Tuesday, November 24, 2015

Housing, A Series: Part 89 - Low Interest Rates, the Housing Supply Constraint, and Picketty's Concern

Recently I speculated that the sharp fall in long term real interest rates since the crisis is a product of the housing bust.  After looking at the supply conditions of various cities, this issue now seems to me to be very clear - in fact, obvious.

Here is the supply and demand curve I looked at in the last post.  In a market that allows supply, we should expect demand expansion to feed real housing expansion with very little change in home prices.  Ironically, this is the basis for much of the commentary from people who have been leaders in identifying the housing market as a "bubble".  The idea that high home prices must be temporary because of this supply response is central to the "bubble" story.  This is not controversial.  And, in fact, this is what we see in the majority of the country where housing supply is not severely constricted.

In places where housing is constricted, new demand cannot induce supply through a price mechanism.  So, where housing is constricted, demand leads to rent inflation and asset price inflation.  And, this is what we see in cities where the issuance of new housing permits is low.

Now, let's think about the effect of housing supply conditions on capital markets.  In Houston and Atlanta and most of the non-coastal areas, increased incomes or falling real long term interest rates create real housing expansion.  This requires capital.  Existing homeowners don't experience any capital gains or increased rental cash flows.  But, new homes are very capital intensive.  In conditions where demand comes from easier credit terms or low interest rates, the average household increases their real housing consumption, so the new housing stock that is developed leads to higher rent payments in absolute dollars, because there are more, or better, homes.  But rents on the existing stock tend to go down.  Owning a home in this environment is like owning a floating rate bond.  When rates go down, your cash flows go down.  Since home prices are moderated by the cost of new building, the face value of existing homes remains stable while net rental income declines (adjusted for inflation).

This is what happened in most of the US.  This is how the world works in the model used by economists like Robert Shiller.  And, in these areas, there was no asset price bubble.

The natural outcome of these market forces is to channel capital into our most durable type of investment - housing.  This should naturally push interest rates higher.

But, what happens in housing constrained locations?  There is no use for new capital.  The effect of new demand, either because of the value of the location to potential tenants, or because of increasing incomes, easier credit, or lower long term real interest rates can only be to produce higher incomes to existing owners and to produce capital gains for the existing housing stock.

Any transaction of housing stock in this context is simply a transfer from one owner to another.  So, I'm not sure we need to even think through the complications of mortgage markets and credit creation.  At the end of the day, whether the ownership of a property contains a debt component or not, the level of capital has increased, and income to capital owners has increased.  This is the opposite of what happens in Open Access cities, and it seems to me that this could only lead to a decline in interest rates.  Capital owners would find themselves with a windfall of capital gains and income, and housing which would be a natural recipient of that capital would not be available.  Other potential investments on the margin would be bid to higher prices and lower yields.

This leads to an unstable equilibrium.  Lower yields mean that at given rents, home prices must rise, increasing the incomes and asset values of capital owners, pushing yields yet lower, etc.

This should be obvious.  Freedom of entry is the bedrock of a functional economy and an equitable society.  Where freedom of entry is impaired, income flows to capital owners who are protected from new entrants.

Here is a table of home values, by type of city, from 1995 to 2005.


These are broad estimates, using median home values from Zillow and population from the Census Bureau

The closed access cities represent about 15% of US population.  In 1995, they held 27% of housing stock, by market value.  By 2005, as a result of NOT building they had 39% of housing stock, by market value.  Open Access cities, which built 12% of the homes during that period, despite being home to only 6% of the population saw the total market value of their real estate remain level as a proportion of the total.  This is despite a net migration of approximately 1% of the US population out of the Closed Access cities into the Open Access cities, so that by 2005, they represented 14% and 7% of the US population.  This is what happens when we limit access to a good with inelastic demand.  Our Closed Access cities are the problem Thomas Picketty is writing about.  More and more capital is held by a select and protected class of owners, which leads to a sharp increase in capital/national income and a decrease in yields.  But this is happening because these cities are closed.  This is happening because these cities have eliminated the natural response of a capitalist system.

The solution to this problem isn't taxing landlords in California.  Their tenants have inelastic demand for that housing.  The tenants will be paying that tax.  Redistributive taxation will simply feed that unstable equilibrium.  The solution is in Houston and Dallas.  Free entry is the solution.  Dare I say, capitalism is the solution.  Our poorest households have already voted on this matter with their feet.


So, we might say that when low interest rates and credit expansion increased demand for housing, a country dominated by open access would have produced a natural counter-reaction which would have pushed interest rates back up.  Instead, the demand for housing in our most economically dynamic cities was strong enough that these closed access cities became dominant and pushed long term real interest rates down.

Housing imputed yield compared with real long term interest rates.
Source
I have only been talking about the 1995-2005 period so far.  After 2005, the entire country has been a Closed Access country, with regard to new housing.  But, now the limited access is happening through credit markets, so rents are rising throughout the country, but home prices cannot respond.  This still has the effect of pushing down yields in other asset classes, like treasury bonds, but the implied net yield on homes is high.  So, since 2005, there is still an excess of capital that is searching for an investment outlet, and housing now can't be that outlet very well anywhere in the country.  So, home prices are out of equilibrium and owners aren't booking capital gains, but they are still earning higher rents, so we still have an economic context where we are funneling more income to asset owners.  It's just that the prices of one class of assets - houses - don't reflect those cash flows, so the market for selling homes is illiquid.


So, closed access policies in a few cities led to slightly lower long term real interest rates in the 2000s, and national closed access credit policies since 2006 have pushed long term real interest rates even lower.  A valid response to this is that those treasury rates are part of a global market where real long term rates have fallen even in countries without our housing credit problem.  I haven't dug deeply into that issue, but I think it is reasonable to believe a gap of something like $20 trillion in the market value of US real estate, including more than a trillion dollars worth of new homes which have not been built, relative to any reasonable historical trend, could lead to global yield responses.

We might say that low interest rates or lenient credit led to a housing boom which could have been expressed as an expansion of real housing stock, but instead, was expressed as an asset price bubble because of the prevalence of some cities in our economy with closed access housing policies.  But, I don't think that takes the issue to its logical resting point.

Given the prevalence of these closed access cities, and given the economic draw they have due to their geographically captured productive industries, the boom was inevitable.  The fundamental cause of this wasn't the decline in interest rates after the 2000 recession.  The cause was the prevalence of these closed access cities, itself, and the rising rents that this produced.  The housing supply problem created the declining long term real interest rates.  Keep in mind that the problem of closed access housing policies is widespread around the Anglosphere.  Our mortgage problem isn't global, but the closed access problem is - in Sydney, Vancouver, London, etc.

In the US, rent inflation began to rise in 1995.  But, in the late nineties, incomes were rising strongly, so rent remained affordable.  Incomes were rising faster than rents, even with the shortage of housing where we needed it.  And, capital was finding many avenues outside of housing in the tech boom.  In fact, I would argue that the finance and tech sectors share these closed access characteristics with housing, because these limits to access in these cities also limit access to the industries that utilize these highly valuable networks of skilled labor.  So, not only were landlords earning excess rents, but tech firms and finance firms were earning excess profits, and those highly skilled workers were earning excess wages.

But, the hard limit on population in these economic centers puts a cap on how long this can continue.  Eventually, gains will go to the bottleneck in supply, which is housing.  By the 2000s, rent in these cities continued to rise, but incomes moderated.  So, these tensions were building up in the 1990s, but rising incomes meant that ratios of rents and prices to incomes didn't begin to rise until the 2000s.

I submit that the relative costs created by the constraints on housing in the closed access cities had reached their limit by 2005.  Nationally, rent/income ratios topped out at the end of 2004.  If we had pursued monetary and regulatory policies that had allowed housing starts to remain strong, rents would have continued to moderate, and long term real yields would have bottomed out, and possibly risen on their own.  This could only happen naturally with the continued real expansion of the housing stock, even if that had to occur in suboptimal locations.  Instead, we choked off sources of mortgage credit and pushed up interest rates to above their natural levels, until the system of housing finance became dysfunctional and unable to fund a housing market at functional prices.

Instead of recognizing this clear disequilibrium and its relationship between low home prices, high rents, and low new housing starts, many people seem to want to explain this new context by claiming that the huge, outrageous, and now nearly decade-long drop in housing starts was a necessary counterbalance to that relatively minor and short-lived bump above the long-term average rate.


Ironically, what we need to do to reduce these transfers to capital is to bring in more capital.  Build homes.  From 2000 to 2010, population in the Closed Access cities rose 4%, compared to just under 10% for the rest of the country.  This caused their real estate to grow from 27% to 39% of the total US market value.  What if they had grown by 10% or 15%?  That is to say, what if they had increased their real housing stock by 10% or 15%?  Is there any doubt that by 2005, the total value of their housing stock would have been closer to 27% than 39% of the US total?  More capital inflows would lead to lower total capital values.  Is there any doubt that this would lower rents for workers?  Is there any doubt that this would draw in capital and pull up real interest rates?  Is there any doubt that this would lead to more equitable incomes, adjusted for cost?


One pattern that seems to be clear in international economic development is that labor compensation tends to rise as a proportion of national income when capital markets are fair, free to new entry, and stable.  So, we should see these same patterns among Closed Access and Open Access cities.  Closed Access cities should have lower labor compensation and higher capital income.

Matt Rognlie has shown how recent increases in capital income in the West are entirely accounted for by housing.  The intuitively difficult issue here is that, since housing demand is inelastic and housing supply is frequently also inelastic, this increase in rental income comes from declining real housing consumption, relative to other spending.

Fortunately, for my analysis here, the BEA does track incomes by metropolitan area, broken out by Earnings (wages and proprietor incomes); Dividends, Interest, and Rent; and Transfers.

Here are graphs of the proportion of each metropolitan area's income, by type, from 1969 to 2014.  I have included San Jose here as part of the Closed Access cities, which are in shades of red.  Open access cities are in shades of green.

Generally, in the Open Access cities, earnings claim a little more than 5% of total income than they do in the Closed Access cities.  This income generally goes to capital income in the closed access cities.  According to Zillow, median rents in the Closed Access cities have tended to claim from 5% to 15% more of median incomes in Closed Access cities than in Open Access cities, so this very likely represents higher net rental income.

There is no systematic difference between Closed Access and Open Access cities in the level of transfer income as a proportion of total income.  For all cities, we can see that while capital income has risen slightly over time, most of the decline in earnings, as a proportion of total income, has been paired with an increase over time in transfer payments.

If we look at changes in these measures over the entire length of the available data, we see that earnings have declined similarly across these cities.  In Closed Access cities, more of that income was claimed by capital.  In Open Access cities, more of that income was claimed by Transfers.

If we look at changes just over the period of time generally associated with the housing boom, trends look somewhat different.  During that period, earnings grew across cities, generally, as a proportion of total incomes.  Transfers declined slightly.  And, the most significant shift during this period was a sharp drop in capital income in Open Access cities.  Building homes is associated with rising wages relative to capital incomes.

I think there are several interesting things to consider here.  The trends in transfer payments seem to be another piece of evidence that our impression of political movements in the US right now is based on a backwards causation.  Maybe red state (Open Access) and blue state (Closed Access) political postures are a product of their contexts, not the cause of them.  The red states have seen rising transfer payments while costs remain low and low income households migrate in.  The blue states have seen stable increases in transfer payments, high costs, and out migration of low income households.  So, Open Access citizens see a functional economy where families moving in tend to rely more on public support, but Closed Access citizens see a dysfunctional economy where families that remain fail to keep ahead of their rising costs, but don't qualify for public support.

But, I think the most difficult factor to think about here is the changing capital income share during the 1995-2005 period, because the characteristics of housing turn our intuitions on their heads.  Why did capital income in the Open Access cities decline so much during the housing boom?  Because, they built a lot of houses.  Because they increased their real housing stock.  This reduced rent expenses.  In the Closed Access cities, capital income was stable during that period.  Why?  Because they didn't build many houses, and real estate owners captured a large amount of rental inflation as growing income.

The hard fact of the matter is that cost-adjusted incomes will rise in Closed Access cities when they start building housing units, and only when they start building housing units.  Until then, the Closed Access cities will have the income distribution of a less developed economy - high income inequality and high capital income.

I see a lot of commentary and man-on-the-street observations in these cities that amounts to, "I feel really strongly about the housing crisis, and I really support attempts to build more housing units.  So, as long as they don't block the views of the existing buildings, as long as they meet a range of demands from local activists, and as long as they build units in the range of rents that seems fair to me that include many units at below-market, I say, build away."  That is the Closed Access policy.  Builders in Phoenix don't have to set aside 30% of their units for below-market rents, and like some sort of perverse magic, market rents in Phoenix somehow are more affordable than below-market rents in San Francisco.

10 comments:

  1. Hear, hear. Though I do think, at least in SoCal, you will find plenty of "conservative" Red Statish types (Orange County, San Diego County, Ventura County) who zone to the max (even minimum acreage requirements) to prevent "highest and best use", which is what you are talking about.

    My guess is that Ventura and OC would have condo booms too.

    Downtown L.A. actually has large housing projects underway

    It is homeowner groups that stop development, largely.

    ReplyDelete
    Replies
    1. Yes. That is true. On the other hand, there are homeowner groups in Houston, Dallas, and Atlanta. Why do they only act to limit access in blue states? The answer to the Closed Access problem has to be something about California that makes it different than Texas. Maybe it's just some mundane detail about the political process that sets a series of political trends down a path that leads to obstructionism. But it's got to be a detail that doesn't exist in most of the country.

      Delete
  2. Hi Kevin,

    Have only picked up reading the tail end of your work on housing and migration (looking forward to working through more of it over the long weekend), but find it quite thought provoking. Recent bloomberg brief (here: http://newsletters.briefs.blpprofessional.com/document/2cz1hfilorhzpm732h/regional-focus-us- ) looks at US housing price trends, with the key figure here: https://pbs.twimg.com/media/CUqs-akW4AAh3v5.jpg

    The geographic pattern is pretty striking, and obviously there's some shale oil beta driving prices in a couple of those states, but it shows that states more representative of open access have seen more price appreciation of housing, even though the ability to expand stock is less constrained. Would you see this as a result of migration flows out of the closed access areas driving larger supply-demand imbalances in open access areas? I find it somewhat surprising that CA is still down relative to the peak, but I guess the state binning makes the effects a little squirrely, too.

    One outlier relative to your proposed open/closed analysis seems (to perhaps my uninformed view) to be Nevada. It would seem that geography, at least, is not exactly constraining the ability to build out there, and I wouldn't think of it as "closed access". What, in your view, explains the sharp reversal in the NV housing market around the 08 peak? To me, this does seem more like the spec driven bubble story that is currently accepted.

    Thanks,
    Andy

    ReplyDelete
    Replies
    1. Thanks for the input, Andy.

      I think several things are going on here. First, I do think there was a very small part of the boom, in 2004-2005, in Arizona, Florida, Nevada, and the Inland Empire, where some confluence of factors pushed home prices above fundamentals. So, compared to the peak, these areas will probably show more decline.

      As for the rest of the country, I believe that because we mistook the supply problem for a demand problem, we destroyed the mortgage credit market, and it remains weak even today. This has put a huge obstacle up to new homebuilding. So, now the entire country is "Closed Access". Rents are high and returns to home ownership are high.

      If mortgage markets ever heal, housing starts should surge, interest rates should rise, home prices would rise a little, and rents would fall relative to other prices.

      Delete
    2. By the way, Andy, I think the first sentence from your link is a good example of how insane and self-destructive we have become.

      "Earlier this month, the Boston Fed's Eric Rosengren wondered aloud in a speech to the Portsmouth, Rhode Island, chamber of commerce whether the swelling number of cranes dotting Boston's skyline should be a source of worry."

      The amazing thing is that this comment is taken as reasonable. Because we have been wrong about everything, we literally have become afraid of progress. We literally have public leaders who see progress (full stop) as dangerous. And we beg them to make it stop. I wish I could take everyone who is currently within reach of the levers of public economic policy management and ship them away some where. They just need to stop. Stop. Stop. Anyone who thinks there is too much support for real estate development now should have their PhD stripped away and be given a job cleaning toilets or bussing tables somewhere, where they might do some good. They are worse than useless. And the public loves them. Asks for more.

      Delete
  3. You cannot build away in SF. Maybe you can in the other cities, but in SF the risk of going up is in the possibility of a gigantic earthquake. I am sure this fear of the big one is not lost on the leaders of the city.

    ReplyDelete
    Replies
    1. Kevin, perhaps greed trumps earthquakes. That was certainly true of Japan, where they built nuclear plants on subductions zones and rationalized their safety. One other point, You fail to understand the massive demand of long bonds, also leading to low interest rates. And if the dollar is backed by collateral, as the Fed has claimed, how can the Fed allow the excess reserves the banks hold to get out into the economy without some collateral being fronted. That is a serious question and I don't know if you have the answer.

      Delete