Friday, December 2, 2016

Housing: Part 190 - Rent and Mortgage Affordability

Zillow maintains measures for both rent and mortgage affordability, which basically measure the amount of the median household's income required to afford to rent or to buy the median home with a conventional mortgage.  I think this chart of housing affordability in the largest 100 cities does a good job of telling parts of the story.  (About 1/3 of the country isn't represented in this graph.  That rural portion of the country tends to have lower rent and mortgage costs, similar to the left end of the series shown below.)

From 1981 to 2001, the main change was that lower inflation premiums in interest rates made mortgages more affordable.  So, from 1981 (purple) to 2001 (blue), the relationship maintained its basic shape, but mortgage affordability moved down quite a bit.  (The trendlines are 2nd order polynomials.)  Even in 2001, rent affordability was relatively normal and mortgage affordability responded to rent affordability at less than a 1 to 1 basis.

Then, the relationship changed.  The change in the relationship coincided with rising rents.  (The entire set of red dots shifts to the right, and the high rent cities especially shift.)  Notice that mortgage affordability where rents were low did not shift.  The shift in mortgage affordability was strongly related to rent.  The higher rent expenses were in a metro area, the more mortgage affordability shifted up from 2001 to 2006.

This is why when researchers looked at national data, it looked like prices were rising sharply with little relationship to rents.  This relationship is non-linear.  So, when prices were aggregated among these cities, the high prices in the very expensive cities in 2006 moved the aggregate price estimate up more strongly than they moved the aggregate rent estimate.  A time series of the national data looks like prices unmoored from rents.  But, clearly the change in the 2000s was that home prices became much more responsive to rent levels, at the MSA level.

An expansion of credit availability may have helped push home prices up where credit allowed home values to reflect rents in ways that they couldn't before.  In a city where rent affordability is up to 35%, there is no federal agency that regulates rental contracts, refusing to approve rental agreements that take more than 30% of the tenant's income.  But, there are federal agencies that enforce the amount of income mortgage contracts might claim.  This constraint declined during the private securitization boom, which probably did allow home prices in high rent areas to be bid up.  But, note, this is a significant change to the story.  This would imply that home buyer markets were becoming more efficient during the boom, not less.

I'm not sure how much that explains.  Expected future rent inflation can explain the non-linear relationship between price and rent.  Credit might have helped facilitate those price levels, but, as I have shown in previous posts, given rising rents and prices, home prices across each MSA were rising in a surprisingly systematic way.  On a Price/Rent basis, home prices in zip codes we might expect to have been more credit constrained did not move above the level we should have expected, given the rising rents they were experiencing.  Rent levels had an effect on prices, expected future rent levels had an effect, and credit availability probably accommodated some of that, but I'm not sure how much.

If one believes that credit was the causal factor in home prices and pushed them away from efficient prices, one might be able to come up with an argument that explains why credit only flowed to certain cities, and only pushed prices up where rents were high.

This is where the experience of housing markets since the boom shines a light on things.  We severely cut back on the mortgage market.  We killed the private securitization market, and the GSEs severely curtailed the number of loans they make to the bottom half of the market.  (FHA did make up for some of that, but they also tightened standards.)  Yet, killing the mortgage market did not shift the relationship between price and rent back to the previous regime.  Since the end of the boom:

  1. Rents have risen even more.
  2. Prices have declined everywhere, so that mortgage affordability at the cheap end of the spectrum is extremely low.
  3. High rent cities still have high prices, and there is a non-linear relationship between price and rent.
We still have cities with very high rent.  We still have cities with high rent inflation expectations.  (And we have high rents that justify the rent inflation expectations implied by the boom prices.)  We don't have a generous credit market.  Yet, the new relationship remains.  Screwing up the credit markets didn't do anything to flip the regime back.  It just hamstrung the housing market across the board, whether mortgage affordability was 20% in your MSA or 60%.  Because mortgage availability wasn't the cause of the regime flip to begin with.  The collision of highly valuable core city labor markets with incapable homebuilding markets in those cities was.

PS.  Note the odd movement in rents and mortgage expenses from 2006 to 2015.  Those dots all moved sharply down and to the right.  Since the housing market has become defined by these overwhelming limits to supply and credit, we have an ironic condition where the only way to create more affordable housing is for home prices to rise.


  1. Another great post.

    I quibble with your courtly use of the phrase "incapable homebuilding markets."

    It is not the builders or markets that are incapable. It is government intrusion into free markets through ubiquitous property zoning and other stipulations.

    Just say it out loud!

  2. Another great post.

    I quibble with your courtly use of the phrase "incapable homebuilding markets."

    It is not the builders or markets that are incapable. It is government intrusion into free markets through ubiquitous property zoning and other stipulations.

    Just say it out loud!

    1. You're right. That was a poor choice of words.