One of the main problems with selling housing deregulation is the perception that new construction “only benefits the rich.” Rich developers of course, but also rich home-buyers. It’s easy to see where casual observers get this idea. New housing is usually nice housing, because over time technology improves and capital depreciates. Since richer people are more willing to pay the upcharge for nicer housing, the future residents of new construction are usually well-to-do.
So what do casual observers miss? They miss the big picture: People who move into new construction are moving away from older construction. When they move, those older units become available for others. While those others probably won’t be drastically poorer than those they replace, they tend to be slightly poorer. Think: “one rung down.” When these slightly poorer people move, their prior dwellings will tend to be taken over by those who are a further rung down. And so on, in a great chain reaction. Allowing new construction really does help the whole income distribution.
Since this is hard to visualize, picture a game of musical chairs. With one key difference. A normal game of musical chairs starts out with one chair per person, then subtracts a chair every turn. The result: Faster, aggressive kids push out everyone else, until the fastest, most aggressive kid wins. In my variant game, we start out with fewer chairs than people, then add a chair every turn. The result: Slower and more pacific kids start getting places to sit, until there are enough chairs for everyone.
Both games feature a competitive scramble. In conventional musical chairs, however, the competition gets more and more cutthroat and in the end almost everyone loses. In my reverse musical chairs, in contrast, competition gets milder and milder and in the end everyone wins.
I don’t advocate playing reverse musical chairs at children’s birthday parties. Games and stories where everyone wins are notoriously dull. The struggle is part of the fun. In real life, however, the Alice in Wonderland outcome of “All have won, and all must have prizes” is a dream come true. Deregulating construction won’t instantly deliver high-quality, affordable housing to everyone. Instead, it’s like my game of reverse chairs. Every new structure built makes the competition for housing a little milder, until practically everyone comes out a winner.
READER COMMENTS
Phil H
Mar 22 2021 at 9:32am
I don’t think this is a great analogy or a useful model of what markets are. There is much more housing per person in the United States now than at any previous period in its history; but the competition for housing is not milder. Quite the contrary, prices are high, and in the hotspots, astronomically high.
Increasing the size of the supply and the market is good for everyone. But the mechanism by which things get better is emphatically not reducing the level of competition. Market competition remains equally fierce at all times, but the quality of housing improves.
robc
Mar 22 2021 at 9:50am
I think the analogy is better if he each new chair is fancier and more desirable than the last. The fastest, aggressive kids get the fancy chair and the slow kids without a chair get the plain one.
Christophe Biocca
Mar 22 2021 at 10:13am
If you’re talking about the average/median house-size metric, that’s probably a poor measure of actual availability for a few reasons (beyond the obvious issues with averaging data over the entire country):
Laws limiting occupancy, minimum lot sizes, and the like, cause this number to go up, while making it harder to get housing.
Destroying housing, as long as it is below the median/average, makes the number go up.
Unoccupied housing that no one wants still counts. All the abandoned homes in Detroit still count towards the metric.
As a thought experiment, if you bulldozed 146 McAllister Street in SF and replaced it with a parking lot, this metric would record it as an improvement, and vice-versa.
KevinDC
Mar 22 2021 at 1:29pm
This struck me as a surprising claim, so I spent some time digging through data on population growth and housing unit growth in the United States over the last few decades. And, it turns out, this is false. I got data from the St. Louis Federal Reserve on the housing stock over the last 20 years, and US population data from the UN World Population report, and did some quick calculations. As of 2020, there were 2.35 people per housing unit in America – the same ratio of people per housing unit as there was in 2005, and in many of the years in between, the ratio of people per housing unit was lower than it is at present. The average ratio over the last 20 years is 2.36 people per housing unit – so you could say that the moment the amount of housing per person in the US is very slightly lower the recent historical average, but your claim that there is much more housing per person than at any previous period isn’t true.
However, that’s all besides the point. The point of the post was about the benefits of allowing supply to grow in order to meet demand. The per-capita availability of a particular good doesn’t shed any light on whether supply has in fact been growing fast enough to meet demand. You can have a good where the per-capita availability is higher than it’s ever been, and also have supply at its lowest relative to demand than it’s ever been, at the same time. Those are distinct concepts. If in 2021 we found that the housing stock and population both grew by exactly 3%, that would in no way allow you to infer the housing supply is growing on par with demand. That doesn’t follow. You could have a 10% increase in the housing stock and a 10% decrease in population, and still have supply growth fall short of demand growth. A growing population is one factor which can increase demand for housing, but it’s far from the only factor. If you’re looking for data on all this, you can check out this report to start. Key conclusion:
Phil H
Mar 22 2021 at 11:20pm
I was thinking of a longer timescale than 20 years, and housing units are getting bigger.
More importantly, though, if this is true:
“The per-capita availability of a particular good doesn’t shed any light on whether supply has in fact been growing fast enough to meet demand.”
Then it negates BC’s claim even more strongly than it negates mine:
BC said: “Every new structure built makes the competition for housing a little milder”
Both your point and mine demonstrate that this is not right. You by suggesting BC missed a factor (demand growth); me by suggesting he’s got the model completely wrong: in fact, in a market, the competition *never* goes away.
KevinDC
Mar 23 2021 at 8:30am
Unfortunately, the FRED data for housing units only went back 20 years, but I’m happy to dig into other data sources. Although, if you want to save me the trouble, could you please cite the specific data sources you used which led you to this conclusion?
You have misunderstood the argument. Everything I said is perfectly consistent with the claim “Every new structure built makes competition for housing a little milder.” My point was that increases in per-capita availability do not allow you to infer how much supply has grown relative to demand. That in no way implies that increases in supply don’t moderate the level of competition compared to what they otherwise would be.
Imagine a country with a fixed population. In this country, as people get wealthier, demand for housing goes up. The population is 1,000,000, and there are 200,000 housing units. In the next year, to meet increased demand, an additional 100,000 housing units would need to be built. Now consider a few scenarios.
In the first scenario, only 10,000 new houses are built. Now, technically, per-capita availability for housing would be at the highest it’s been at any point ever! And yet, competition for housing will be extremely fierce, and much more so than it was the previous year, when the per capita availability for housing was lower. Given that demand has grown much faster than supply, prices will rise significantly to make the quantity demanded equal to the quantity supplied.
In the second scenario, 80,000 new houses are built. Demand growth has still outpaced supply growth in this case. And yet, it’s perfectly obvious that with the additional 70,000 houses compared to the first scenario, the competition for housing will be milder than it was in scenario one, and while prices will rise, they won’t rise by as much as they otherwise would have.
In the third scenario, 100,000 new houses are built. Supply growth has matched demand growth. And competition for housing will be much milder in this scenario than in either of the other two – it will be on par with how it was in the initial scenario. Because supply growth has matched demand growth, the quantity supplied and quantity demanded will clear the market at the same (real) price as before. (Obviously I’m simplifying a lot of things here – maybe the new houses will be more expensive because they’re built with more expensive material or the new land used is highly expensive, so prices still rise etc – but I’m assuming all that away for now because the topic here is just how changes in supply and demand affect competition compared to per capita availability.)
In all three scenarios, per capita availability of housing is “higher than ever”, but the point I was making was that per capita availability of some good is independent from relative shifts in supply and demand for that good, and info about one doesn’t give you any basis to draw conclusions about the other. So, the data point you cited, even if true, is totally irrelevant.
But as I hope I also made clear, that in no way negates the idea that increasing supply moderates the level of competition compared to before. In scenario two, the competition for housing would be more intense than it was in the initial setting – but that doesn’t change the fact that the additional 70,000 housing units would make the competition more mild compared to scenario one – and if even more houses were built, it would become milder still.
Also, you seem to end on a bit of a straw man – you suggest that “in fact, in a market, the competition *never* goes away,” but you also quote Caplan as saying “Every new structure built makes the competition for housing a little milder.” There’s a world of difference between saying “competition will get a little milder” with claiming “competition will go away.” Very different things there! (I suppose you could be claiming that the level of competition in a market is somehow exogenous to that market and exists at a permanently fixed level, and doesn’t change regardless of other market conditions. That would equally rule out both the moderating of competition and the disappearance of competition, but that would be a very uncharitable reading to attribute you with a view that silly and ahistorical.)
Alabamian
Mar 23 2021 at 8:53am
It seems like we need an explanation of the portion of demand that isn’t driven on a per-capita basis. I am sure that there are others but there are a few that come to mind:
Household size has generally been decreasing over the past few decades. If the same number of people are split into more households, there is a greater demand for housing.
As the absolute level of wealth rises, demand for housing increases due to more people owning multiple houses. In a wealthier society, more people can afford — in real terms — the costs associated with having vacation homes, condos in the city, etc.
Foreign demand. It seems likely to me that this would primarily be at the very top-end, so probably marginal and something that can be controlled for.
Speculative: Services like AirBnB, VRBO, etc. have increased demand for rentable space, and it is disproportionately in places where there is a lot of friction associated with adding additional housing. So increased demand without new supply.
KevinDC
Mar 23 2021 at 10:47am
Alabamian –
All very true. These are just some of the reasons why housing available per capita doesn’t provide meaningful info on supply and demand in the housing market. They’re separate concepts and conflating them only leads to confusion.
Phil H
Mar 23 2021 at 1:28pm
“the fact that the additional 70,000 housing units would make the competition more mild compared to scenario one”
*Not a fact.
Not historically informed – as I pointed out above. Not theoretically backed up (you don’t have a theory of what competition intensity even means). Not empirically supported.
”a view that silly and ahistorical”
Cheers. That makes you sound very intelligent.
KevinDC
Mar 23 2021 at 3:18pm
Hello again Phil –
Help me out here – where did you point that out above? You made an unsupported claim that there is far more housing per capita in the US than at any other point, but from the data I’ve been able to find that’s simply not true. I did ask what data sources you were using to derive that claim, but so far you haven’t shared the source of your information, which leaves me with fairly little to go on. In addition, I explained why per-capita availability of housing isn’t a relevant measure for relative supply and demand in the housing market, or as a measure for how competitive a market is – and you haven’t offered any rebuttal to that either. So…I’m still left wondering – help me out?
And if you think that the idea that increasing the supply of housing lowers the relative level of competition in the housing market is somehow not theoretically backed up nor empirically supported, I don’t know what to tell you. It’s probably one of the least controversial and most supported ideas you can find among economists, and there is a huge theoretical and empirical literate in support of it. I already provided you with a link to one very recent report showing data to that effect, and you offered no response to any of that data either. If you want more empirical and theoretical support, I’m happy to point you towards more – there is a lot of it out there. I mean, a lot! You can start here, or here, or here, if you like. If you think you’ve managed to overturn this entire body of work that’s been built up over decades, I would encourage you to make your work publicly available, because that would be a huge contribution to field of economics.
Steve Fritzinger
Mar 22 2021 at 10:13am
Hermit crabs figured this out eons ago. It’s a great, real life example of how the housing market should work.
https://www.facebook.com/watch/?v=402896140417078
Tiago Santos
Mar 22 2021 at 10:49am
Great analogy and post.
Matthias
Mar 22 2021 at 11:15am
I wonder if people who are into the kind of fallacy you suggest believe that building fancy housing creates rich people to occupy it ex nihilo?
(Though even that would be great. Just create a bunch of rich people and tax them to pay for your public services.)
Christophe Biocca
Mar 22 2021 at 11:46am
I think it’s usually viewed as “importing” rich people rather than creating them. But that doesn’t make much sense either: those people are perfectly capable of bidding for existing housing stock, and while more stock can increase the inflow it can only do so by shifting equilibrium along the demand curve (ie. by lowering prices).
Some people have taken the position that increasing density shifts the demand curve right in and of itself by creating new industries and employers, but if you take this position then you get to reject “more housing will make housing cheaper” only by accepting “more housing will create enormous increases in real wealth”.
robc
Mar 22 2021 at 11:54am
I think it is people that cannot think past first order effects.
They think “Poor people need more housing options. But all I see is housing being built for rich people. There needs to be new housing for poor people.” They don’t walk thru the steps Caplan talks about. They are oblivious to second order effects. I think almost all problems “regular” people have with economic issues comes down to That Which is Seen, and That Which is Not Seen.
.
Laron
Mar 22 2021 at 12:26pm
Mahalos for this. I like this analogy and will use it the next time this exact topic comes up (this induced demand argument against housing development is common in Hawaii).
Do you happen to have any research to cite for this reverse musical chairs/moving up the ladder type of migration among tenants/home buyers? I’d love to be able to cite some papers/articles but don’t even know what to call this.
Alabamian
Mar 22 2021 at 12:42pm
I think that this is all true. It’s too bad that the trickle-down logic has gotten a bad reputation and there is a popular aversion to any use of it, even when it’s accurate.
I’ve always liked the counterfactual: “Yes, let’s stop the market rate housing from being built! That way, the rich who would have bought it will . . . . [decide they don’t want nice housing?!?] [go elsewhere?!?] . . .”
You can either build nice new housing that the wealthy want to live in, or the wealthy will buy not-so-nice not-so-new housing and make that what they want to live in. People have to be housed somewhere.
Philo
Mar 22 2021 at 12:58pm
Another metaphor: Trickle-down. Build expensive housing for the rich, and housing will trickle down to the poor. (But I suspect that deregulation would also allow some new high-tech cheap housing to be built, directly for poor [or stingy] buyers and renters.)
zeke5123
Mar 22 2021 at 2:04pm
What is meant by new builds being largely for the wealthy?
It would be odd to me that in terms of housing we are talking about the 1% or even the top 10%. Presumably, new builds are a bit more egalitarian because a contractor should be able to get rich providing needed housing to e.g., the middle class.
Is that not happening? If not, why (e.g., maybe the US housing stock is over-supplied so there isn’t much profit in building anything but mansions, maybe regulations make land prohibitively expensive making new builds more expensive)?
Lizard Man
Mar 23 2021 at 11:44am
It varies a lot from state to state and metro to metro, but this “regulations make land prohibitively expensive making new builds more expensive” is largely what’s going on in the metros with large median multiples. In metros with lower median multiples (and more undeveloped land and less onerous regulations/permitting processes), developers are buying and developing greenfield sites with housing for middle class households.
Frank
Mar 22 2021 at 7:45pm
With a constant population, and extra house that is more expensive will lead to a reduction in the price of all other houses. 🙂
Drea
Mar 22 2021 at 10:51pm
I have a memory of a study, which I think I read about in Reason, that traced the chain of tenants back from both “affordable” housing and market rate housing, and found that the market rate housing units brought 30% more people out of homelessness than affordable housing.
But now I can’t find it on Reason, and my google-foo is similarly weak. Does anyone have a citation, or am I just pushing confirmation bias back into my past?
Christophe Biocca
Mar 23 2021 at 8:49am
I can’t find a reason article covering it, but there was a paper ~2 years ago with some coverage that matches your description.
The actual link in that post broke in the in-between years, but it still be can be found here.
Tim
Mar 26 2021 at 11:45am
@Bryan, cross-posting a comment from Alex’s MR post.
As someone who spent several years doing real estate development in Charlottesville, VA, I’d offer it as an example of when a good violates 101 supply and demand curve models.
Land is scarce in funny, non-linear ways. In certain cities or in certain constrained physical areas, increasing real estate development supply a little can increase demand more than a little, and can therefore increase some prices overall. Like a herd effect.
Parts of Charlottesville’s housing and commercial development sectors follow this counter-intuitive pattern for several reasons: (a) there is large, pent-up demand from current residents who cannot access starter or mid-tier homes, (b) there are a small number individuals who buy up as much housing and commercial real estate stock as possible and drive market prices above competitive equilibriums, and (c) there is enormous pent-up demand from non-residents (UVa Alumni for example) who want to move back or retire.
As a result, if you built a certain number of units in certain places, you would both increase supply and drive up the prices of the surrounding units.
Note: you might come back with many reasons for why there should be greater structural changes for why things like this shouldn’t happen. And you would make great points! Nevertheless, for boots-on-the-ground real estate development, the model I describe is real.
Comments are closed.