Wednesday, October 26, 2016

Housing: Part 182 - The Only Solution for Closed Access Cities is to Build Luxury Market Rate Units

A common complaint in the Closed Access cities is that they can't allow housing expansion at market rents because all the developers want to build is luxury apartments.  When they try "trickle down" housing policy, it just attracts more high income residents to the luxury apartments and just puts more pressure on working class tenants.

The surprising thing is that, if you think about it for a bit, they are actually making an arguable point.  Or, at least, it would appear to be an arguable point to anyone who was simply trusting their own experiences and observations about what was happening.

But this is one of those Bastiat things, where the good stuff is all happening outside our field of vision.  Here's why.

As a city's housing supply becomes constrained, the first response of households is to reduce their real housing expenditures.  We can see this effect, really, in just about all cities, in the "drive until you qualify" distribution of home values.  As you move into the city core, more amenities and shorter commutes mean that you get less house for the same money.  The same adjustment happens across the board in a city as the entire city develops constrained housing.  The house has to get smaller or the drive longer in order to keep costs manageable.

In cities like Seattle, Washington, DC, Minneapolis, etc. where there are supply constraint issues, but not to the level of the Closed Access cities, households generally can make those adjustments across the income spectrum.  The median household tends to feel comfortable with rent in the 20% to 30% range.  This has moved up to 25% to 35% because we have imposed a housing shortage on the entire country by killing the mortgage market.  But, as the first graph here shows, up until 1995, rents settled around a common range.  After 1995, cities with especially high incomes (which can only be maintained by exclusion) also developed very high rents - even as a proportion of those high incomes.  This is because housing became so constrained that households couldn't downsize any more, and for households that valued remaining in those cities, they had to allow housing expenditures to rise as a portion of their incomes.

Washington retains normal rents, as do Seattle and other cities that are regarded as expensive because they still are functional enough that most households can downsize enough to keep housing in that comfort range, even if the home itself is not as physically comfortable as they would prefer.

Here was my attempt at graphically describing this effect.  As housing becomes constrained, households begin to allow it to take on more of their household budget.  Demand becomes inelastic.  But, at some point, income effects become too strong, and demand becomes elastic again for the current residents.  At some level, there is a maximum amount that households can spend on housing, where demand maxes out.

This is why Closed Access cities have a peculiar migration pattern of low income households moving away in such large numbers.  At some point, there is potential in migration, and these in-migrants have less elastic demand for housing the current residents.  In other words, they have higher incomes, so when rents rise, they can accept the higher rents on the given housing stock, so they can bid up rents, forcing the previous residents out.

Across the distribution of incomes, households move along that demand curve to the right.  The higher their incomes, the less they need to spend on rent as a portion of their incomes.  They achieve this by reducing their real housing expenditures, as households in many cities do.

This is another old graph of mine that describes this phenomenon.  At low incomes, rent pulls in a large portion of Closed Access incomes.  As incomes rise, households in Closed Access cities spend a similar amount as households in other cities do, on rent.

So, at the two extremes, there are households with very elastic housing demand, in terms of the proportion of their incomes - but convex at one end and concave at the other.  If we imagine an expansion of supply in the graph above, rents will decline.  For low income households, they will tend to remain in the same unit, and will allow their rental expenses to decline from 60% to 50% of their incomes.  For high income households, their rent spending is already at a comfortable 20% of their incomes, so falling rents will allow them to increase their real rent expenditures until they can once again spend 20% of their incomes on rent, but for a more comfortable unit.

Thinking about this in the aggregate, notice what this means regarding the effect of new supply on housing consumption.  The pent up demand for housing is focused within the high income households.  They are naturally the households that will expand their real housing consumption.*

So, if a Closed Access city manages to expand the supply of units, the overwhelming natural demand for those units is going to come from high income households.  The measure of how much that new supply helps low income households isn't going to come from seeing low income households move into new units.  It has to come indirectly.  First, by seeing that they have more money to spend on non-housing consumption, because that's what they have been cutting back on.  Second, by seeing that they stop fleeing the city by the thousands for lack of affordable housing.  Closed Access cities are so dysfunctional that expanded real housing consumption among low income households is way down the road in any scenario that addresses the problem.

Since the direct effect of new housing - what the locals see - is high income households buying them up - expanding their real housing consumption, it looks to any reasonable observer like "trickle down" economics.  Then, they react by obstructing new market-rate units.  And, ironically, it is the obstruction that continues to make their city dysfunctional.  It is the obstruction that makes those units "luxury" units.  It is the location.  The reason the location makes those units status goods is because of the constricted supply.  So, again, Bastiat might point out to us that across the city, low income households are forced out of existing units that have now become a little more "luxury" because of the obstructed supply.

The irony is that many of those forced out households move into units in Phoenix or Dallas with higher ceilings, more square-footage, etc. than those "luxury" units in the Closed Access cities, because in cities that don't obstruct building, developers have an abundance of choices for all types of households.

* In fact, the opposite of this is what happened in the housing bubble.  More flexible mortgage terms allowed high income households in the Closed Access cities to expand their real housing consumption.  In the Closed Access cities, where home prices were by far the highest, this was not a phenomenon focused on low income households.  Low income households were being forced out of the Closed Access cities by the hundreds of thousands during the bubble, because high income households were able to use those mortgages to outbid them.


  1. Terrific post. SF and LA should construct literally hundreds of thousands maybe millions of new luxury units.

    Eventually the supply would cut housing costs.

    An interesting idea:Raise gasoline taxes a dime and cut bus and subway fares by same. Repeat until traffic thins out.

  2. Terrific post. SF and LA should construct literally hundreds of thousands maybe millions of new luxury units.

    Eventually the supply would cut housing costs.

    An interesting idea:Raise gasoline taxes a dime and cut bus and subway fares by same. Repeat until traffic thins out.

    1. It will be interesting to see if cities allow the high levels of innovation in transportation tech to lead to private replacements for public transportation.

  3. Kevin: Hey I am all for private vans, self-driving taxis or privatizing subways etc.

    In Thailand there are private pick-up trucks with slats over the wheel wells (the long ways) and a simple roof that provide "mass" transportation. (We have to strangle trial lawyers first, btw).

    But the kind of density I am talking about probably requires the sky trains or subways. Even places like Singapore go to mass transit. There is just a practical matter of moving millions of people around, and goods, with a limited amount of space.

    In short, I think LA and SF and OC would densify relentlessly in free markets, and just swamp the already swamped roadways.

    Remember, as housing became cheaper, even more people would move in, and businesses would start hiring again. The thing could snowball for decades.

    You could have metered pricing for use of roads.

    But probably, lots of sky trains and subways. Pushing people out of cars through gasoline taxes is one approach, as is cutting mass transit fares.

    BTW, if your one-hour commute to the office become a half-hour due to gasoline taxes…you benefit too. But not captured in official economic data.

    But hey, ain't going to happen. The propertied classes of America have embraced property-zoning and feel that the demerits of the minimum wage are much more interesting to talk about.

    1. It seems like there will be opportunities for innovation in (mass) transit, but that we will need to divorce the ideas of public transit and mass transit, like the trucks you describe in Thailand. As will AirBnB & Uber, American polities will probably be obstructionist. On the other hand, I am surprised at how quickly driverless autos are developing.

  4. An amazing article it’s nice to read a quality blog post. I think you made some good points in this post.
    bookkeeping UK

  5. Kevin: The driverless autos are amazing. Unbelievable! Still, three million driverless cars on the roads of L.A.---all trapped in traffic. There is only so much road space.

    Metering roads--$20 per vehicle or something--might work, as the vans with 10 people would make economic sense. Just keep raising the metered rates until traffic clears off enough to run.

    But mass transit has been universally adopted in every very dense city. I think it must be a practical response. Even Singapore.

    BTW Bangkok had a "free market" response to heavy traffic for many years. It was called 3-hour traffic gridlocks, five days a week. They have been building Skytrains, which are crowded.

    There are times when ideology fails reality.

    1. Driverless busses seem like the most immediate and useful way to use driverless tech - mass transit. But I think it will develop most quickly where it isn't run by the municipality.