What really drives housing prices? It definitely isn’t housing...

@kaseyklimes
kasey@kaseyklimes
32 views Jun 12, 2025 ~3 min read
Advertisement
1
What really drives housing prices?

It definitely isn’t housing density.

Let's dig into the data…

(feat: lots of charts!)
🏙️💰📈

kaseyklimes.com/notes/2023/5/2…
2
1/ A few days ago Scott Alexander argued that “Density Increases Local But Decreases Global Prices” on the logic that a mobile subset of Americans prefer big cities and that a city will become more desirable to this population in proportion to its size.
astralcodexten.substack.com/p/change-my-mi…
3
2/ By this logic, a city that builds more housing makes itself more expensive to live in.

It’s true that there’s *some* correlation between density and housing prices, but the causal mechanism here is backwards…
Media image
4
3/ The critical missing piece in Alexander's argument: Jobs!

Money doesn't follow housing; housing follows money.
5
4/ Three variables drive housing prices:

1. The number of available jobs
2. The salaries of those jobs
3. The amount of housing available for those workers

Aggregate income (total income from all jobs) & housing supply are tightly coupled...
Media image
6
5/ But if we zoom in, we see that there's a difference between the cities on either side of the fitted line.

High income to housing:
San Jose, San Francisco, and DC.

High housing to income:
Detroit, Tampa, and Phoenix.

[interactive chart📈] kaseyklimes.github.io/incomevsupply.…
Media image
7
6/ We can turn these two measures into a ratio to see the relationship more clearly:

Aggregate income per unit of housing.

How much wealth is being generated and how many homes are available it to flow into?

This single metric explains the majority of median home values.
Media image
8
7/ But wait, outliers! How do we explain these?

These are exceptions that prove the rule.

Key West, Kahului, & Ocean City have unusually high property values because the income that drives those values up comes from jobs in other places—these are vacation home markets.
Media image
9
8/ On the other side we find places with unusually low values, like Los Alamos NM.

Los Alamos is a company town for Los Alamos National Laboratory, which employs 14k in a city of 13k.

High-paying jobs, but with only one employer in town, the pool of potential buyers is small.
Media image
10
9/ Let's cut outliers out and focus on metros with populations larger than 1M.

An r-value of 0.9 suggests that—among these metros—over 81% of median home value can be attributed to aggregate income per unit of housing.
Media image
11
10/ We now have an explanation for why San Jose is more expensive than San Francisco despite having a lower housing density—it's largely *because* of its lower housing density.

Too-much-money flows into not-enough-housing.
Media image
12
11/ OK, but cities are different!

What we really want to know is whether a city can improve affordability by building enough housing to keep up with job/wage growth.

Looking back at the last decade, we find that... literally no city did this.

However, the relationship holds:
Media image
Media image
13
12/ For a way more rigorous analysis of a single city over time, check out @enf's brilliant work collecting and analyzing 30 years of for-rent ads in San Francisco:

experimental-geography.blogspot.com/2016/05/employ…
Media image
14
13/ No one wants aggregate income to fall, so how do we build more housing?

How do we get more cities into the "Wealthy & Affordable City Zone"?
Media image
15
14/ Tactics for making cities affordable:

🚫🏠 Eliminate single-family zoning
🚫🚗 Ban parking minimums
🏗️🏢 Allow single-stair apartment buildings
🪙🏙️ Tax land values
👷🗽 Increase immigration

But most importantly:
End the vetocracy.
theatlantic.com/ideas/archive/…
16
15/ OK *but how do we do all that?*
I regretfully turn to my former profession, city planning.

A lot of this mess is because we gave up.
We fucked up, then we gave up.

Fixing the housing crisis will require planners to step up as the adults in the room.
placesjournal.org/article/jane-j…
17
Thanks to @slatestarcodex for inspiring this dive into the housing data.
18
.@slatestarcodex published a followup w a softer stance, acknowledging there may be a lag b/w building new housing & new demand created by new jobs created by new residents.

Reasonable, but all the more reason to build housing and *keep building housing* astralcodexten.substack.com/p/highlights-f…
19
@MarkusHasiPhone instantiation of which*
Actions
What You Can Do
  • Export as PDF or Markdown
  • Batch Export to Notion
  • Bookmark & Highlight
  • Screenshot Tweet
Create Free Account

Includes 7-day Premium trial

Advertisement