Wealth is critical for financial well-being and mobility. It enables families to manage day-to-day finances, remain resilient in the face of economic shocks, and pursue opportunities for upward mobility—affording tuition for education, launching a business, buying a home, or retiring with enough savings. Wealth provides a critical path to the middle class.

But in the US, wealth isn’t equally distributed. Historic and current policies have prevented Black and other people of color from accessing the same wealth-building opportunities white families have had. In 2019, the typical white family had eight times the wealth of the typical Black family and five times the wealth of the typical Latinx family.

Policymakers, program managers, community leaders, and funders looking to address the racial wealth gap and improve economic outcomes need good, local data on how households are faring by race and ethnicity so they can target interventions to communities. However, publicly available wealth data that are disaggregated by race and ethnicity have several key limitations. Improving data on racial wealth and making it accessible are critical steps to understanding and addressing the wealth gap.

Limitations to public racial wealth data

Existing data disaggregated by race and ethnicity are limited by slow timelines, lack of granularity, and inaccuracy.

Foundational sources such as the American Community Survey (ACS) only publish data annually and at a lag, limiting the ability to conduct real-time, high-frequency analysis. Though some public sources, such as the ACS, Survey of Consumer Finances (SCF), and Survey of Income and Program Participation (SIPP) disaggregate by race and ethnicity, this often comes with caveats, such as the use of broad racial categories that miss important granularity within Black people (for example, ethnicity) and other communities, or time and geographic constraints.

Many public data sources like the ACS also have limited sample sizes, meaning that even if the data are disaggregated by race and ethnicity, economic outcomes for certain subgroups and smaller or rural geographies cannot be reliably estimated. And public surveys, which are common sources of wealth data, often have high nonresponse rates among certain communities (for example, underrepresentation of young Black men in SIPP data), as well as errors in sampling populations or in measurement that hinder data accuracy, particularly for smaller and more rural (PDF) communities. Without combing through different datasets, a typical data user may not even be aware of these nuances.

Additional tools and resources can improve racial wealth data

Private-sector data assets could help fill wealth equity knowledge gaps. Urban’s landscape scan of private-sector data found that private-sector data on topics such as home assets and credit health can provide real-time, frequent insights that can fill gaps in publicly available alternatives. Stakeholder interviews revealed that private administrative data, such as credit card transactions, can offer more accurate information about net worth and assets than self-reported answers on a questionnaire like the SCF.

Still, private-sector data can have similar and more severe limitations in quality, granularity, and utility to racial equity analysis. We found most private-sector companies either don’t collect race and ethnicity identifiers or don’t share that information publicly, whether for legal, compliance, or customer privacy reasons.

A couple analytical techniques that take advantage of public-private partnerships hold promise in adding race and ethnicity identifiers to private data—with some caveats.

• Imputation allows data users to fill in missing data using statistical methods. Algorithms such as Bayesian Improved Surname Geocoding (BISG) developed by the RAND Corporation can impute race based on one’s name and address. BISG is strongly predictive of self-reported race for Black populations, though its estimates are uncertain at the individual level and perform poorly for other racial groups. Urban also used multiple imputation techniques to affix racial identifiers to credit bureau data, while implementing “checkpoints” to assess their algorithm for racial bias and inaccuracy at each step of the process. Urban’s Financial Health and Wealth Dashboard similarly used a machine learning model–based imputation approach to estimate emergency savings and net worth. These are among the data offered at the Public Use Microdata Area and city levels, allowing leaders from government and philanthropy and policy practitioners to influence financial well-being at the local level.
• Linking rich private data sources that don’t have racial identifiers with public sources that do is another fruitful way for the public and private sectors to collaborate. The JP Morgan Chase Institute linked Chase banking data with public voting records data in states that contained Chase branches and collected self-reported voter race information. Their analysis answered questions about financial earnings, assets, and behaviors that no public data source could measure as accurately. Similarly, the Urban Institute’s Financial Well-Being Data Hub is creating a new data asset that will merge public and proprietary consumer data.

However, these are not perfect solutions because they raise ethical concerns, such as the risk of reidentifying individuals in the data, lack of informed consent as to how data will be used, and violation of personhood of those represented in the data. With thoughtful treatment of these risks, these methodologies are worth more exploration.

Finally, a crucial part of improving data on racial wealth includes gathering data and insights from communities of color themselves in a non-extractive way, particularly when the best alternatives fail to capture local, granular context. Urban’s Community Engaged Methods Guidebook offers practical strategies to improve community participation in existing and new surveys, determine what data should be collected, and interpret findings.

Strengthening racial wealth data can reduce inequities

Improving data collection and analysis will help decisionmakers answer questions such as: What are typical asset-debt ratios, and how much does this vary within Black communities and other communities? How are asset and debt holdings related to other outcomes, such as health inequities? What systemic interventions—such as baby bonds and reparations—can help address racial wealth disparities?

Having the data necessary answer to these questions is the first step to advancing evidence-based policies and practices that root out systemic disparities and narrow the wealth gap.