Good old central place theory.  Here's a Vox essay that may be missing something.
People with similar skills and similar levels of education make a lot more money in New York and San Francisco than they do in St. Louis or Cleveland. You might expect these differences to even out over time, as workers relocate from low-income areas to high-income areas to take advantages of the opportunities there.
To a first approximation, they do. But net of adjustment costs, and subject to the bid-rent curve, the differences might already have evened out.
But that hasn't been happening. Over the last half-century, income differences between metropolitan areas have actually been increasing. One study estimated that this effect is costing the country hundreds of billions of dollars in lost income every year.

It's not hard to figure out why people aren't moving. While the wage difference between St. Louis and San Francisco is large, the difference in housing costs is even bigger. A programmer in St. Louis might get a big raise by moving to San Francisco to take a job at a technology company there, but he might still be left worse off thanks to the much higher rents there. High housing costs make it hard for companies in high-cost areas to attract workers, stunting the growth of some of America's most dynamic industries.
So far, so good. But what, apart from geographic constraints, might be contributing to disequilibrium?  (Left to the reader as an exercise: is the New York area so powerful a Marshallian industrial district for finance, or the agglomeration economies in Silicon Valley so significant, that Chicago or Minneapolis or Sioux City might as well forget about any expansion in finance, or much of flyover country just empty out?)
How much does this cost the economy? In a study released as a discussion draft last year, economists Chang-Tai Hsieh and Enrico Moretti tried to figure that out. They imagine a world where the housing markets in America's most productive cities became more elastic, meaning that increasing demand led to more construction rather than higher rents.
That's not as easy as it looks.  Brad DeLong picked up the abstract to an earlier version of the paper (now replaced?)
Differences in the nominal wage across cities reflect differences in the marginal product of labor across cities which, ceteris paribus, lower aggregate output. We show that the dispersion of the average nominal wage across US cities increased from 1964 to 2009 and may be responsible for a 13% decline in aggregate output. Changes in amenities appear to account for only a small fraction of this output loss, with most of the loss likely caused by increased constraints to housing supply in highly productive cities. We conclude that welfare gains from spatial reallocation of the US labor force are likely to be substantial.
Perhaps, although the losses to vested interests are also likely to be substantial.
For the last couple of years, San Francisco has been erupting with periodic protests aimed, rather imprecisely, at a nexus of grievances related to gentrification, affordable housing, transportation, the tech industry, newcomers to the city, its changing skyline and Silicon Valley to the south. The city is screaming, although at what its protestors seem a little confused.

"In my view, the whole debate here misses the point," says Enrico Moretti, an economist at the nearby University of California at Berkeley. "People aremarching against Google buses when they should be marching for more housing permits."

At the root of San Francisco's tension is a mismatch of supply and demand: Affluent workers have been flocking to the area for its tech jobs, but as the number of jobs in the region has grown, the number of housing units to accommodate people taking them hasn't remotely kept pace. As a result, rents are going up. Low-income residents are pushed out. Landlords who see more lucrative opportunity in condo conversions have ramped up evictions.

"Once I started seeing what was going on in the San Francisco public debate, I got appalled by the lack of understanding of basic economics among the general public, the protesters," Moretti says. "And it’s even more problematic among policymakers."

The culprit here isn't really the tech industry. It's much-harder-to-protest land-use policy. And from Moretti's point of view, the rest of us should care about how San Francisco and big cities like it restrict new housing because the economic repercussions of such local decisions stretch nationwide.
Perhaps so, although that's why land rents adjust.  Professors Moretti and Hsieh argue that existing constraints on land use contribute to the rising land rents in some cities.
If housing wasn't so expensive in coastal cities, a lot more people would move to New York, Washington, Boston, Seattle, and the San Francisco Bay Area. The data suggests that — even after controlling for factors such as education — workers in these cities are more productive than in other metropolitan areas. The study doesn't try to explain these productivity differences, but possible explanations include better infrastructure, opportunities to learn new skills, and a culture that encourages entrepreneurship.

Hsieh and Moretti estimate that moving American workers to higher-productivity cities could increase the income of Americans by a stunning amount: more than $1 trillion. That amounts to a raise of several thousand dollars for every American worker.
I'll have to locate the working paper and look at it carefully, as there might be agglomeration economies at work to further boost the value of living in those already crowded cities.  Perhaps, though, in addition to relaxing the constrains on relocation to those cities, there are unexploited gains from trade in working around those constraints.  For instance, might a company based in Rockford, Illinois, be better off paying a compensating differential to get people to locate there, rather than paying a higher salary that reflects Chicago or San Francisco or New York costs of living?

No comments: