Census Bureau Releases Coverage Estimates for Housing Units from the 2020 Post-Enumeration Survey: Were Connecticut’s Housing Units Accurately Counted?

On August 16th, the U.S. Census Bureau released coverage estimates for housing units from the 2020 Post-Enumeration Survey (PES). These results provide insight into the accuracy of the 2020 Census counts of housing units (not to be confused with counts of people). The coverage estimates are provided nationally, by region and state, and by type of housing unit (e.g., single unit versus multiunit and owner- versus renter-occupied) and characteristics of the householder (e.g., race and ethnicity).

The Census Bureau estimates the net coverage error of the 2020 Census housing unit counts by comparing them to independent counts generated as part of the PES. The Census Bureau’s definition of housing units includes occupied and vacant units, single and multiunit homes, apartments, and mobile homes or trailers, but excludes group living quarters such as college dormitories and correctional facilities.

 

Key Findings at The National Level

Nationally, the estimated net coverage error for housing units was 0.04% (or 60,000 housing units), which was not statistically significantly different from zero. This suggests that the 2020 Census produced a reasonably accurate count of the total number of housing units in the nation.

However, there were some significant national net coverage errors for particular kinds of housing units and householders:

  • Occupied housing units were slightly overcounted by 0.33% and vacant units were undercounted by 2.6%.

  • Rented units were overcounted by 0.85% whereas owner-occupied units were not statistically significantly over- or undercounted.

  • Small multiunit housing units (those in a two- to nine-unit building) were overcounted by 5.1% and trailers and mobile homes were undercounted by 4.3%, whereas single-unit and large multiunit housing units were not significantly over- or undercounted.

  • Housing units with a householder who was Black or African American (0.87%), Asian (1.37%), Native Hawaiian or Other Pacific Islander (2.64%), or Some Other Race (0.58%) were overcounted, whereas units with a householder who was White, American Indian or Alaska Native, or Hispanic or Latino were not significantly over- or undercounted.

 

key Findings For Connecticut and Other States

The net coverage error for all housing units in Connecticut was not statistically significantly different from zero, suggesting that the count in Connecticut was reasonably accurate.

In contrast, the following states had a significant overcount:

  • Alabama (2.7%)

  • Massachusetts (1.4%)

  • New Jersey (2.5%)

  • New York (3.7%)

  • Ohio (1.2%)

  • Rhode Island (1.6%)

  • Utah (0.8%)

Two states had a significant undercount:

  • South Carolina (2.5%)

  • Vermont (4.1%)

 

For More Information

See the Census Bureau’s full report for more details, including components of coverage errors (that is, the estimated number of units erroneously counted by the Census due to duplication and other reasons, and the number of units erroneously omitted by the Census). For in-depth information on how the Census Bureau generated the 2020 Census housing unit coverage estimates, see the Methodology Report.

For information about 2020 Census population coverage estimates, see the Census Bureau’s report on the 2020 Post-Enumeration Survey coverage estimates for people by state, or check out our prior blog post for an overview of those results.

You can keep up with us by subscribing to the CTData newsletter and following us on Twitter, Facebook, and LinkedIn.

Census & ACS, HousingCynthia Willnercensus