Housing quiz

By Felix Salmon
April 30, 2010
Richard Florida:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Chart of the day comes from Richard Florida:


He’s re-running the numbers on the rent-vs-buy debate, and the first thing to notice about this chart, if you just look at the distribution along the y-axis, is that there really aren’t all that many cities below David Leonhardt’s cut-off of 20, and there are precious few indeed below Dean Baker’s cut-off of 15. Most of the country, it seems, is still pretty expensive on a rental-ratio basis.

But be careful with those numbers: Florida, here, is comparing house prices in these cities to rents in these cities. But that ratio is always going to be higher than the ratio between the price of any given house and the annual cost of renting it, since houses for sale tend to be bigger and more expensive and located in nicer neighborhoods than houses for rent.

That, in turn, helps to explain at least a little bit of the startling positive correlation we see in the chart: the cities at the top-right are more likely to be ghettoized into expensive neighborhoods where everybody owns and cheap neighborhoods where everybody rents. Meanwhile, I’ll happily hazard a guess that the cities at the bottom-left will see much less in the way of that kind of differentiation.

But I think there’s something real going on in this chart as well, and it’s rational, to boot. Let’s ignore the difference in housing stock between purchase and rental properties for the time being, and create an oversimplified model where prices are set, in a rational market, at the net present value of future rents. Does it make sense that the prospects for future rent increases are much more robust in LA, NY, and SF than they are in Tampa and Atlanta? Of course it does: vibrant urban centers are much better placed, economically speaking, than endless crumbling exurbs.

In turn, what that says to me is that the market does a pretty good job with respect to pricing in future appreciation or depreciation in rents. There’s a clear baseline cluster around the 20 mark, and then an adjustment is made to that number according to the rental outlook from city to city.

Still, the top five cities are clearly outliers and far too expensive: it would be foolish, I think, to buy in any of them. Which raises an important question for Richard: What are the other two cities in that cluster of five up in the top right hand corner?

I’ll drop Richard an email to find out, but in the meantime put your guesses in the comments, and I’ll try and get his publicist to send a copy of his new book to the first people who get the answer right.

Update: Quiz over. It’s Honolulu and San Jose.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

Boston and Portland?

Posted by singsingsolo | Report as abusive

Boston and San Jose.

Posted by Tierra | Report as abusive

Honolulu looks like a sure thing. The other maybe San Jose?

Posted by stevenlyons | Report as abusive

San Jose & Seattle

Posted by kmitchelson | Report as abusive

off that, San Jose & Honolulu – it helps if you look at the map…..

Posted by kmitchelson | Report as abusive

But other things suggest Las Vegas and Sacramento.

http://washingtonindependent.com/82864/i s-it-really-a-good-time-to-buy-a-house

Huh, I hope you find out the answer.

Posted by Tierra | Report as abusive

Honolulu and San Jose

Posted by petertemplar | Report as abusive

Ugh, I keep telling my friends in costal cali to resist the urge to stop renting and buy…falling on deaf ears I think they close may 1. And Property taxes are still such a killer with Prop 13.

anyway I’m gonna guess San Jose and… Boston?

Posted by wolphkaat | Report as abusive

In re, “expensive neighborhoods where everybody owns and cheap neighborhoods where everybody rents,” in San Francisco 68% of occupied units are rented. In LA, 65%. In New York, 72%. It’s hard to ghettoize vast majorities like that.

Don’t get me wrong, prices here are insane. Our nice, rent-controlled apartment costs less than half the after-tax-subsidy ex-principal-payments costs of ownership. But my wife wants to buy anyway.

Percentage of occupied units in her childhood home that are owner-occupied? 77%.

Posted by wcw | Report as abusive

I don’t know if the figures are calculated very differently, but rental yields in London are about 5% (ie 20 ratio), with an average proporty price of about $550,000.

(see here http://www.londonpropertywatch.co.uk/s/p h?pc=LON&t=1-4b&c=y)

This would be very much an outlier on your chart in terms of price, but not insofar as it is 20 like most of them. Are there rent controls in the US, or is it a measurement difference/

Posted by mjturner | Report as abusive

You heard it here first – there are no crumbling exurbs in the LA area. Because Riverside County is a completely different part of the world from LA. Oh, and California’s finances are as sound as a pound, so no worries.

Posted by johnhhaskell | Report as abusive

Boston and Honolulu

Posted by david3 | Report as abusive

Honolulu & San Jose. Outside chance one is Boston, but Boston is probably the other point down a bit lower with Seattle & San Diego.

One other thing, your comment about ghettoizing neighborhoods just doesn’t hold water. So, New York and LA are more likely to have renter ghettos and owner neighborhoods than Atlanta or Tampa? Have you BEEN to Atlanta or Tampa? If anything, it’s the other way around… New York has tons of high-end renters. It’s one of the few markets in the country, because it is just so expensive, where you have people with very high incomes who actually still choose to rent.

Posted by ACS | Report as abusive

Rather surprising that the graph shows Washington DC is more ‘reasonable’ than Baltimore. Something odd there.

Posted by MattF | Report as abusive

Bostona and San Jose were my initial feelings, but I’ll switch San Jose to Portland.

Posted by MitchW | Report as abusive

MattF, DC is almost assuredly a more expensive city in which to rent than is Baltimore. With similar house prices, that makes DC look “cheaper.”

Posted by MitchW | Report as abusive

Boston and Charleston, SC

Posted by creed223 | Report as abusive

I would be careful with regard to any number coming out of David Leonhardt’s interactive device. First, he doesn’t seem to understand that $1 today is not the same as $1 seven years from now. As future cash flows are not discounted and much of the benefit from ownership often comes from house price appreciation years in the future, this will significantly overstate the benefit of ownership. Second, he discusses the opportunity costs, but I can’t see where they fit into his calculations.

Posted by framed | Report as abusive

Nice chart! Real data! This is been one of your best blogging weeks ever! You are right about the apples-to-oranges comparison; rental housing stock is dramatically different from owner-occupied housing stock.

This also in part explains my earlier disagreement with you, Felix, regarding bubbliness of the housing market. You look out the window and see New York while I look at the DC area, which is one of the least expensive ‘vibrant urban centers.’

In our new age of American statism, the capital is the place to be, much like Moscow or Brussels or Beijing are top places to be in those countries.

Posted by DanHess | Report as abusive

Boston and Honolulu

Posted by drewbie | Report as abusive

Boston and Portland

Posted by finandistboy | Report as abusive

We do need to figure out a way to normalize for quality on rentals vs. owner occupied homes, but these ratios are a great start at affordability analysis. This chart is a little funny to me though — especially fitting a linear trend to it. In what case would we expect home prices to not be correlated to a ratio that also contains home prices?

I’ve done a little work on affordability myself. Using the FHFA home price index and the census median income and rent figures I’ve calculated price-to-income and price to rent ratios for a bunch of metros here:

http://www.deptofnumbers.com/affordabili ty/

Posted by BenE | Report as abusive