Homeownership During the Third Quarter: Technical Issues

Facebooktwitterpinterestlinkedinmail

The Census Bureau’s Housing Vacancy Survey (CPS/HVS) reported the U.S. homeownership rate declined to 67.4% in the third quarter of 2020, from 67.9% in the prior quarter. The elevated homeownership rates are line with increases in both existing and new home sales in the second quarter. The national rental vacancy rate increased, while the homeowner vacancy rate remained unchanged.

However, technical issues involved with data collection changes limit useful comparisons of the second and third quarter data with the prior data series. This may change in the fourth quarter. We have noted the second and third quarter homeownership rate data with a dashed line below to denote these technical changes and encourage readers to consider these data points separately from the existing data series.

 

The data collection changes during the COVID-19 pandemic period likely affected the estimates of homeownership rates, vacancy rates, and the components of housing inventory, with the exception of total housing inventory unaffected. Compared to the second quarter when all in-person interviews were replaced by telephone contact attempts, in-person interviews were gradually resumed in the third quarter of 2020. Only 39% of in-person cases were allowed in July, but 50% in August and 100% in September. The overall response rate for the third quarter of 2020 was 71%, 4 percentage points higher than the 67% response rate for the second quarter 2020. However, it was much lower than the 83% response rate for the third quarter 2019.

The current CPS/HVS weighting and methodology, however, cannot compensate the differences in response rates in the subgroups, such as rental versus homeowner units, and vacant versus occupied units. If the suspension of in-person interviews increased the challenges of collecting data for rental units more than homeowner units, response rates for rental units would decline further than homeowner units. Under this circumstance, the effects on CPS/HVS estimates would be a lower estimate of the number rental households, a higher estimate of the number of homeowner households, and a higher estimate of homeownership rate.

Similarly, if the suspension of in-person interviews disproportionately affected the response rates of the HVS-eligible vacant units compared to occupied units, the vacancy rate estimate would be lower. The homeownership rate by mode in 2019, before the suspension of in-person interviews, indicates that a lower share of homeownership and a higher share of rentership under in-person interviews than other survey modes across all regions.

All things considered, the CPS/HVS data collection issues make comparing second and third quarter estimates to previous quarters challenging. Furthermore, it is impossible to tell whether the changes in homeownership rates, vacancy rates and changes in housing inventory components are due to the effects of the COVID-19 pandemic or the effects of data collection issues.

Looking forward, the Census Bureau hopes, but is not certain that in-person data collection will be allowed nationwide in the fourth quarter.



Tags:

1 reply

  1. Weekly initial jobless claims declined slightly last week, and continuing claims continued a downward trend. The modest decreases in initial and continuing claims show that labor markets continue to recover from the COVID-19 pandemic but the pace is slow.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: