Kevin Erdmann on U.S. housing supply

He writes,

Just a few cities are at the heart of the housing supply problem, most notably New York City, Los Angeles, Boston, and San Francisco, which I refer to as Closed Access cities. There are two very different housing markets within the United States: the Closed Access market, where new housing is highly constrained, rents rise relentlessly, and households are forced to make difficult choices as housing expenses eat up their budgets; and the rest of the country, where homes can generally be built to meet demand, housing construction is healthy, and housing expenses remain at comfortable levels for the typical household.

Pointer from Tyler Cowen.

This is an important point, which I would like to see stressed over and over. We do not have just one housing market in the U.S.

For the nation as a whole, Erdmann makes two claims.

1. There was no significant widespread overbuilding during the housing boom. This claim is contrary to many people’s impression, but pay attention to his statistical support.

2. There has been under-building since 2007. This claim strikes me as undeniable. But I saw an essay on Medium recently that tried to deny it, using what I thought were inappropriate and misleading indicators.

I strongly recommend reading the entire paper, or at the very least skimming it to get all the main ideas. He is doing important work.

9 thoughts on “Kevin Erdmann on U.S. housing supply

  1. Handle-bait if there ever was one. If I can get some time (not easy, new addition to the family), I’ll write up a more thorough critique.

    But how’s this for a contrarian position. Erdmann says this recovery was not “exceptionally strong”, and most folks seem to want to say that the recovery was abysmal or weak or disappointing and so forth.

    But maybe people were just being impatient and failing to take the long view and falling into the characteristic exaggerated sensations style of contemporary reporting of anything. A lot of people seem very invested in glass half empty interpretations and dismissing good news with a lot of “on the other hands” which don’t dispose of the issue.

    Because nine years later, the US recovery looks kind of amazing. I don’t want to jinx it and provoke a downturn out of God’s need to make fools of us, but think about how this good news would have been reported with ebullience in the recent past.

    Consider: Weekly initial unemployment claims are the lowest they’ve been for nearly half a century and that’s without correcting for a larger work force, and much lower than at any point in the four previous expansions. Indeed, this is the second longest expansion since the early 80’s, and on long term graps is now starting to look like a continuation of “The Great Moderation” with only particularly bad cyclical downturn. The unemployment rate is at 4.1, also near historical lows, and which in the recent past would have been thought to be around the NAIRU / about as good as it could get for the U.S.

    Typical measures of inflation are low but stable and close to target. The market indexes are at record highs, NASDAQ is up nearly 8.7% annualized (nominal) since it’s 2007 peak.

    Small business optimism is high, major reported problem, “difficulty of finding workers”. Oil is up on strong demand. Average hourly earnings have been inching up at a decent growth rate. The employment-population ratio for 25-54 year-olds has recovered to post 1988 normalcy, and the overall participation rate for this age cohort has bounced back from its lows and has been climbing for two years straight (though there is apparently a male-female gap, but then again, if that’s based on data older than 2017, which seems plausible, then it may be missing the latest positive developments.

    “Part time for economic reasons” and “unemployed over 26 weeks” have declined steadily are at at the “good times” levels of the prior expansions.

    Personal Consumption Expenditures have been trending up so steadily the Great Recession looks like a tiny blip, so even if one attributes that to “stabilizers”, it’s hard to argue those stabilizers didn’t work very well.

    Household debt service ratios are at 40-year lors, auto sales have recovered, residential contruction spending is now where it hasn’t been except between 05 and 07, single family serious delinquency rates are below 1% again after 8 years of mostly steady decline, national nominal house prices are now above the bubble peak, with good positive gains and price to rent ratios look roughly like they’re climbing on the pre-bubble trend.

    I could go on, but you see my point.

    Fact is, it’s been an incredible recovery! Ok, tt was just slower than people wanted, and from a lower starting level.

    But, most people recognize that there have been a lot of big and important structural changes in the economy, many of which were part of long trends that just happened to have the GFC in the middle of it. Many people think it was a “slow” recovery, but they are offering judgments based on comparisons to times when things were structurally different. The truth is, we don’t know the new possible range of different quality recoveries in our new economy, and our instincts based on the past are not reliable guides in an era of significant transition.

    And that’s not even counting all the “macroeconomic mood affiliation” (or maybe “normative macro assessment” to complement “normative sociology”), which biases people to feel a particular way about the news based on partisan political calculus instead of based on a consistent approach to judging the facts.

    So my heterodox position is that this recovery has not been disappointingly lackluster, but, contra Erdmann and everybody else, “exceptionally strong.”

    • Very few economists will agree with you that this has been a strong recovery. It is close to being the second longest on record (may already be), but has not achieved the median level of total growth of historic recoveries. In other words, it has been a long recovery, but has not achieved the same rates we have previously seen. This matters, because economic growth rate, in the long run, is the most important factor.

      • What is the most appropriate or useful way for us to define the meaning of “strong” in “strong recovery”? It may seem like mere semantics, but in fact, there is a subtle conceptual problem here.

        Are we to pick some particular growth rate or index and just assign qualitative terms to numbers? 5% = “strong”, 4% = “normal”, 3% = “weak”? Create some kind of “Expected Bounceback Time” metric? Take the set of all past recoveries are judge where we are based on standard deviations away from mean? Are we dealing with policy counterfactuals and comparing to model-based estimates of “macroeconomic potential”?

        What if we are in a “new normal” now, transitioning to a different kind of economy where comparisons to former recoveries are inapposite or as unsuitable as crude comparisons with less developed nations? Maybe we can’t even use “return to prior trend pre-crisis” without the caveat that maybe some of that prior trend – perhaps a pre-bubble-peak ramp-up / speculative frenzy era – was unsustainable or illusory?

        What about “Misery index * duration”? Maybe when a financial or economic crisis has an international character, we can compare growth rates for other countries of comparable levels of development and proportional to their “exposure” to the downturn?

        I tend to be very suspicious that those who use the words “strong” or “weak” in connection to the recovery like to do so in order to feed into “partisan macroeconomic assessment and normative analysis.” That is, they want to preserve the option of characterizing the overall economic situation as something which could be better or worse if someone else were in charge, and if some other policy were implemented, if some social problem were “fixed”, all meant to serve as an input for an argument as to “what is to be done,” which is typically and of course what the author seems to always want to be done, day or night, rain or shine.

        The view from a decade later shows that while this recovery may not have been rapid, for however “disappointingly slow” it may have been at times, it has also been steady and solid and put certain important measures in historically excellent territory for an extended period.

        I think that slower growth rates are the new normal around the world (something I explained earlier deriving from the increased predominance of particular kinds of services), and given that, I say what we’ve seen in the U.S. is indeed a strong recovery, and about as strong as anyone could reasonably hope for.

  2. I see several key points:

    1) I am surprised house size is not included here. Over the decades that has not been constant (I believe increased until 2007 it went from 1,700 sq ft. to 2,300 sq ft. or so.) That impacts the the price and space a lot and this should be account for. In terms of building 2004 – 2006 the bubble had more high end houses (2,700 sq ft. at $500K) and not enough of 1,400 sq. ft. at $310K in SoCal.

    2) In SoCal we are witnessing a lot more stock of multi-family units. I wonder how that is impacting the economics.

    3) It might wise to state there was a housing bubble in the 1980s and do average without 1983 – 1990 data. (The S&L crisis was more commercial but I knew people in underwater in CA in 1992 and our S&L were hit hard.)

    4) I wish Kevin had accounted for housing units for population 25 – 65 as well. I suspect more seniors today (where living separate housing diminishes) might show today not as low.

    • Looking at the house building for the US, it appears US home builders appear to have US Auto Manufacturer disease in which in good time they start only focusing on the most larger sales instead of basic bread & butter sales. I think local regulation plays a role here but it limit the impact of local deregulation.

    • Those are some good points, and some of that is data I just haven’t had the time or ability to cover.

      On home size, you have to be careful though. In terms of price, during the boom, new homes rose much more moderately than existing homes. That is because the boom was mostly facilitating a mass outmigration from the cities that can’t build. Really, if you think about it, that’s all it could do. So, we were building 3,000 sq. ft. homes in Dallas for $300,000 instead of 1,000 sq. ft. units in San Francisco for $1 million.

      As Arnold points out, because of the different types of housing markets that coexist in the US, and because capital repression (through local building obstructions) is an important factor, normal intuitions sometimes don’t play out the way they should. This is one of those cases. Increasing home size was actually a part of a process of reducing real housing consumption through shifting location.

      That’s what’s happening in Phoenix, which I mention in the linked piece. Phoenix, the city, looked like it was being overcome by a speculative bubble. And, it was, in a way. But, those marginal households moving into Phoenix were largely from California, and they were reducing their nominal housing expenses substantially. In fact, that was overwhelmingly their motivation for making the move. That is the case even though they were generally moving into units that were larger, with more amenities.

  3. “There has been under-building since 2007.”

    With as much government intervention in the housing market as we have, I don’t think we can have any idea of what the proper (i.e., free market) level of building is.

    • It is true that there is some government subsidy, mostly through income tax benefits, which mostly cause high end housing to be over-consumed. But, property taxes are also a tax that specifically targets real estate. On net, this might be somewhat neutral, but highly regressive. Subsidies of the GSEs are generally overstated. It amounts to a fraction of what income tax benefits do, if anything.

      Since 2007, there has been an additional government intervention, which has been a multi-sigma shift in credit regulations that mostly serves to prevent marginally qualified borrowers from buying homes. This is part of what has prevented new supply from coming on line. So, prices (especially in low tier markets) have dropped considerably while rents rise. Part of my work (which wasn’t included in the short linked piece) has been to demonstrate how lending to these households was not a significant factor in the housing bubble but the removal of this credit was devastating to low end home equity after 2008.

Comments are closed.