Access to the vital community conditions for health and well-being differs across place and population. Disaggregating data—by geography (e.g., county, neighborhood), demographic group (e.g., age, gender, race, and ethnicity), or by other subpopulations—helps identify underlying patterns of variation.
Disaggregating data refers to the process of breaking down aggregated data into smaller informational units, in order to examine a characteristic or dimension.
For example, we can “break down,” “disaggregate,” or “stratify” health outcome or behavior data (such as cancer incidence, or receiving a flu shot) by race and ethnicity to gain clarity on existing disparities and examine issues of health access and opportunity in communities of color. Disaggregating data provides an additional layer of information that helps to focus health improvement efforts that minimize disparities, and leverage investments to advance equity.
Explore an Example
Disaggregating poverty data allows us to see areas of concentrated poverty and the disproportionate impacts of poverty on certain priority populations. Use the interactive visualization below to examine disaggregated poverty data in your County.
The map displays the percentage of the population for whom poverty is determined who are below the Federal Poverty Level for each Census Tract within a County; the bar graph shows the percentage of the population for whom poverty is determined who are below the FPL by race (e.g., Asian and Asian Americans, Multiracial), ethnicity (e.g., Hispanic and Latinx), and age group (e.g., Older Adults).
In Boone County, Missouri (the default geographic selection) we can see that poverty is concentrated in certain parts of the county’s urban core–the City of Columbia (IP3’s hometown!). We also see that higher proportions of Black and African American people and Native Hawaiian and Pacific Islander people experience poverty than other groups. These insights can help guide community stewards when making decisions about how and where to best invest time, resources, and programming to advance equitable well-being.
For example, geographic insights about poverty can inform decisions about siting community services, or prioritizing transit and affordable housing investments in low-income neighborhoods. Information about groups that are disproportionately affected can inform program planning and evaluation activities, community engagement priorities, and outreach and communication efforts. By using disaggregated data, community stewards can make better decisions, leading to effective, efficient, and sustainable change.
Characteristics of Disaggregated Data
In order to explore disaggregated data for geographies or subpopulations of interest, we must collect the appropriate data. To see data disaggregated by geography, geographic identifiers must be present in the dataset. To see data disaggregated by demographic group or other subpopulations, relevant demographic variables must be present in the dataset.
Geographic Identifiers
Disaggregation by geography is possible with geospatial data. Geospatial data refer to location-based data, which are often displayed on a map but can be used in many ways and facilitate geographic and location-based analyses. Geospatial data require geographic identifiers, like addresses or latitude/longitude coordinates, or geographic entities, such as States, Counties, Census Places (e.g., Cities), Census Tracts, and Census Block Groups. In the embedded visualization above, values for individual Census Tracts within a County are displayed on the map.
Demographic Variables
Disaggregation by demographic group or sub-population is possible when relevant demographic variables are available. Demographic variables are variables or factors such as age, gender, or race, used in the study of human populations. In the embedded visualization above, values for various demographic groups are displayed on the graph. Demographic variables are often 1) self-reported from individuals in a population of study or 2) derived from administrative records.
Self-reported information - Data collection methods like surveys, focus groups, and interviews typically collect information from individuals within a study population through questionnaires which may contain a set of demographic questions that may ask respondents about race, ethnicity, gender, age, and more. These elements can then be used to support data analysis for specific subpopulations. Note that definitions of demographic classifications may not be consistent from source to source due to differences in data collection.
Derived from records - Certain datasets, like vital records, patient health records, and administrative records of social service agencies may use methods other than self-report to derive demographic variables. For example, age may be calculated from birthdate data. Sometimes demographic variables are derived from observations as well, for example preferred language spoken.
Considerations
Despite the benefits of disaggregating data in advancing equitable well-being, it also has some risks. Disaggregated data can reinforce stereotypes, contribute to stigma, and perpetuate harmful narratives about specific groups and communities. Care and consideration are needed when working with disaggregated data, especially in sensitive use cases. Take the example of people who are involved with the justice system. Devoid of the context of systemic racism, disaggregating arrest rates or counts by race can center narratives that are harmful to the communities experiencing the inequities. For this reason, it is recommended that some datasets are not disaggregated.
On the other hand, presenting disaggregated elements may obscure issues of intersectionality—the acknowledgment that multiple aspects of a person's identity, such as race, gender, and socioeconomic status, intersect and influence their experiences of health and well-being. For example, income data disaggregated by romantic or sexual orientation does not account for intersecting issues of gender and race within the broader LGBTQ+ population, which is essential for a comprehensive understanding of income disparities. Weaving available local data together with story and qualitative data adds nuance and context.
Disaggregated data can also create privacy concerns when populations are small in size. When populations are small there is a risk that disaggregated data may identify a specific individual or small group of people. Data suppression is an important data privacy strategy to protect individuals from being personally identified and prevent discriminatory use of data and other data misuse. Data sources will often adopt data suppression standards wherein values are not made available when population counts are below a threshold. For example, the Centers for Disease Control and Prevention suppresses rates and counts if there are fewer than 16 total cases or deaths in a population. For this reason, certain demographic breakouts and/or geographies may not be available from a data source. In addition to privacy, data suppression helps ensure data are reliable and stable and discourages misinterpretation or misuse of data.