity, a framework for applying an intersectional, anti-racist equity lens in the agency’s data practices. This included identifying resource needs around social justice and racial and gender equity-informed data analysis and communications, identifying initial guidance for staff working with data, and developing recommendations for leadership on the types of tools and trainings that should be offered.
What we learned:
- To be a truly anti-racist and intersectional public health agency and effectively eliminate health disparities requires recognition of the subjectivity of data and the power of data to dictate and reinforce narratives, accompanied by intentional reform of data practices.
- Addressing systematic oppression in how data are collected, analyzed, and shared must be an explicit part of intersectional and anti-racist public health practice.
- Equity in data is an essential foundation of the national conversation and of initiatives towards data modernization.
- While the public health community has made strides in foregrounding racial equity in public health rhetoric, few resources that address how to incorporate equity principles in the collection, analysis, and reporting of data that influence public health decisions are available. Other health departments can use the model we created to design similar institutional reform initiatives.
- In addition to staff training and reforms to data collection and analysis, high-level leadership commitment is essential to ensure that equity is embedded into actual planning and execution of analyses, report compiling, and other uses of data whether it be internal dissemination, program or resource allocation decisions, or publications. This embedded planning and execution must be ongoing and intentional.
Recommended principles for working with data in an equitable manner:
- Reframe ‘best practices’ as ‘better practices.’ Practices shift and change over time as we learn and unlearn. Additionally, people and communities change and language shifts to include new ideas and identities.
- Embed data reform within institutional reform and personal reform. Each person’s own lived experience and position changes how they name and interpret oppression. The process of acknowledging our biases and need for growth is ongoing at individual and institutional levels.
- Present all data in their rightful historical, political, and experiential contexts. Data are always the result of various historical, political, and social contexts. For example, race/ethnicity categories not only add a sociopolitical lens to the data being collected but also reflect histories of privilege, racist social structures, and racist policies.
- Collect and analyze data intersectionally. Truly inclusive data are intersectional. The impact of racism and social injustice on people of color cannot be fully understood unless we understand the specific impact on people at the intersections of oppressions, including women of color; people with disabilities; transgender, gender nonconforming, and non-binary persons; and many more who have been historically and systemically marginalized.
- Value people’s lived experience and treat it as data rather than anecdote. Lived experience is required for understanding the meaning of data. Identities captured in demographic data are social constructs that reflect complex and ever evolving hierarchical power structures. Data collection instruments with poorly or broadly defined demographic categories often do not match people’s lived experience, making it hard to accurately identify inequities with quantitative data.
- Be transparent and consultative in methods across the data lifecycle. Methodologic decisions (e.g., reliability estimates) are largely inaccessible to people whose data are being analyzed. For example, decisions about how to aggregate small groups within a dataset happen long after the data have been collected. This ends up creating categories that do not reflect lived experience accurately.
- Always interrogate who is missing from the data, why they are missing, and what that means for the analysis and dissemination. When key populations are left out of data collection, it is rare that their absence is noted except for a cursory line in a limitations section. These absences have a compounding and inequitable effect.
We are encouraged by the increasing interest in equitable data work in public health over the last two years and hopeful that this will lead to continued integration of data practices in overall health equity work. We also encourage all public health practitioners – from people who knock doors and collect survey data to people in leadership positions who use data to advocate for desperately needed structural change — to make an ongoing commitment to transforming themselves, the ways they work with data, and the structures within which they work.
Elimination of health disparities requires recognition of the subjectivity of data and the power of data to dictate and reinforce narratives, accompanied by intentional reform of data practices.
Read Our Article in the Journal of Public Health Management and Practice:
