Population and Housing in 2021

The 2021 American Community Survey confirms that major population shifts took place due to the pandemic. But those shifts aren’t necessarily reflected by declines in housing prices in cities and regions that lost population. Indeed, prices rose almost everywhere, and usually faster than incomes.

Americans are moving out of big cities to smaller towns and the suburbs, but so far this hasn’t made big-city housing more affordable.

Numerically, the big loser between 2019 and 2021 was California, which saw the net departure of almost 275,000 people. That was just 0.7 percent of the state’s population, but the only state that lost a greater percentage was Mississippi. Other states that lost residents were Arizona, Louisiana, and West Virginia, plus the District of Columbia lost more than 25,000 people, or more than 5 percent of its population.

Despite its population decline, California housing prices grew by 14 percent, compared with a 6.6 percent increase in median family incomes. Prices also grew in every other state by more than incomes. Only in DC did prices, which grew 3.6 percent, not keep ahead of incomes, which grew 4.5 percent.

The result, of course, is that housing became less affordable almost everywhere, even many places that lost population. Home prices grew slower than incomes in Puerto Rico and a few rust-belt regions, but these were rare exceptions to the rule.

California’s biggest population losses came in Los Angeles County, which lost more than 200,000 people (130,000 from the city of Los Angeles). Yet home prices increased by 11.8 percent against a 6.9 percent increase in incomes. San Francisco lost 66,000 people, Santa Clara County (San Jose) 42,000, and San Diego County 52,000, yet all three saw home prices grow faster than incomes — twice as fast, in San Diego’s case.

Population losses weren’t confined to California cities. San Antonio lost 95,000; Dallas 55,000; Boston 40,000; Phoenix 56,000; Detroit 37,000; Houston 30,000; Miami 28,000; Memphis 23,000; Milwaukee 21,000; Seattle 20,000; Austin and Denver 15,000; Portland 11,000. (Note these are all changes in city populations, not counties or urban areas.) This shows that the trend is that people have been leaving big cities regardless of state, but in states such as Texas (whose population grew by well over half a million people), they have been moving to suburbs and smaller cities while many in California simply moved out of state.

Regardless of population declines, housing prices nonetheless grew faster than incomes in all of these cities. Curiously, places where home prices did grow slower than incomes often didn’t see population declines. For example, Lake Charles, Louisiana, saw incomes grow by 25 percent and its population grew by 3.5 percent yet home prices grew by only 15 percent.

These numbers are based on tables B01003 (population), B19133 (median family incomes), and B25077 (median home values). I’ve posted a spreadsheet with all of data from these three tables for both 2019 and 2021 for about 1,900 geographic areas including states, counties, cities, urban areas, and the nation as a whole. For some geographic areas, the Census Bureau reported data for 2019 but not 2021 or for 2021 but not 2019. In general, I’ve retained the 2021 numbers but deleted the 2019 numbers when there are no corresponding 2021 numbers.

In the spreadsheet, columns B, C, and D show 2021 median home values, median family incomes, and value-to-income ratios; columns E, F, and G show the same data for 2019; and columns H, I, and J show percent changes in each of these values. Column K shows the change in value-to-income ratio as a whole ratio rather than a percent so that places that became more affordable could be shown in red.

Population data are in columns N (for 2021) and O (for 2019), with numeric change in P and percentage change in Q. For sorting purposes, column R shows the state (for counties and cities) and column S shows the type of geographic area (nation, state, county, city, urban area). The latter are preceded by a number (1 for nation through 5 for urban area) so I could sort in the same order I usually use: nation, state, county, city, and urban area.