The conventional wisdom (based on Census estimates) seems to me that urban cores have lost population since COVID began, but are beginning to recover. But mid-decade Census estimates are often quite flawed. These estimates are basically just guesses based on complicated mathetmatical formulas, and often diverge a bit from end-of-decade Census counts.
Is there another way to judge the popularity of various places? Perhaps so. I just uncovered a database of real estate price trends from Redfin. Because housing supply is often slow to respond to demand trends, housing prices probably reflect changes in demand. What do they show?
First let’s look at the most expensive cities: San Francisco and New York City where I live now. If conventional wisdom is accurate, I would expect to see stagnant or declining housing prices in the city and some increase in suburbia. In Manhattan, the median sale price for condos and co-ops was actually lower in 2024 than it was in mid-2019, declining from $1.25 million in August 2019 to $1.05 million in August 2024.* Similarly, in the Bronx multifamily sale prices decreased slightly (though prices for single-family homes increased). By contrast, in suburban Westchester County, prices increased by about 30 percent (from just under 250k to 325k). Similarly, in Nassau County prices increased from 379k to 517k, an increase of well over one-third. So these prices suggest something like a classic suburban sprawl scenario: stagnant city prices, growing suburban prices.
In San Francisco, by contrast, property values declined everywhere. City prices declined from $1.2 million in August 2019 to just under $1 million today; in suburban Marin County, the median price declined from $633k to $583k. So sale price data certainly supports the narrative of flight from expensive cities.
What about places that are dense but not quite as expensive? But New York might not be typical. Let’s look at Boston, another dense, transit-oriented city. Condo prices in Boston have actually risen, from $625k in 2019 to 719k in 2014, about a 14 percent increase. In Middlesex County, prices rose from 485k to 615k- about a 27 percent increase. So prices rose more in the suburbs, but they rose in both city and suburb, indicating stable or growing demand for urban life. The same is true for other relatively dense cities like Philadelphia (19 percent increase city, but over 40 percent increases in suburban Montgomery and Chester Counties), Washington (6 percent city, 21 percent suburban Montgomery County, and Chicago (16 percent increase in city, over a 50 percent increase in suburban Lake County).
What about the Sunbelt? In my home town of Atlanta, city condo prices increased by over 30 percent, from 245k in 2019 to 321k in 2024. In suburban Cobb County, prices rose even faster, from 180k in 2019 to 300k in 2024. Similarly, Dallas prices increased from by about 25 percent (205k in 2019, 266k in 2024), while prices in Collin County to its north almost doubled (153k to 285k). So city prices grew at hefty rates but still lagged behind suburban growth rates.
What about prices for single family houses? Here, the picture is much more complex. In Philadelphia, price patterns for houses were pretty similar to those of condos. But in Boston, there was not much difference: city prices rose by almost 34 percent (from 649k to 870k), and Middlesex County prices by about 41 percent (600k to 850k). And Washington city prices grew by over 50 percent (from 777k to $1.3 million) more than in Montgomery County (about 30 percent) or Fairfax County (35 percent). In metro Atlanta, city and suburban prices both grew by about half (299k to 499k in the city, 287k to 465k in Cobb County).
So it seems to me that the real story might be that there is unmet demand for multifamily housing in suburbs, and unmet demand for houses in some cities.
