Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Home prices rose everywhere
Are rising prices in Kalamazoo a symptom of “climate refugees” moving to the cool, Rust Belt uplands? That’s a hypothesis put forward by a friend who works in the climate-insurance-housing cost nexus. It’s a provocative hypothesis. How else can one explain why demand for homes in Rochester, Syracuse, Youngstown and other snowy, rusty places has risen so much?
To kick the tires on this idea, I ran the number: six year home value growth rates by metro, from Zillow (full size map here). If climate is the central piece of the story, it should show up as low price growth in sea-level, hurricane-exposed states, and high price growth inland.
The clearest takeaway is that there’s nothing special about Kalamazoo. That metro has experienced nationally-average price growth.
To be sure, the worst-performing housing market in the U.S. is in Louisiana, where actual hurricane destruction hammered the physical housing stock and the insurance market alike. But Miami – which is a sinking ship by many accounts – has had higher price growth (much more construction) than, say, Greater Boston or Philadelphia.
Other regions with low price growth include Colorado, Texas, and Northern California. But even in those areas, the rates are not very far below average. Mostly, the nation was hit by broad inflationary factors. The local deviations from that are small – perhaps even smaller than usual – and swamped by the national story.
None of this proves that climate isn’t a meaningful underlying factor. Perhaps prices in Florida would have doubled without rising fears and insurance rates. Perhaps Youngstown and Syracuse would have had below-average growth without climate refugees. But unless that’s proven, we should stick to stories that can explain why prices rose about the same in Kalamazoo, Deltona, OKC, and Seattle.
It seems to me that the “climate refugee” idea is the “peak oil” of the 2010s. My fellow urbanists wanted to believe that Sunbelt sprawl would go away on its own, so they talked themselves into believing that the world was running out of oil. Now some of the same people want to believe that climate change will make Sunbelt sprawl go away.
The 2008 mortgage crackdown pushed home prices deep into a disequilibrium so that existing homes were too cheap. That required excess rent inflation of something close to 25% on average to get prices of existing homes back high enough for new homes to “pencil” for builders. In fast growing regions, that happened more quickly, and by the time Covid threw an extra wrench into supply chain capacity, prices had risen enough in fast growing cities to trigger some increased construction.
Kalamazoo, and most of the Midwest, is still bridging the gap back to equilibrium, so new home permits in Kalamazoo are flat as a pancake still at about 30% of what they were pre-2008.
https://fred.stlouisfed.org/graph/?g=1NSIB
So, in the Midwest, there is perennial excess rent inflation. There is no supply response in multi-family because they are treated as a nuisance. There is no supply response in single-family because rents are still bridging the gap to get back to equilibrium prices, though I think most of those markets are getting close.
Despite the low rents, Kalamazoo is in fact building. Per capita building permits have been right around 2 per 1000 since 2013, about half the US rate, and on par with the Chicago, Philadelphia, and Santa Barbara MSAs. As a cross-sectional matter, that isn’t hard to understand. But why hasn’t production risen from 2013-2025 as home prices rose? The unresponsiveness in land-rich places suggests that hard costs play a large role.
https://www.fhfa.gov/research/papers/wp1901
According to this paper from the FHFA, in the 4 ZIP codes they track in the Kalamazoo metro area, the average residential lot value increased from $41,000 in 2012 to $63,425 in 2022, which was almost twice the rate of inflation. There is a role hard costs are playing in allocating capital among undersupplied regions, but it seems unlikely that they are binding if land prices are rising. Rent inflation in Kalamazoo has cumulatively been nearly 50% since 2015. It would be odd for that much of an impulse to fail to trigger a supply response.
Mostly what is going on in Kalamazoo is that home and land prices were pushed too low, and most of that rise in rents has been from a deep disequilibrium condition. I suspect they will see a bit more rent and land inflation as we continue to work out of the temporary national supply constraint on new construction. For now, oddly, their rent inflation is sort of a consequence of their slow growth. Rent inflation and the return to an equilibrium condition that could trigger new supply happened first in the fast growing cities. So, now, rents are levelling out in some of those places, and Kalamazoo rents are still likely rising back to a point that makes new building on the city’s outskirts valuable enough to purchase farmland at marginal prices under post-2008 market conditions.
5 years from now, rent inflation will have flattened in Kalamazoo, and construction will be at least twice what it currently is.
I was skeptical of this – if Kalamazoo prices are really rising spectacularly, they must be far above previous highs, and that can’t be explained by the rebound from previous declines.
The data are sort of in-between on this. Prices in Kalamazoo (PCE deflated) have only risen 23% since Jan 2000. So most of the post-2011 rise is just recovery, which fits your story. But the post-pandemic rise broke new ground, pushing above $250,000 ZHVI. My guess is there really is a demand increase driving that.
I think what is happening is that the long decline in construction permanently wiped out production capacity which has to be slowly redeveloped. Basically, the construction industry has been dealing with hysteresis. There is a limit to the second derivative. Growth in new construction has been rising, and will continue to rise, but when staking out novel capacity levels, it seems to top out at 100,000-200,000 units in construction growth annually. Covid only made that growth more difficult. Some inputs are still binding new growth. Electrical transformers for new community development are still slowing sales, for instance. So, there is cost pressure on inputs. Builders are choosing where to allocate those inflated inputs. As those constraints loosen and construction continues to rise, construction costs will normalize and builders will put more attention to markets like Kalamazoo.
The implied value of land has risen in every market because of supply-related rent inflation, and there will be a relatively predictable overshoot in each market related to the scale of local demand. Some markets have already overshot a lot and more construction will start to reduce land prices and rents. Others, like Kalamazoo, are going through the process at a lower scale and more slowly. I think developers that find the right Midwestern markets to build an inventory of lots where they are ready to build as construction capacity grows and input costs normalize will have a lot of potential for profitable construction.
I think we agree, then
Generally, yes.
Sorry if my comments came off as “disagreement”. They were intended more as “adding color”.