Saturday, July 28, 2018

Housing: Part 314 - Costs will rise to meet price

I recently saw some research that suggested that the cost of building in the San Francisco area was high - that the typical new unit costs about $600,000, and most of that is construction costs.  Unfortunately I have lost the link.  But, I want to address this idea.  The conclusion this would lead to is that building in San Francisco is just naturally expensive, and that it is those natural costs that create high rents and prices rather than zoning or political obstructions.

But, I think that is incorrect.  Here, I think the comparison to Detroit is useful.  Building cars used to be very profitable in Detroit.  Detroit had natural advantages which were magnified as supply chains developed around those advantages.  The natural evolution of the industry led to a context where certain firms who had survived within the marketplace captured excess returns from these path-dependent advantages.  But, those advantages were geographically captured, so they could be siphoned off by local governments, unions, etc.  Eventually, an hour of labor in an automotive factory in Detroit cost more than an hour of the same labor in other locations.

That wasn't disruptive as long as the advantages of operating in Detroit were strong enough to share some of the economic profits.  But, the problem was that other places - notably Japan - developed their own economic advantages, and eventually the costs of operating in Detroit outweighed the advantages compared to other locations.  Now, it is to the point where automotive production is expanding in many areas other than Detroit, in spite of Detroit's previous advantages.  That is because claims on economic rents are more difficult to minimize than natural costs are.

But, it's not like those costs are separated out nicely.  It's not like you can easily look at the cost of building a car in Detroit and say, "OK, this $100 is natural, this $100 is economic rents." etc.  All those costs are bound up together with natural costs.  Who can say, with precision, what portion of local taxes are legitimate and what portion were due to wasteful spending that grew because the availability of excess profits could fund it?  Some of the cost of building the car is due to high wages.  But, much of those wages may also have been claimed by a myriad of other local costs, so that, even if workers were willing to give up their portion of the fading economic profits, they wouldn't be able to because their costs would be higher, and it would be difficult to determine which of those costs were reversible.

So, it would seem reasonable to say, "It is expensive to build cars in Detroit."  But, that would sort of be the opposite of the truth.  It was never naturally expensive to build cars in Detroit.  Yet, once a geographically captured source of economic profits develops, costs will inevitably fill the gap.  In fact, one could say that the reason Detroit has become less competitive is because it was too economical to build cars there, and sticky rent seeking filled in the gaps where production costs were low.

This is how it costs $600,000 to build a housing unit in the San Francisco area.

It really only costs, say $200,000, just as it would anywhere else in the country.  But, political obstruction has created an oligopoly in San Francisco housing, boosting the market price to $600,000.  That difference would go to land, some might be claimed in development taxes and fees, some would go to queuing and financing costs, etc.  Many of those costs would get rolled up into the actual cost of construction.  Local laborers require higher wages to pay for their own housing, for instance.  And, in addition to adding fees and taxes to new developments, local governments can add requirements which make building more expensive.

There is no natural impediment to adding those costs because it doesn't affect the incentive for developers to build or the market price.  It comes out of queuing costs or the cost of the land.

I sort of half jokingly wrote about California requiring solar panels on new homes, concluding that while that might be a dumb policy in most places, in California, those solar panels might produce some benefits, whereas the capital that would go toward the solar panels would otherwise have just been wasted in queuing or would have been transferred to land owners.  That is an instance where these costs get folded into building costs.  It is an example of how none of these issues reflect a Machiavellian collective policy structure.  There are just various groups with various forms of ideology or self interest who are vying for access to these economic rents.  On this case the solar panel interests plus some ideological environmentalists got a piece of the rents for their own purposes.  And, those rents now are reflected in construction costs instead of taxes or queuing.

If none of these costs were imposed, the difference between price and cost would all be waste or inefficient transfers, as I argued in the solar panel post.  So, in a way, fees and taxes, and even regulatory demands, are an improvement over the status quo at any point in time.  But, the nice thing about queuing costs and land prices is that they aren't sticky at all.  If added supply brought prices down, queuing and land costs will adjust quite easily.

So, if these economic rents have been changing the local landscape for long enough for these economic rents to have been rolled into costs, that is really bad news for the San Francisco area.  That means that there is now no functional way for new supply to bring costs down to a functional level.  It means that "market rate" supply, and even publicly funded units, really can't bring down high rents.


  1. Need to remove the political barriers?

  2. Great post though on some levels I think I disagree with the conclusion.

    If San Francisco actually embraced no zoning , and people developed the PC fever that more housing must be built in order to help poor people and that any rules that obstructed that socially just goal were bad, then I think you would see supply rising enough to eventually curtail market rents.

    I am talking that you see steel container housing and 80-story condo towers in perfusion.

    But, as a practical matter, the scenario you paint is much more realistic.

  3. If there is no land, say a tract of land, in SF, you are stuck with a million dollar vacant lot. Interesting thing on Zillow is, some lots are being foreclosed on. Bubble may be popping.