Racial disparities in the housing market have been stark for nearly a century. Because of the pervasive influence of structural racism, people of color have experienced the greatest barriers to accessing and maintaining homeownership. Moreover, these disparities are expected to widen, given that the COVID-19 pandemic has had a disproportionate impact on households of color.

Amid increased attention on these racial disparities, policymakers, local stakeholders, and financial institutions are designing and implementing solutions to expand homeownership and wealth-building opportunities for households of color.

To measure the success of these efforts, a timely estimate of the racial homeownership rate is critical. It provides a common understanding of the current reality and can help housing stakeholders set a common goal for defining and measuring progress.

Though this may sound simple, variations in data sources make it complicated. Different data sources have different ways of describing the gap, so experts are not always working from the same benchmarks and numbers. To reduce the confusion, it’s vital to be clear about what data source is being referenced, understand how data sources differ, and use the correct terminology.

Why different data sources provide different numbers

Three commonly cited sources—the Current Population Survey/Housing Vacancies and Homeownership Survey (CPS/HVS), the American Community Survey (ACS), and the decennial census—measure the racial homeownership gap and give different results. Though the datasets all come from the US Census Bureau, the homeownership rates show conflict because they vary in frequency and use different sampling techniques, definitions, and weighting methods.

The CPS/HVS

The CPS/HVS provides quarterly information for the United States, census regions, states, and the 75 largest metropolitan statistical areas (by population). It employs the Current Population Survey sample, which has about 72,000 housing units. 

One of the CPS/HVS’s greatest strengths is timeliness: the US Census Bureau releases estimates within a month or two of the end of every quarter. This enables us to quickly understand and respond to changing market conditions, especially when there is (or has been) substantial market disruption. 

But the quick data collection and release come at the cost of accuracy and detail. The CPS/HVS has a small sample size, which produces large fluctuations, especially for small subgroups. For example, the Black homeownership rate—which is based on a smaller sample than the sample of white households—is volatile.

The small sample size also means it is not possible to analyze homeownership in smaller geographic areas (e.g., states, counties, zip codes, and census tracts) or by housing characteristics (e.g., number of rooms or units in structure). This weakness of small sample size was pronounced in the second and third quarters of 2020, when response rates were depressed because of the COVID-19 pandemic (PDF), especially among renters, pushing homeownership rates improbably high.

Thus, the homeownership rates based on the CPS/HVS are best for someone who wants to know the most up-to-date trends at the national or regional level.

The Decennial Census

The decennial census counts 100 percent of the US population and housing, so it is the most complete and reliable source. It is conducted every 10 years, and as of today, the most up-to-date estimates are for 2010. It usually takes two years for the Census Bureau to compile and release the results. The 2020 Decennial Census data have only been partially released, and tenure information is expected to be released in mid-2022.  

Beginning in 2010, the decennial census only used a “short-form” questionnaire that includes limited household and housing characteristics. But this makes it impossible to calculate racial homeownership rates by income, education, and employment—trends that research seeks to understand. 

The decennial census is best for someone who wants the most accurate picture of racial homeownership rates or homeownership rates in smaller geographies, but because it’s conducted so infrequently, it does not include the most up-to-date trends.

The ACS

The annual ACS is positioned between the other two. Its sample includes about 3.5 million housing units, and thus its one- or five-year estimates provide a sufficient sample size to produce estimates for smaller population subgroups or geographies.

For example, using the ACS, we can find the homeownership rates of Asian and Hispanic (PDF) households by country of origin. But there is still a substantial lag in the release of the ACS estimates (usually about two years), so the latest available data are for 2019. And recently, the Census Bureau announced that it will not release its standard one-year ACS estimates because of the pandemic’s effects on data collection.

The Pros and Cons of the Most Commonly Cited Datasets that Measure Racial Homeownership Gaps

The ACS is best for someone seeking to understand homeownership rates by different population subgroups or across smaller geographies, but it is less accurate than the decennial census and less timely than the CPS/HVS.

Don’t confuse percentages with percentage points

When talking about the racial homeownership gap, a common misstep is to conflate percentage-point differences with percentage differences.

Using 2019 ACS data, we calculate that the white homeownership rate is 30 percentage points higher than the Black homeownership rate (72.2%-42.3%), and white households are 71 percent ((72.2%-42.3%)/42.3%) more likely to own their homes than Black households. In other words, the white homeownership rate is 71 percent higher than the Black homeownership rate.

It is important to use the correct terminology, because when we don’t, the size of the racial homeownership gap is unclear. Without that clarity, we can’t begin closing it.

Racial Homeownership Gap: Percent vs. Percentage Points

Source: 2019 ACS

Users may want to consult different datasets, depending on their goals

The homeownership rate differences across datasets may confuse housing market analysts, community stakeholders, and policymakers who want to assess homeownership trends or the effectiveness of policy interventions aimed at closing the racial homeownership gaps. Which data source should we use? It depends.

If stakeholders need the timeliest information, it’s best to use CPS/HVS data. But they should be aware that the numbers may not be highly accurate, be cautious about presenting them, and not be overly heartened or discouraged by fluctuations.  

If they want to know what is going on at the state or metropolitan level, they should use ACS data. For smaller geographies, it is safer to aggregate multiple years of ACS data to get accurate homeownership rates for the population they are interested in.  

Once the 2020 Decennial Census data come out next year, we will get the most accurate information on the racial homeownership rate. But to track trends, we will need to continue relying on CPS/HVS data and ACS data. At Urban, the Housing Finance Policy Center generally prefers to use the ACS. 
Each survey has pros and cons, and users should consider these and find the data source that best serves their purpose, refrain from mixing different sources, and indicate which source they are using. This will help ensure accuracy, which can lead to stronger analyses and more effective policy solutions.

The Urban Institute has the evidence to show what it will take to create a society where everyone has a fair shot at achieving their vision of success. 

Show your support for research and data that ignite change.  

DONATE