Asset Pricing and Agglomeration Economics

By Fester

At the Creative Class Exchange, Richard Florida is promoting an interesting angle on Philadelphia's strength based marketing and economic development.  Interesting, but I think wrong or at least incomplete.

Philadelphia has a secret weapon. Housing everywhere, from the urban core to its terrific suburbs, remains affordable. The biggest challenge in the leading mega-regions of the world is escalating housing prices. Wharton's Joseph Gyourko and colleagues dub this the "superstar-city" phenomenon. Prices in other key parts of Bos-Wash (not just Manhattan, but also Boston and Washington) have skyrocketed.....

This housing-cost advantage is a huge edge for Philly's future. Philadelphia has plenty of challenges.

As disclosure, I worked and studied with Dr. Florida in grad school.  As a professor, he was great for me. He consistently challenged my thinking, showed where the holes were, and encouraged me to run with a couple of very interesting projects post-grad school.  But I think that he is getting the economics wrong on Philadelphia for two reasons.

Using the same logic Detroit or Gary, Indiana are very well positioned as they have a massive housing cost discount working to their advantage.  I think Florida is inverting cause and effect in his analysis here.  High housing costs occur in highly desirable cities as supply and demand reach equilibrium at a high point.  Less desirable cities have lower housing costs. 

Assuming that one is not in an irrational bubble assets, in a fluid and reasonably transparent market, are priced according to their long run net value.  A house's purchase price is influenced by its expected resale value (higher the expected resale value, the higher the initial price), the cost of money (lower interest rates lead to higher initial prices) and the  carrying costs of depreciation/maintenance and taxes/insurance (higher the annual costs, the lower the initial price.)  City regions have minimal control over the cost of money; isolated examples of below market rate financing are always available as incentives, but they are not prevalent.  City regions do have different growth profiles and different cost structures.  All else being equal, New Yorkers discount future prospects far less than Philadelphians and this is what produces the price differential.

As a micro-example of the different pricing regimes, is where I live.  Pittsburgh as a city region has recently experienced below national home price appreciation and it is distributed widely and unevenly throughout the region.  One neighborhood about 2.5 miles from my house has effectively null or negative home values.   It is one of the tougher and more decrepit neighborhoods in the city region.  However my neighborhood is seeing home values,based on recent sales, increase in real and inflation adjusted terms.  The prospects here are that the school system is decent, and as a cost shifting measure the neighborhood has high density, good local services and is a hell of a lot cheaper than comparable neighborhoods within city limits.  So prices are being bid up.  But I digress.

Philadelphia's relative pricing discount is only present because future assessment of economic activity in its city region is comparatively lower than the economic activity in New York, Boston or Washington.  As soon as Philly gets moving economically, that advantage collapses. 

Furthermore, the agglomeration aspect of the economics are being stretched a bit too far.  Analyzing super-metro regions such as the Bos-Wash corridor is good for broad social and economic trends, but has a scale problem. It can also be useful for comparative analysis but time/transportation problems intefere with grand scale economic development analysis. 

Local economic development tends to be either the transplanted big box of low innovation and creativity, or the more volatile, and more economically valuable creation of new or at least new to you work.  Most new or new to you work is based on locally pre-existing work and is a logical progression of something that is already being done in the local economy.  This creates clusters of related work, and more importantly, clusters of common relationships, specialized knowledge and expertise, a shared language, and massive tacit and formal knowledge sharing.  Silicon Valley and Route 128 around Boston are the two classic American examples of these types of industrial clusters in high tech, while Pittsburgh was the center of the steel/heavy metal/coal cluster which supported logical spin-offs in computer science, psychology (training and assimilation), banking, chemicals etc.

These clusters occur when there is both a social and physical mileau in which the right sorts of people can cheaply, easily and frequently meet and run into each other.  This allows for new ideas to be bounced and strengthened, techniques to be disseminated, and efficient sorting of the labor/capital markets to take place.  The research has shown that these quasi-random interactions need to be frequent and need to be varied.  This strongly indicates reasonably contained high innovation geographical spaces as most people will not travel two hours to introduce an aquaintance with the rumblings of a good idea to a potential investor.  Philadelphia's economic region in which it can exert that type of gathering and jumbling system is at the most fifty or sixty miles in diameter. The most innovative regions are much, much smaller for the high intensity change.  Once the change becomes standardized, it can go to the periphery of the innovation region or beyond. 

Philadelphia as a social mileau similiar to Boston, New York or Washington (major East Coast city) can compete as it offers a near substitute for those experiences ( or at least near compared to Columbus, Ohio et al) but the housing costs are a discounting indicator of where people believe Philly will be relative to the other near peer competitors and the high density social mileau needed for innovation can not by its character stretch up to Scranton or down to Baltimore so its existence within Bos-Wash allows Philly to contribute and benefit to super-regional externalities and infrastructure, but not to neccessarily gain from near-peer changes.