Alchian-Allen and Agglomeration: Explaining Economic Inequality and Nice Golf Courses in Vegas
The Alchian-Allen Effect is one of my favorite economics party tricks. The theory is that whenever there are two versions of a product, one of which is more expensive, any increase in the fixed cost of the products will cause consumers to substitute for the most expensive one.
To take the classic concrete example, suppose a normal orange in Florida costs $0.50, and a great orange costs $1.00. The exchange rate between good oranges and bad oranges is 1:2; you give up two okay oranges for every good one you consume. Now, imagine that it costs $1 to transport any orange to New York City. The costs are $1.50 for the okay oranges and $2.00 for the premium variety, for an exchange rate of 3:4. Since the exchange rate is so much more favorable, New Yorkers will tend to consume mostly good oranges. So if you want a good Florida orange, stay away from Florida.
Destined to be eaten thousands of miles from here.
Like all economic models, it’s true if you grant the assumptions and only useful if you question them. For example, it assumes oranges are some sort of abstraction with well-defined gradations in quality. It also assumes that the travel process doesn’t affect quality; a robust model of orange consumption would note that shipping oranges will likely degrade their taste and texture. But the model gets its revenge: suppose there’s an expensive climate-controlled transportation process that does a great job of preserving quality. Alchian-Allen tells us that this process will be used disproportionately for the fancy oranges, so once again, New York wins on quality but loses on cost.
As a general model, it’s quite powerful: cities with high fixed costs tend to import the most valuable version of whatever it is that they import. The restaurants in New York, London, and Tokyo are unusually fancy; you can get Italian food anywhere, but you only get Carbone in New York.
You can see this more clearly if you look at cases where a given location has one particularly expensive input. In Las Vegas, water is not cheap. So the cost of a crummy golf course or a fancy golf course will be elevated — but in percentage terms, the fancy one’s cost gets elevated less. From a resource-usage perspective, a great golf course seems wasteful, but from an economic perspective an anything-but-great golf course is profligate.
In many large cities, the universal cost-increase constant is the price of real estate. And high real estate costs exert a relentless upward pressure on the fixed cost of everything else, which makes them a relative subsidy for the best version. This interacts with agglomeration effects to produce regional industries: if you’re building a financial services company, New York is an expensive place to start it, but hiring from a non-New York labor pool is an even more expensive decision. And finance is a global industry; whatever you do in one location can scale globally.
This cycle is self-reinforcing: when there’s a profitable industry in a city, it raises real estate prices in that city, which is a tax on everybody but a steeper tax on people who aren’t as skilled. So over time, several things happen:
Non-leading industries get priced out
The skill/comp cutoff within successful industries ratchets up
More marginal industries have to adapt — either find a way to be a proxy for the main industry, or build your business around having a high cost structure and commensurately higher revenue, but accept that your economics won’t be extraordinary.
This geographic concentration gets further skewed by culture: when we talk about an industry, it’s easier to handle the abstraction if we talk about a specific place. Everyone knows what business Wall Street, Hollywood, and Detroit are in. And this causes people who care about that industry to show up even if they don’t have a job offer. It’s a standard story that someone who wants to make it big in movies will move to LA and start waiting tables and going to auditions. To compete with LA in movies, you need your own supply of thousands of naive young people. And to do that, you have to have already solved the problem you’re trying to solve.
(This is why the other big centers of moviemaking — Bollywood and Nollywood — are so far away, both geographically and culturally. Bollywood and Nollywood economics are partly defined by the fact that they’re emphatically not competing with Hollywood. Different stars, different financial structures, different distribution channels.)
This dynamic explains why globalization has paradoxically made knowledge industries even more concentrated: to compete with New York, you actually need New York level rents to price out non-New York-ready talent.
As industries mature, the effect of wages on real estate tends to get nonlinear. Since real estate is a fixed cost, you can have a situation where moving to NYC is an even trade if it raises your after-tax income by $20k and your rent by the same amount. But in percentage terms, that means your rent has gone up. So we’d expect to see that in cities with successful industries, real estate is not just expensive, but disproportionately expensive compared to salaries.
This, however, creates some vulnerabilities. As rents rise, it places a higher premium on certainty: if the market bid for a San Francisco apartment is set by entry-level workers at top tech companies, it means that people earning lower take-home pay have to either spend a lot of time commuting or accept a negative personal burn rate.
The way this plays out will depend on how compensation in an industry is structured: in finance, where the variable component of compensation is typically paid out annually and where startup times are quick, you’ll still see people quitting to strike out on their own. In the Bay, where the variable part of compensation is made up of stock options with an uncertain and long-delayed payoff, higher rents are a more direct tax on startup formation.
So the net result of an Alchian-Allen based analysis of industry centers is that New York can retain its dominance in finance for a long time, while the Bay is at risk. In the short term, this won’t be visible in aggregate data. In fact, a decrease in the number of startups will make legacy companies look even better: they’ll have fewer competitors on the revenue side, and less trouble retaining their employees. But those large companies were startups not too long ago, and the entire ecosystem relies on new companies showing up. When the big success stories are increasingly in places like LA or Provo, it’s a warning sign: venture capitalists don’t want to get into a bidding war with Big Tech over real estate prices, and they don’t want landlords to earn more from their investments than their LPs do. So they’ll leave.