From housing economist Tom Lawler: ACS-Based Household Growth Slowed Last Year: Homeownership Rate Estimate Up, But Barely

The Census Bureau released the results of its 2016 American Community Survey, which provides a wide range of estimated statistics about people and housing. On the housing front, the ACS-based estimate of the number of occupied housing units (or households) for 2016 (yearly average) was 118,860,065, up just 651,815 from the estimate from the 2015 ACS. The ACS-based homeownership rate for 2016 was 63.1%, up just a tad from the 63.0% in 2015.

The growth rate of nonfamily households outpaced the growth in family households last year, with the fastest growth coming in nonfamily households with two or more people. This growth in part reflected a sharp jump in roomers and boarders.

ACS-Based Household Estimates
20162015Change% Change

In terms of the composition of people in “family” households, last year there were relatively big gains in (1) the number of “adult” (18+ years old) children living with their parent(s); (2) the number of “other relatives;” and (3) the number of “nonrelatives” excluding the householder’s unmarried partner.

Number of People in Family Households
by Relationship to Householder, ACS
20162015Change% Change
  Under 1864,448,63864,499,542-50,904-0.08%
Other Relatives23,949,69023,665,436284,2541.20%
  Unmarried Partner3,092,3573,205,347-112,990-3.53%
  Other Nonrelatives4,432,2664,332,33599,9312.31%

Before going further, I do need to qualify the table above. The 2015 ACS results did not incorporate the late 2016 downward revisions in population estimates, but the 2016 ACS results do reflect these revisions. As a result, the 2015 ACS results overstate the total resident US population by about 522 thousand. The 2015 ACS estimates for the number of households, however, would not have been impacted by the downward population revisions (if they had been known at the time), because ACS household estimates are (effectively) controlled to independent housing unit estimates. If the 2015 ACS results were adjusted to reflect the updated lower population estimates for that year, the result would be that both the average household size and the average family size (as estimated by the ACS) would be lowered for 2015. As a result, both the average household size and the average family size adjusted to reflect population revisions for 2015 increased last year.

For “young” adults (18-34 years old), the % of young adults living with parents increased to 34.25% in 2016 from 34.11% in 2015, while the % of young adults living with other relatives rose to 13.30% in 2016 from 13.20% in 2015.The % of young adults who were either (1) a householder, (2) the spouse of a household, or (3) the unmarried partner of a householder declined to 32.95% in 2016 from 33.46% in 2015.

As readers know, there are numerous and conflicting estimates of the number of and the growth of US households based on various surveys, and this “household conundrum,” while identified and for a brief bit “prioritized” by Census earlier this decade, has not gone away. However, the “latest” data from the CPS/ASEC, the CPS/HVS, and now the ACS all suggest that household growth as decelerated, for reasons that are not crystal clear but which cannot be explained by “cyclical” forces.

I’ll have more on the ACS data later this month.