An Update on Housing Tenure Trends
The U.S. Census Bureau this week released its Q3 2020 Housing Vacancy Survey (HVS) estimates of households and housing tenure trends. Similar to our findings from the Q2 data release, the latest estimates suggest owner and renter rates that, while closer to pre-pandemic levels, show a significant break with expected trends. For example, the national homeownership rate in Q3 was estimated to be 67.2 percent on a seasonally-adjusted basis, a full percentage point down from the Q2 rate. However, this share is still nearly 2 percentage points above the 65.3 percent estimated in the first quarter of the year.
There are two likely drivers of this higher-than-expected rate of homeowning households: 1) changes in HVS data collection and response rates that overestimate owners and underestimate renters; and 2) market dynamics – including low mortgage interest rates, rebounding home sales, and spiking rates of rental evictions – that are having real impacts on the distribution of households by tenure. This blog explores both factors and discusses the potential impact of each on the reported HVS rates. How much of the apparent increase in homeownership is based on methodological versus actual trends, however, remains unknown.
The Role of Data Collection Procedures
As noted in our blog on the Q2 HVS data, a large part of the skepticism around the reported homeowner and rentership rates involves changes in how the survey was conducted amidst the pandemic. Specifically, in-person data collection, which was suspended entirely in the spring, was gradually reintroduced through the three months of the third quarter, rising from 39 percent of interviews in July to 100 percent in September. Overall, this means that 63 percent of the HVS survey interviews were conducted in person last quarter, and the remaining interviews were conducted by telephone.
HVS Data Collection Methods and Response Rates, Q3 2020
|Share in-person Interviews||39%||50%||100%||63%|
Source: U.S. Census Bureau, “Frequently asked questions: The impact of the coronavirus (COVID-19) pandemic on the Current Population Survey/Housing Vacancy Survey (CPS/HVS)”, Q3 2020 update.
With fewer in-person interviews, the response rate for households in the HVS sample was again below historical averages. Only 71 percent of those in the random address sample were reached last quarter, versus 83 percent for the same time frame in 2019. As the share of in-person interviews increased over the quarter, so to did the response rate, reaching 79 percent of the survey’s random address sample in September – though still below the 83 percent response rate documented in September 2019.
The precise impact of fewer in-person interviews and lower response rates on tenure estimates are not conclusively known, though we suspect these changes inflated the share of homeowners somewhat above what would otherwise have been observed. The Census Bureau notes that in-person interviews are more likely to reach renter households than those conducted by telephone. Non-response rates are also higher for renters than for owners, meaning that there were fewer renter households identified and surveyed in Q3 relative to prior quarters. As such the number and share of renters were likely underestimated.
Census’ own comparisons of in-person data collection responses bear this out; in September 2020 alone, when 100 percent of interviews were able to be conducted in-person, the national homeownership rate was estimated at 66.1 percent, or 1.3 percentage points below the estimate for the entire third quarter. In geographies where in-person interviews were allowed through all three months, the national homeownership rate was an even lower at 65.5 percent. The list of such geographies skews towards the Northeast, where lower virus rates over the summer meant that 95 percent of all interviews in the quarter could be conducted in-person, versus 45 percent in the South and 61 percent in the West. The homeownership rate in the Northeast also tends to be lower than that of other regions of the U.S., which may explain part of this lower estimate relative to the national quarterly rate.
The Role of Market Dynamics
While changes in the HVS data collection and response rates are likely contributing to the higher national homeownership rate in Q3, some indicators of recent housing market activity may also suggest potentially higher shares of homeowning households.
Since the early days of the pandemic, mortgage interested rates have been holding steady at record-low levels, most recently coming in at 2.8 percent for a 30-year fixed rate loan, according to Freddie Mac. Homebuyers that can take advantage of such cheap borrowing costs appear to be doing so, as existing home sales, which initially fell in the spring, hit a 14-year high last month, per the National Association of Realtors. Demand for more space as many households have moved to remote work and school, as well as increased locational flexibility, are also driving increases in home purchases.
However, the share of homebuyers who are new homeowners has declined to 31 percent last month. This means more than two-thirds of recent home sales have been to existing owners, who do not add to the national homeownership rate. With home prices also rising dramatically amid low inventory, affordability is likely to present a challenge to many first-time buyers.
Even if the number of homeowners is not growing substantially, other factors may be elevating their national share. In particular, a decline in the number of renter households would reduce the total household count in the denominator of the homeownership rate, making the share of the remaining households that own appear larger. Indeed, there appears to be some evidence of this, as the surge in rental evictions that occurred this summer when most state and local moratoria ended may have reduced the total number of renter households. Meanwhile, the CDC-ordered national eviction ban, which was announced in September, does not appear to have substantially reduced the pace of evictions in many parts of the country. Declines in renter households may also come from some voluntarily giving up their unit and moving in with family or friends to save money or avoid an anticipated eviction.
How much either of these trends are contributing to estimated higher homeownership rates – and whether the effects will last beyond the pandemic and recession – remains to be seen. There are still plenty of reasons to take reported tenure trends with a grain of salt, for now.
Stay tuned to our blog as we continue to track and evaluate pandemic-related housing market outcomes. You can also check Enterprise’s Resilient Futures resource page for more information on housing and policy responses to the pandemic, and subscribe to our daily and bi-weekly policy newsletters for more information on Enterprise’s federal, state, and local policy work.