Wednesday, December 24, 2014

Housing policy - please do the opposite

I recently saw this column by Robert Shiller, where he made this comment:
since 1890, the average appreciation of inflation-corrected home prices in the United States has been only a third of 1 percent a year. That’s why housing hasn’t been a great investment. And in 10 years, it may be almost equally likely that real home prices will be higher or lower than they are today.
That is kind of a shocking statement to me.  That's like saying bonds are a terrible investment because the redemption value will be the same as the initial face value.  You don't buy bonds for capital gains.  You buy them for income.  Likewise, you don't buy a house for capital gains.  You buy it for the rent.

Some people do buy bonds or houses as speculative activities, but of course speculation is a zero sum game.  That doesn't have anything to do with whether they are good investments.  How can Shiller make this statement?  The question is, how much does the house cost, how much would rent be (corrected for homeowner expenses), and how does that compare to alternative investments?

In fact, the fact that home prices in the US have roughly tracked inflation suggests that thinking of a home as an inflation-adjusted bond is a pretty good first step for looking at aggregate home values.  There is no way that 30 year TIPS bonds are paying a higher return now than the average rental home is.  This has nothing to do with what home prices will do in the next 10 years.


In that column, Shiller also argues against the mortgage tax credit deduction.  This is an interesting issue.  I agree with Shiller about this.  And, I think this would be the perfect time to phase it out.  Affordability is not the binding constraint in housing right now.  Access to capital is.  Households with capital or credit can purchase homes.  The homes are underpriced, so for the few that can buy a home with a large mortgage, they are earning excess rents.  The mortgage deduction is just adding to those rents.  Normally, there would be a fear that ending the mortgage deduction would lead to a drop in home prices that was steep enough to cause an economic dislocation.  But, real estate credit has been too hobbled for the mortgage interest deduction to lead to higher prices.  Home prices are low enough to be profitable for investors, and at least until very recently, cash buyers have been dominant, so if the mortgage deduction was ended now, cash and institutional investors would keep prices from declining significantly.

But, public housing subsidies are interesting to think about.  According to Modern Portfolio Theory, a tradable asset or security that is widely accessible should be bid up to a market price where there are no risk-adjusted excess profits.  Optimized portfolios will be diversified, so they will still be exposed to market risk.  But any exposure to idiosyncratic risk related to individual securities will not have any excess returns in the aggregate, because that risk can be diversified away.

But, with housing, there is limited access, due to the all-or-none form of ownership that is typical, and there are potential gains from idiosyncratic risk, since such large portions of the market are not, and cannot, be diversified.  Also, high transaction costs create a liquidity premium.  So it is likely that there are excess returns from home ownership, especially when a home is held for a long period of time, minimizing trading costs.

But, this issue, as is often the case, gets turned on its head.  Since home ownership provides excess profits, and these profits tend to go to households with the most access to capital, we tend to think, wouldn't it be fair if everyone could get access to those profits?  This is wrong-headed.  Profits (accounting for liquidity and idiosyncratic risk) only exist because there is limited access to the market.  If everyone gets access, the profit goes away.  The solution is to get rid of the profit.  But the irony is that, if we get rid of the profit, then moving households into home ownership is not necessarily a benefit.  The only equitable outcome for housing policy isn't to subsidize more home owners, it's to make all households indifferent to home ownership.

This is a very difficult distinction for the consensus to accept, because in social policy, we tend to associate middle class behaviors with social improvement, and it is natural to assume that social progress comes from nudging the lower economic classes into that behavior.  If middle class behavior involves the accretion of economic rents, universality of middle class behavior is a mathematical impossibility.

A more universal and equitable housing market would come from lower transaction costs, more access to credit, more pathways to ownership, etc.  This is a good example of how hard it is to have good public policy in an IMH world.  The housing market of the 2000s was a great example of a context where the housing market was more universal and equitable.  Low interest rates, low down payments, investor diversification through securitization - all of these trends were pushing down excess returns in housing and expanding the pool of potential home owners.  All great things!  In a world where rent payments will be fairly stable, how will lower excess profits (economic rents) be manifest?  Through higher asset prices!  And, what kind of reputation do the housing market and the financial industry of the 2000s have?  Public views of the housing market are definitely an example of strong form IMH.

On all sides of the political spectrum, there seems to be a consensus that the Fed is the lap dog of Wall Street, making sure the financial elite earn economic rents.  And there also seems to be a consensus that the era of equity and universality in housing was a monstrosity that had to be beaten down.  And the consensus complaint is that the financial intermediaries who made that housing market possible, who for the most part are bankrupt, reorganized, or a fraction of their former selves, have been "bailed out" because the Fed dares to inject some liquidity into the chasm that used to be a credit market.

The mortgage tax deduction really made everything worse.  For those who could capture excess profits, the deduction increased those profits.  And, on a macro level, it creates a two-tiered market, where there is a general price level for landlords, and a higher price level for owner-occupiers who can benefit from the deduction.  This means that supply and demand forces for landlord owners will translate into higher rents.  This also means that the market for single family homes, especially homes with higher nominal values, is much thinner than it would normally be.  There is a market of non-diversified, owner-occupiers with equilibrium prices above the equilibrium price for landlords.  When that market broke down in a context where (1) excess profits had been bid down because of wider credit access and (2) the mortgage deduction continued to provide added profit for owner-occupiers, prices had to fall significantly before landlord investors were willing to add support.

So the mortgage deduction creates a less stable, less equitable market.  I wonder what the counterfactual would have been if, during the 2000's, we hadn't had the mortgage interest deduction, but we still had low interest rates, securitization, and all of the other accommodations that came out of low real interest rates, low inflation, and financial innovations.  Home prices would have been somewhat lower.  Home ownership rates would have been lower.  There would have been much broader landlord demand for homes.  Rent inflation would have been lower.  And, I think that it is plausible that there would have been more supply of homes as a result of all of these factors.

Partly, what was going on in the 2000's was that very low long term real interest rates were pushing up the intrinsic value of homes - the value of homes as an investment.  This was pushing up the landlord owner equilibrium price of homes.  The equilibrium price for owner-occupiers was naturally above that price, at least partly because of the mortgage deduction.  And, low nominal mortgage rates pushed that price even higher, as low monthly payments meant that constraints on demand coming from mortgage credit access were much lower than they had ever been in the modern era. (This is a separate effect from the effect of low rates on the actual nominal value of homes as a durable asset.) But, liquidity issues, transaction costs, the microstructure of the realty market, and the inability for buyers in the owner-occupier market to expand their holdings, meant that there were a lot of frictions and price stickiness in the owner-occupier market.  This created the opportunity for a lot of speculative activity.  But, it also might have kept supply from rising quickly enough to meet demand, because home builders were generally limited to finding buyers among that thin, friction-filled market of owner-occupiers.

If we hadn't had this two-tiered market, there might have been a more robust market for home builders to build for the renting market.  The relatively lower amount of frictions in the market might have allowed quantities to more quickly rise to meet demand, even as average prices would have remained lower without the mortgage deduction.  Landlord buyers would have been able to build large numbers of homes at prices they could profit from.  With the mortgage tax deduction, home builders could hold out for price levels higher than the landlord price level, but they had to find buyers one at a time from the owner-occupier market.

For many reasons, there has been a deluge of capital searching for low-risk investments, and the housing market served as a useful conduit for that capital.  In a more landlord dominated market, there would have been a much more robust set of saving opportunities through real estate.  There would have been a much larger industry of REITS and mutual funds placing investments in landlord institutions.  Savers would have been able to utilize real estate to meet current savings demand without trying to stuff so much of it through the conduit of owner-occupiers and small-time real estate investors.

Put another way, the equilibrium price of houses is the Price to Rent ratio that creates a return equal to other investment opportunities (including non-financial considerations).  In a normal real estate market, taking away the mortgage deduction means the average Price to Rent ratio will decrease.  This will lead to more quantity supplied from landlord owners, and rents will decline.  The declining level of rent, with a stable Price to Rent ratio, will mean that prices would decline further.

In addition to the lower prices this change would encourage, the mortgage market should allow low income households to establish ownership with few obstacles.  Down payments should be low, and various payment options should be available.  There is no reason to publicly subsidize these things.  Anyone who does establish ownership will likely already be earning profits relative to renting households.  Private mortgage insurance and securitization were available in the 2000's.  There is no reason why we need to bring in the moral hazard problems of public subsidization.  Deregulation would suffice.  In effect, housing policy should be a policy of creating lower housing prices and then not encouraging households to be home owners.

The main problem with this prescription is that there will still remain this deep cultural association of home ownership with the middle class.  As long as we have that association, on the margin, there will probably be households that establish a real estate position when they probably shouldn't.  That would subside over time, and, in fact, may already be less of an issue after the collapse of 2008.  Generally, the positive cash flows from mortgaged home ownership come after years of rent inflation, so there isn't generally a short-sighted incentive to switch from renting to owning.

So, please, take your mortgage interest deduction and give us back the real estate market of the 2000's.  There may never be a better time to do it.  This is what so-called consumer rights activists and working class political heroes should be pushing for.  So, Who's with me?....I said, Who's with me?!.....

...I think I have one more post worth of this nonsense, for those of you still hanging with me....


  1. I'm still hanging with you. Thank you for taking the time to write up your thoughts. You think about this stuff more clearly than any other source I've found. Sometimes I wish there was a way to take people like you and let them write the rules.


    Kenneth Duda
    Menlo Park, CA

  2. Great blogging.

    And I think you are right---Shiller is waaayyyy off base on housing, from the perspective of an individual buying a house.

    So the house only keeps up with inflation. But if your mortgage is fixed, all that inflation becomes equity! That is, if house prices double in 10 years due to a bout of inflation or appreciation, and you never refi-ed, now you own half the house!

    Then, as you mention, there is the mortgage tax deduction.

    If your house is large enough, there are other options, like renting rooms out.

    The tax code and inflation make buying a house a good move.

    Side question. For a while, I have been thinking that house prices, as measured overstate inflation.

    Some housing become more valuable due to location, and that is influenced by where people are moving. So, a 1000-sf condo in Manhattan is worth X, and 10 years later it is worth X+Y.

    Simple to say, "Yes, inflation."

    But the buyer is getting more for their money, even in this extreme case. They are closer to desired amenities, work, other people they like. The condo-buyers get status--and people pay for status, if you think not explain luxury brands.

    And there is something about people with rising incomes competing against against for the limited high-status housing stock.

    All this is my longwinded way of saying I think the CPI overstates housing inflation.

    1. I've wondered that, too Benjamin. Like with Ed Glaeser's work about the value of cities, if the city grows up around a house, the house becomes more valuable. That must happen, on net, naturally. It all depends on how the hedonic adjustments work with shelter inflation. I don't know anything about that. It does seem like the kind of thing that wouldn't be accounted for.

  3. I'm following you, too.

    I had to think long and hard about the trades on the margin between renting and home ownership. After much thought, I figure all the additional expenses I have as a home owner, such as maintenance, taxes and insurance to protect the structure, are included in the price of the rent that a landlord would charge a renter for a single-family home like mine.

    With that realization, I conclude that the short-term economic incentives are clearly in favor of home ownership, considering the mortgage interest deduction first, and the distant second benefit of the homestead exemption on state and local taxes that likely wouldn't be priced into rent I would pay as a renter. The only economic incentives I could identify for a renter are in unique situations: (1) an existing long-term renter in an area with rent control, or (2) a transitory living arrangement that through renting avoids transaction costs associated with the sale of a home (real estate commissions, closing costs, etc.).

    As an aside, I’m wondering about the analogy you are making with owning a home with a mortgage to holding a bond, the latter of which pays a return during the life of the bond. Wouldn’t this analogous “return” during the life of the mortgage be the various tax deductions I get as a homeowner, such as the mortgage interest deduction, homestead exemption, and property tax deduction? Or were you thinking of something else?

    The analogous financial instrument I think of for home ownership is a deferred annuity. I pay for an annuity up front, with the guarantee of an income stream for life later. Paying a mortgage leading to owning a home free and clear provides a future income stream in the rent savings that I would incur otherwise. This is the major factor influencing my choice in choosing home ownership over renting, inspiring me to pay off my mortgage before I retire. Shiller seems to ignore this very important factor in his article.

    On eliminating the mortgage interest deduction, my biggest policy concern is unintended consequences arising as a result. Twenty-five percent of filers take the mortgage interest deduction, who are mostly young and wealthy households. For example, how many of those 25% would be pushed from itemizing to taking the standard deduction, and thus forgo other deduction incentives such as charitable contributions? (The latter is the number three tax deduction behind the mortgage interest deduction and the deduction for state/local taxes.)

    Clearly, there is a big challenge in crafting a revenue-neutral change affecting tax rates, the threshold for the standard deduction, etc. that takes into account first-order and reasonable second order effects. But what other unintended consequence might arise as a result of eliminating the mortgage interest deduction, analogous to the unintended consequences that arose from implementation of the mortgage interest deduction itself? Would they be better? Or worse?

    Thank you for a thought-provoking blog post. I found it from the link from David Henderson’s blog. As my Economics professor, he challenged us to think on the margins, and your post definitely fits that topic.

    1. Thanks Joe.

      Regarding your comments about mortgages, I find it helpful to think about home ownership as a cash purchase, then to separately consider the many implications of mortgage debt. I think the conflation of these factors has led to an overestimation of the importance of mortgage rates on home prices, especially since mortgage rates tend to correlate highly with the real long term discount rate one would use to determine the value of a home from the present value of future rents. If all homes were bought with cash, very low long term interest rates would still increase the intrinsic value of homes.

      The mortgage interest deduction is probably more important for the aggregate market than it is for most individual buyers - acting as a subsidy for a certain class of buyer and pushing up the equilibrium price. But, I don't want to over-sell this. The scale of the effect on price probably isn't huge - well less than 10% probably. But it creates these dislocations in the housing market.

      But, for an individual buyer, other factors will usually be more critical - how long you will be in the house, your risk aversion, your savings level, your expected income life cycle, etc. Since we usually have to buy real estate in an all-or-none transaction, the obstacle of attaining credit lowers access to the market, so that it is usually beneficial to buy if you have access to mortgage credit. You gain access to excess returns from an inefficient market. The mortgage interest deduction mostly subsidizes these excess returns. The landlord owner market is scalable, so it is probably more efficient. I think it is possible that if we didn't have the deduction, a small decline in prices could push a lot of potential buyers from owning to renting. The owner-occupier price would decline, and the landlord price would bid away the excess return to ownership for some households.

      Your comment on the homestead exemption is very good. I haven't mentioned that, because I've mostly been looking at relative home values over time, as opposed to looking at the actual trade-off level between buying and renting. But, you're right. That's a significant legal advantage to owner-occupiers.

      Good points about policy consequences.

  4. Finally, Kevin, a post on housing with which I can find nothing seriously disagreeable! I think there is room for a debate on the semantics of Shiller's comment (I doubt he would find anything disagreeable about your post, here, other than your negative characterization of his comments about housing being a poor "investment"), but that's kibble.

    Otherwise, I am willing to grant you that if there are excess economic rents captured by the landowning class due to market distortions, the erosion of those rents would likely come more from rising asset prices than falling rental rates due to sticky prices. It is not clear to me that this was the principal or even a significant contributor to the rapid price appreciation immediately prior to the recession, but it was probably a factor (inarguably, the homeowner market of the 2000's was more egalitarian than it had ever been, FOR BETTER OR FOR WORSE).

    In practice, as you say, there are numerous public policies which both encourage homebuying, subsidize owner rents, and support real estate prices. Together, these have the effect of keeping home prices AND rents high, so that unwinding them towards a market equilibrium becomes a difficult task, indeed. Maybe a competitive real estate market would see home prices fall and rents rise in the short run, followed by the reverse in the long run towards a stable long run equilibrium where homes acted more like income-generating bonds with little or no (real) price appreciation: a type of asset for a particular class of buyer, rather than the "something for everyone" public policy goal we pursue today. This is likely much closer to the historical norm, particularly in other countries, but the US still wouldn't look like Europe (which I think is the primary fear of housing market subsidy advocates): the abundance of land, capital, and household income here mean we would still have much higher homeownership and square footage per capita rates.

    1. Thanks for the input, Glenn. It's definitely a complex issue. I agree that the interest deduction wasn't a large factor in the price increases of the 2000's. Although, thinking about it, the decline of other market frictions in the 2000's probably increased the aggregate effect of the interest deduction. But, you're right. It has been a longstanding distortion, and at most would increase some home prices by something in the single digits (%). The effect it might have on average home prices, in and of itself, is not the main problem with the policy.

  5. I must thank you for this excellent read! It was absolutely enjoyable throughout. I have you bookmarked to check out the new content you provide
    Good way of explaining. Now it's time to avail bus tours Washington dc for more information.