Lawler: Update on the Household “Conundrum”

"An actual decline in rents next year would be a reasonable base case"

This is from housing economist Tom Lawler:

The pandemic wreaked havoc on some of the government surveys designed to measure the number of and characteristics of US households (though there were measurement issues before the pandemic), making it difficult for analysts how much of the extraordinary strength in the housing market during 2021 and early 2022 was related to “demographics” as opposed to behavioral and preference changes.  (Here is the November 2021 piece on this issue).  Nevertheless, it may still be useful to see what these reports are showing.

In this regard, below are estimates of the number of US households by age group from the Current Population Survey Annual Social and Economic Supplement for March 2022 compared to updated estimates (incorporating the Census 2020 results) for March 2021 from the 2021 Income, Poverty, and Health Insurance Reports.

These estimates, if accurate, suggest that the number of US households increased by1.958 million from March 2021 to March 2022, with especially large % gains in the number of young adult householders. 

CPS/ASEC household estimates are “controlled” to independent population estimates, and these latest estimates are based on Census’ “Vintage 2021” population estimates, which included projections for 2022.  In any given year the CPS/ASEC household estimates are based on the latest population estimate “vintage” available, and Census does not typically adjust historical household estimates to reflect updated historical population estimates.

What is striking about the estimated household gain is that it is significantly above – in fact, about double -- what one would have projected if one had assumed that so-called “headship rates” in each age group (the number of householders in an age group divided by the total population of that age group) had remained the same over this period.  Moreover, headship rate increases were concentrated in the 15–34-year-old categories.

If in fact over the 12-month period ending March of this year the number of households increased by almost two million, then household growth significantly outpaced housing production, as total housing completions plus manufactured housing shipments totaled only about 1.44 million.

In terms of household growth, there are several reasons to expect that household growth has slowed sharply since the beginning of this year and will continue to be well below the latest reported rise.  First, concerns about the pandemic have eased considerably since last year, while at the same time many firms have shifted to a “return to work” policy.  Second, and very importantly, housing costs – both for buyers and for renters – are substantially higher than was the case last year.  And finally, of course, base population projections would suggest a significant slowdown in household growth if headship rates stopped increasing.

If headship rates from March 2022 to March 2023 were unchanged across age groups, then base population projections would suggest that household growth over than period would be around 975,000, or less than half that experienced from March 2021 to March 2022.  At the same time, it would appear as if housing production – completions plus manufactured housing shipments – will be somewhere in the range of 1.5 to 1.6 million, suggesting that housing production will significantly outpace household growth over that period.  Next year should see an especially large increase in the supply of new rental units, given the sharp increase in multifamily housing starts and the long lags between start and completion.

If in fact this dynamic shift between household growth and housing production is taking place, then it, combined with (1) the unprecedented increase in housing prices over a 2-year period from mid-2020 to mid-2022, and (2) the unprecedented surge in mortgage rates this year, would suggest that a non-trivial decline in home prices from the middle of this year to at least the middle of next year would be a logical “base case.”  This shift, combined with (1) the unprecedented increase in rents from late 2020 to the middle of this year and (2) the significant increase in rental units coming to market over the next year, suggest not only that rent growth should soon decelerate sharply (in fact, some indicators suggest that it already has), but that an actual decline in rents next year would be a reasonable base case.

Of course, there are several valid reasons to be leery of the CPS/ASEC household estimates.  For one, the estimates are based on a relatively small sample.  Second, the survey continued to be plagued by low response rates this year.  E.g., the response rate in March (the CPS/ASEC is conducted from February to April, with March the heaviest month) of 2022 was 72.2%, down from 76.2% in March 2021, 73.0% in March 2020, and 83.7% in March 2018. While low response rates do not necessarily bias the survey results, there is evidence to suggest that the characteristics of “non-responders” were different than those of responders. (See How Has the Pandemic Continued to Affect Survey Response? Using Administrative Data to Evaluate Nonresponse in the 2022 Current Population Survey Annual Social and Economic Supplement)

And finally, since 1990 the CPS/ASEC estimates of the number of young adult householders has been higher than, and grown faster than, the estimates from the decennial Census (though 2010; Census has not yet released the characteristics of US households from Census 2020).

Note: Previous articles on Household Formation:

McBride Sept 2021: Household Formation Drives Housing Demand

Lawler Nov 2021: The "Household Conundrum"

McBride Dec 2021: The Household Mystery

McBride May 2022: The Household Mystery: Part II