Sunday, January 14, 2007

Superstar Cities week (and Galveston rail)

This week I'm going to focus on a very interesting academic paper by three authors from Wharton and Columbia titled "Superstar Cities". There's just too much here for one post, so we'll spread it over the week. Today, I'll focus on interesting excerpts so you get a feel for the themes of the paper. Tuesday I'll give my thoughts and describe the larger system of urban forces I think they're describing, and Thursday we'll explore possible solutions. If "superstar city" doesn't sound to you like something that needs a "solution," just wait and you'll see what I mean.

Before we get started, if you didn't see Alan Kolodny's op-ed on Galveston rail in Sunday's Chronicle Outlook section, be sure to check it out. It's related to this post of mine from a few weeks ago. Although I'm usually not a fan of commuter rail, this route has a lot of potential for five reasons:
  1. local transit at both ends to get you to your final destination
  2. highly populated corridor with heavy traffic flows both directions
  3. regular congestion on the existing freeway
  4. tourism potential in addition to commuters (better overall utilization for the capital cost)
  5. existing tracks that make the cost much more reasonable
Getting back to superstar cities (their definition: "some cities and towns have turned into scarce luxury goods, which we label superstar cities"), here's the somewhat academic abstract:
Differences in house price and income growth rates between 1950 and 2000 across metropolitan areas have led to an ever-widening gap in housing values and incomes between the typical and highest-priced locations. We show that the growing spatial skewness in house prices and incomes are related and can be explained, at least in part, by inelastic supply of land in some attractive locations combined with an increasing number of high-income households nationally. Scarce land leads to a bidding-up of land prices and a sorting of high-income families relatively more into those desirable, unique, low housing construction markets, which we label “superstar cities.” Continued growth in the number of high-income families in the U.S. provides support for ever-larger differences in house prices across inelastically supplied locations and income-based spatial sorting. Our empirical work confirms a number of equilibrium relationships implied by the superstar cities framework and shows that it occurs both at the metropolitan area level and at the sub-MSA level, controlling for MSA characteristics.
OK, for those of you out there who don't use "skewness" and "inelastically" in your everyday language, hopefully some of these excerpts will help clarify (highlights mine).
Over time, the gap in house prices between cities can keep increasing. Even if each family individually is willing to pay only a fixed premium for a location, when the absolute number of rich families in the country increases and their incomes rise, there are more families who will pay a higher premium for the same perceived difference between cities. Thus, a changing composition of residents in supply-constrained cities toward higher income families supports the growth in land prices.
Indeed, the inelastically supplied city need not be more desirable on average than any other city, nor do workers in the city need to enjoy increased productivity by moving there. As long as the city appeals to a large enough clientele of families, a growing price gap and a shifting of the local income distribution to the right will occur. We label metropolitan areas and towns where demand exceeds supply and supply growth is limited, “superstars.” These superstar cities and towns are ones in which residents are willing to pay a premium to live and into which high income superstar-earners disproportionately sort. These markets do not allow for increasing density through construction, cannot infinitely expand their borders, and have few close substitute locations.
Consistent with the model’s predictions, once an elastically supplied city or town “fills up” and future supply growth becomes limited, price-to-rent ratios increase, and income sorting and house price dispersion accelerates. A number of jurisdictions in our sample undergo this transition into superstar status. For example, Los Angeles and San Francisco are the first major metropolitan areas to become supply-limited, and that happened between 1960 and 1980. The areas around Boston and New York followed between 1970 and 1990.
The superstar cities model suggests that rising incomes in a city are due to a changing composition of families within superstar cities from an influx of highly productive high-income workers, rather than through gains in existing residents’ productivity.
In this framework, house prices do not rise in superstar cities because there is increasing value from amenities or productivity benefits. Instead, the composition of families living in superstar cities shifts to those who are willing to pay more as high-income families become more numerous.
...most existing superstar cities could expand supply by increasing density, but choose not to. The political economy behind that decision is only just beginning to be studied.

In addition, this dynamic has profound implications for the evolution of urban areas because it implies that even large metropolitan areas might evolve into communities that are affordable only by the rich, just as exclusive resort areas have done. Is such an MSA sustainable, or does it lose the vibrancy that makes it unique? Should public policy ensure that living in a particular city is available to all families or, since superstar cities and towns are like luxury goods, is it reasonable that low income workers can no longer afford to buy homes in superstar cities?
There are also a lot of very interesting graphs worth perusing in the back.

That's enough for today. To be continued Tuesday night...


At 2:50 PM, January 15, 2007, Anonymous Anonymous said...

Re: Alan Kolodny's op-ed

The biggest problem of including Kemah in these Houston-Galveston heavy rail links is that there is no rail bridge across the Clear Lake channel at Kemah. At least according to the satellite photos on and my personal memory. I doubt you could convince the local boating community to suffer a draw bridge or the project would cover the costs of a raised train bridge.

Also the existing train route that travels near Hobby airport continues south to Pearland and Alvin. There is no existing rail link over to Johnson Space Center etc... But if you were going to prettiness, this would be the least ugly route because it goes through more countryside and only about half as much industrial.

The only viable track is the one that runs along highway 3 which does not come very close to Hobby airport, but does go up to downtown. It travels within a mile of JSC. The only reasonable mass transit option from this line to Kemah would be some sort of water taxi or bus service.

At 2:59 PM, January 15, 2007, Anonymous Anonymous said...

Although I think I know what they mean here, I always hate it when I see some variant of the phrase "demand exceeds supply". A bunch of Wharton scholars should know better.


At 3:42 PM, January 15, 2007, Blogger Tory Gattis said...

Alan and I discussed these exact issues. I advocated the Texas 3 route with a circulator shuttle to JSC, Kemah, and League City. A water taxi would be pretty cool if it's viable. I don't think Hobby would be an option, other than maybe connecting to some normal bus route. Nor do Alvin or Pearland work very well.

At 8:16 AM, January 16, 2007, Anonymous Anonymous said...

"once an elastically supplied city or town “fills up” and future supply growth becomes limited, price-to-rent ratios increase, and income sorting and house price dispersion accelerates"

Why don't rents rise as well? If the observed price increase is due to perceived desirability of the location rather than simple speculation, shouldn't rents be climbing in proportion?

At 8:29 AM, January 16, 2007, Blogger Tory Gattis said...

Here's their explanation from p12. R is a superstar city, G is not.

"This prediction assumes that expected long-run risk-adjusted returns to investing in housing in the two cities are equated by the market, where rent is analogous to the dividend paid by a house. If workers anticipate that rents in R will grow faster than agricultural rents, they will bid up house prices so that the overall return for a house in R is equal to the return on a house in G. If R has a faster
growth rate of rents than G, that market will also have the same faster stationary growth rate in
prices. Priced fairly, a house in R earns a lower current yield (lower rent/price ratio), but a higher
future capital gain due to the faster rent and price growth than in G. Thus, when the number of
high-income families is growing, the price/rent ratio should be higher in R than G."


Post a Comment

<< Home