Documenting Recent Convergence in Home Prices Across Local U.S Real Estate Markets
Zillow uses its Big Data on local real estate markets to create interesting data for the public to play with. It creates a Zillow price index that it claims represents an “apples to apples” comparison of quality adjusted homes. This price index allows researchers to study home price variation both within and across markets.
In this Substack, I first document that Los Angeles and Chicago’s real estate markets are not converging in value! I then study all of the U.S urban real estate markets and document recent convergence in home prices.
Consider this graph I made using the nominal Zillow price data for Los Angeles County and Cook County (Chicago).
In the year 2000, home prices in Chicago and Los Angeles were very similar at roughly $200,000. If you calculate the vertical difference over the last 25 years, you will see that a buyer of a home in the year 2000 in Los Angeles earned a much greater return on holding that asset for 25 years than a year 2000 buyer in Chicago. The huge differential in home prices in the year 2024 between my County (Los Angeles) and Chicago highlights that home prices vary greatly across space. A good economist asks; “Why?” . How much of this differential is due to amenity differences? Local public policies and tax differences? How much is due to differences in economic opportunities in the two locations?
Documenting National Convergence Across Local Real Estate Markets
I now use data for all 1,054 counties (including LA and Chicago) and I generate the following unintuitive picture.
Similar to the graph presented above, this picture presents facts from the year 2000 to 2024. While the U.S features more than 3000 counties, Zillow produces county level home price data for 1,054 counties for each of the 25 years from 2000 to 2024. These are the major counties and this includes New York City, Los Angeles, Chicago and many others.
What is a Coefficient of Variation?
For each year from 2000 to 2024, I take the county level data on home price and I construct the standard deviation of home prices. I calculate this by weighting by the county’s population in the year 2000 so I am placing more weight on observations in counties where more people live. Call this variable “Standard Deviation” . I also calculate the weighted average of home prices in each year. Call this variable “Mean”.
The Coefficient of Variation = Standard Deviation/Mean
I calculate this for 25 years and graph this above. Call the Coefficient of Variation (CV) for short. Note that the United States CV for real estate prices increased sharply from the year 2000 to 2018. This has an interesting economics interpretation. This means that the spatial dispersion in home prices was rising over time. As can be seen from the LA/Chicago graph above, the coastal superstar cities boomed and this must have driven much of this variation. Note that after 2018, the CV has declined by 20%.
This suggests that the rest of the nation is slightly converging with respect to home prices with the superstars.
Buffalo!
Consider this recent headline about Buffalo, New York.
Here is a quote from the article;
Buffalo’s home price growth is an example of post 2018 “convergence” such that cheaper local housing markets are catching up. As these low priced markets heat up, the numerator in the CV formula decreases.
Why Has Convergence Recently Taken Place?
I argue that the rise of WFH has contributed to the home price increase outside the Superstar Cities. The U.S population has new opportunities to figure out how to configure their lives. Here are some academic papers for you to read.
Brueckner, Jan K., Matthew E. Kahn, and Gary C. Lin. "A new spatial hedonic equilibrium in the emerging work-from-home economy?." American Economic Journal: Applied Economics 15, no. 2 (2023): 285-319.
Ramani, Arjun, Joel Alcedo, and Nicholas Bloom. "How working from home reshapes cities." Proceedings of the National Academy of Sciences 121, no. 45 (2024): e2408930121.