Bryan Caplan  

Population Externality Bleg

Values Every Scholar Should Ad... How to Spend a Billion Dollars...
Suppose a city's population exogenously rises.  You might think that price theory clearly implies that demand for real estate will rise.  But that's not so.  In theory, higher population could generate a congestion externality so awful that demand for real estate actually falls.

If you're having trouble picturing this, imagine how much you'd pay to live in Manhattan given current conditions.  Now imagine how much you'd pay to live in Manhattan if the streets were so crowded you had a 10% chance of being trampled to death every time you left your apartment.  A sufficiently massive population could drive Manhattan rents down to zero.

Theoretically, then, the effect of population on real estate prices is ambiguous.  In the real world, though, I've never come across a credible example of an real estate market where crowding has blatantly reduced rents.  My bleg: Has anyone got a credible candidate for me?  I'll accept examples from any part of the world in any time period.  I'll even accept cases involving a sudden influx of destitute refugees into an area.

What have you got for me?

Comments and Sharing

COMMENTS (25 to date)
QPR writes:

Kowloon Walled City might be one (at least in that rents would have been collapsing relative to rents in the rest of the Kowloon City area) but is such a bizarre edge case as to likely not be useful.

Himanshu Sanguri writes:

I have observed the effects of congestion externality on the real estate prices where an old, crowded city is surrounded by a number of satellite budding cities. Let me illustrate the scenario in context of National Capital Region of India. Delhi is surrounded by Noida, Gurgaon, Ghaziabad and Faridabad. The story started with Delhi, Ghaziabad and Noida, where all development begin in early 1980's. However, Delhi and Noida were planned cities but story went opposite in case of Ghaziabad. Today, Noida, Delhi, Faridabad and Gurgaon are witnessing an almost 45 degree positive slope curve in real estate prices along with same growth pattern in crowd coming in. But, for Gahziabad, it is not the same case, crowd coming in is still increasing, but people prefer to settle in Noida and Delhi and commute daily to their work places in Ghaziabad. This has resulted in decline of real estate prices in Ghaziabad.

8 writes:

If you look at the price of real estate per sq ft or sq m, and look at the size of homes there, it is very clear that the price of real estate drops as density rises. Or conversely, as wealth rises, people will pay for lower population density. The homes next to Central Park are more expensive, for example.

Tyler Cowen writes:

You need the average return everywhere else to be falling for this to work. That condition is not impossible, but it is far more stringent than the way you describe the scenario and I suspect you won't find such a case, at least not with (relatively) free mobility.

BC writes:

I believe that Yogi Berra once said about a restaurant, "Nobody goes there anymore. It's too crowded."

Brent writes:

So it is congested but no one lives there? This seems like a negative network effect that would reduce some persons' demand, but not overall market demand.

Else, I could decry that the local sports team tickets are too expensive... and there are not enough seats.

Handle writes:

I think I see where you're going with this, immigration, open borders, yadda yadda, ok. But maybe we can get a different take on "externality".

I've got a completely crazy science-fiction / fantasy theory. So unrealistic a thought-experiment it's really just for entertainment, but, please, bear with me.

What if you live in a place, let's call it California, where there is universal voting and democracy ... oh, you heard this one? Well, please, humor me, ok?

And California used to have some actual competition between the two parties, and that competition helped keep the parties saner than they'd be otherwise. Voters are pretty irrational, after all; someone wrote a book about that.

Anyway, let's say a huge number of new people starting crowding in to this "California" with, for some strange reason, a lingering, multi-generational bias or tendency to vote for one of the parties that was skewed from that of the pre-crowded popuation. Crazy! I know! But, let's roll with it, you know, just for the sake of argument.

And that, eventually, that one party achieved a status increasingly similar to permanent one-party rule, and the lack of competition made that party both more Socialist, more extreme, and much more likely to pass and implement awful, anti-libertarian, anti-economic-efficiency legislation. Impossible and Silly! I know! What is this, Tolkien? Will there be dragons too!?

And these policies basically made the state much worse from the perspective of the previous middle-of-the-distribution majority population, who eventually began to flee elsewhere because of the changes.

In fact, in a kind of irony and paradox, the very people who advocated for the newcomers to be allowed to enter and crowd discovered that the government of this "California" was increasingly likely to do the exact opposite of all the other things they thought it should do. And yet people who wanted it to do even more of those opposite things happened to agree completely on only one particular agenda - which was to let in many, many more of these newcomer crowders. Odd.

And while "externality" is a bit sketchy of a concept in some contexts, it sure felt like the new crowders were generating some with the above-described phenomenon.

Ok, we can all go back to reality now.

Philo writes:

(Echoing Brent.) The increased congestion externality would drive *down* the population. What you might realistically have is increased congestion in, say, Manhattan due *not* to increased population in Manhattan but to increased population in the surrounding boroughs

(Echoing Handle.) Besides the congestion externality from immigration there is a *political* externality.

Jeff writes:

BC beat me to the Yogi Berra allusion.

Grant writes:

Could extended protests count? Imagine there is a protest in the capital city that causes thousands of people to fill the streets for months. Did property values fall during the protests in Egypt? Did the Occupy movement lower property values in DC, New York, or elsewhere? Obviously, in Egypt there could be fear that property could be destroyed, so its not so much congestion. For some of the more peaceful occupy protests (here there would be less of a concern that you property would be destroyed), the congestion might have lowered property values in the surrounding areas.

David C writes:

@Handle Almost every state has effective one-party rule. In fact, California isn't even the most liberal, New York, Rhode Island, Vermont, and Maryland all have it crushed. I think California's problems have more to do with the design of its state government. IMO, populist movements favoring high voter turnouts, limited terms, and lots of statewide elections for changes to laws do not a good government make.

David C writes:

Woops! I left out what I think is the biggest reason for California's poorly designed government. California tends to be an early adopter of new policies and ideas. Governments tend to not do so well with policy experimentation.

Handle writes:

@David: Things have changed. And they can change more - things can always get worse.

In the 90's, the State Assembly was nearly even. Even during Clinton's reelection in 1996, Californians chose 37 out of 80 assemblymen as Republicans. There was Wilson and Schwartzenegger. In the last three elections, the number of Republican assemlymen has gone from 32 to 27 to 25 - now giving the Democrats the unstoppable 2/3rds in both houses, the prospect of which they've salivated over forever, but could never quite achieve for some odd reason, but which now seems incredibly solid and secure, for the same odd reason.

The question is not the state but the change. Do we have important demographic mechanisms that correlate with this dramatic and rapid shift with a strong causal signal? We do. Have we seen this before in American history? We have.

But don't take my word for it. Take the word of all the professional Democratic politicians who are licking their chops trying to Californize the whole country as quickly as possible. Hey, who can blame them if they can get away with it - it's a great strategy and victory is so valuable it's worth doing almost anything.

Which is why there's an externality.

Adam_Smith writes:

[Comment removed, awaiting confirmation of email address. Email the to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Salim writes:

There's a tangentially similar notion in highway engineering: as you add more cars to an empty highway, the cars-per-minute throughput obviously rises. But at a certain point, adding more cars (via onramps, etc) actually causes the cars-per-minute throughput to fall. Smart highways use tolls or red lights at onramps to keep the number of cars to a manageable level - and actually carry more passengers.

You might find that demand for a highway causes the amount of toll each one is willing to pay to fall. In fact, that could be an equilibrium: demand rises exogenously (say, a sports event ends), increasing congestion at the original toll. If the toll were a market price, it might then endogenously fall, leading to even greater congestion. This could certainly occur under low-information conditions, and possibly merely when drivers have heterogeneous preferences.

Mark Brophy writes:

According to Wikipedia, the population of Boston fell from 801,000 in 1950 to 562,000 in 1980. Real estate prices fell sharply as Bostonians fled to the suburbs. The population of Philadelphia fell from 2.1m to 1.7m during the same period. New York City population decreased from 7.9m to 7.1m. San Francisco population decreased from 775,000 to 679,000.

Many other cities lost population during the 1950-1980 period because people wanted less congestion. It's likely that real estate prices decreased sharply in all these cities.

Michael Strong writes:

Long-term inflation adjusted property values are remarkably stable. See, for instance, this study on commercial Manhattan office space over a hundred year period,

While not identical with the congestion issue, the authors do suggest that constantly decreasing transportation costs is a factor in the long-term stability of prices,

"The results can also be consistent with the combination of historical population growth and transportation improvements that characterized New York since the early 1800s. From 1830 to 1900, New York City grew from a population of 300 thousand to 1.8 million and began to spill beyond Manhattan. With a tripling of average density, such growth would have necessitated a doubling of the city’s radius – or equivalently of average commuting distances. It is easy to imagine that the introduction of even the inefficient streetcar doubled average commuting speeds, leaving total commute times to the urban “edge” the same. From 1899 to 1999, the City’s population grew roughly fourfold again, and expanded into the full tri-state area. During this period density actually began to decline. Even if the distance to New York’s urban edge had increased four fold – the commensurate greater speeds of underground subways, trains and automobiles could still leave average commute times constant – and hence real land values as well! If construction costs grew only with general inflation during this century – as they have since the 1960s – then asset prices would also not increase in real terms."

Surprisingly, a study of Amsterdam real estate prices over four centuries also showed remarkable stability in inflation-adjusted prices over the centuries,

Peter writes:

"Suppose a city's population exogenously rises. You might think that price theory clearly implies that demand for real estate will rise. But that's not so. In theory, higher population could generate a congestion externality so awful that demand for real estate actually falls."

Are you talking about a subsequent fall in population; or does the population remain high but demand for real estate changes in a way that reduces offered prices despite the high population? In the second case, the need for accommodation is unchanged by the congestion, so if the price is lower then presumably that is only because the congestion is reducing capacity to pay for real estate or changing the real estate demanded from more expensive lower density homes that are less well served by public transport to higher density homes that are cheaper but better served by public transport.

ajb writes:

I think you confuse rising population affecting rents vs population rises in equilibrium. In the worst cases -- say Detroit -- the population does not rise in equilibrium because the worst losers simply flee the inner city. In cases in which they had stayed or found reasons to stay, prices may rise. But some of these rising prices may be compensation for unwanted externalities or the loss of various rents. Rising population can destroy the rents of existing populations but generate enough benefits for those who desire those benefits to push or buy out the incumbents. Where those previous rents were not fully internalized, this results in a net loss for the incumbents who move without gaining from higher prices which may not be enough to offset say loss in school quality.

Handle writes:

@Mark Brophy:

"Many other cities lost population during the 1950-1980 period because people wanted less congestion..."

That's why people left the inner cities in that period? Congestion?

Why didn't people want less congestion before it got congested? Did they want the congestion beforehand, but then change their minds about it?

If a new restaurant opens up and it's really good, gradually more and more people pile in until it's always crowded. Flash forward a few years and you see that there are always plenty of table available now.

Do you conclude it's because the restaurants patrons "wanted lower wait times" (which they happily put up with before), or do you conclude that "perhaps something happened to diminish the perceived qualify of the restaurant itself"?

JKB writes:

I don't think imagining a real risk of being trampled is a good driver. People will adapt, move closer to their work, etc.

But imagine it is so congested that it slows the movement of goods produced by productive activity. This works better with physical goods and from what I've read, a lot of manufacturing has left Manhattan. When a productive activity can't receive supplies or ship product effectively, then congestion will push it out. Before then, owners and workers move closer to the activity to mitigate the risks and costs of congestion driving up real estate prices.

One might consider what they did to Atlanta for the 1996 Olympics. They gave privileged access to the roads to elites, limited movement of supplies and goods to night time and forced everyone else to us public transportation. Oh and they shut down non-Olympic businesses. Now, imagine this wasn't a two-week special event. Imagine that this was done due to ongoing and increasing congestion of the normal workforce. Now, what would businesses, shoppers and people not held in thrall of their Atlanta-based employment do? When might it happen that productive enterprises start moving out of the city due to the need to move goods, recruit the best employees, etc. What might that do to real estate prices.

Of course, as we saw with most major cities from 1970-2000, it doesn't have to be congestion that does this, it can simply be a change in the economic value of the productive enterprises.

I would submit that the driver is when congestion becomes a drag on productive activity, not threat of trampling that is the point where the real estate market turns.

mobile writes:

If enough people are willing to live in a super-crowded Manhattan that they're willing to take on a high risk of trampling, then super-crowded Manhattan must be a super awesome place to live, and the rents must be pretty high.

Or as Scott Sumner would say, never reason from a trampled-in-the-street hazard rate change.

NZ writes:

As other commenters have pointed out, you sure don't see this phenomenon too often on the city- or state-wide level.

But you do see it on a lower level. To perpetuate the restaurant analogy: if you go eat at a self-seating restaurant, you probably scan around for a table that is surrounded by other open tables, and avoid an open table that is surrounded by occupied tables. This is doubly true if the other diners are very different from yourself in some inherent or cultural way.

Steve Sailer writes:

You could start by looking at the densest municipalities in Los Angeles County:

A few dense Los Angeles County communities such as West Hollywood have high home prices, but most of the densest -- Maywood, Bell, Bell Gardens, Hawaiian Gardens, etc. -- are dumps.

Jaime DeVargas writes:

As someone who grew up in a smaller city, I can understand the problems with not wanting to be in a city that's overcrowded. The idea of a city being overcrowded can be unappealing.
-However, I have never seen or heard of a scenario where congestion has lowered rent costs. Looking at prices in New York City which is heavily populated, prices are high. We move over to London, which is the third highest populated city and see that rent is the highest its ever been. So lets say that there is an example were overcrowding lowers rent costs. What would make certain crowded cities capable of having high rent and others would feel the need to lower?
-If anything wouldnt there be more evidence of the opposite effect? Housing rose during the Industrial Revolution while population was increasing(Guess that could also be from inflation)

Comments for this entry have been closed
Return to top