Average is Over for regional job creation

Janet Adamy writes,

in the mid-1980s, 29 metropolitan areas that contained 45% of the country’s jobs were home to half of the national increase in companies after an earlier recession.

Now look at what happened after the painful 2007-09 economic downturn. The aforementioned five metro areas [New York, Miami, Los Angeles, Houston, and Dallas] housed half of the nation’s net increase in new firms and accounted for 17% of employment between 2010 and 2014. Left behind are thousands of small towns and rural areas that stitch together much of America.

She cites a new report, Dynamism in Retreat. I recommend checking out the report, and in particular the section “How did we get here?”

I think it is important to keep in mind that this is a global phenomenon. Commenter Handle has convinced me to consider a scenario in which wealth becomes more highly concentrated in a few key cities.

In traditional economic terms, think of an economy consisting of three sectors. One sector is monopolistically competitive. A second sector is protected somewhat from competition by licensing rules. The third sector consists of new natural monopolies.

Monopolistic competition is what you see as you drive along a road with strip malls. The nail salons, restaurants, and small financial services firms operate on low profit margins, because entry is easy.

The license-protected sector is where you find medical professionals, teachers, and others where credentials are required. If you can obtain a license and you are willing to work long hours, you can make a lot of money in some of these occupations. But nothing spectacular.

The spectacular profits come from the natural monopolies. These are the software-driven firms that exploit network scale advantages.

For a would-be natural monopolist, the difference between success and failure is so dramatic that the savings that might come from setting up in a low-cost area seem trivial. Better to locate in one of the high-growth cities where the best talent can be found.

As the seekers of natural monopoly gravitate toward big cities, the licensed professionals and the monopolistic competitors follow, because that is where spending on services is high. Tax revenue is high there also, which helps to generate government jobs.

The difference between the few winning cities and the loser regions is widened by self-selection. Talented, achievement-oriented people move to the big cities, and they leave the loser regions.

This entry was posted in business economics. Bookmark the permalink.

21 Responses to Average is Over for regional job creation

  1. Andrew' says:

    I don’t consider it a fait accompli. Our cost structures were established before average was over. Now that average is over we should restructure. Until we do that (correctly) we don’t know.

  2. collin says:

    In terms of the Trump election is not just revolt against public elite, but a revolt of the rural/small voters. Read all the Real Clear Politics by Sean Trende and see all Trump’s voting growth v Romney was in small towns while all of HRC voting growth v Obama12 was in large cities. Realize, HRC won some Houston counties against Trump! Trump margin of victory in Iowa was larger than his margin of victory in Texas. Of course, I do really wonder why Iowa that has really benefited from NAFTA and China trade with export sales went so hard for Trump. I do wonder if they switch if grain prices continue continue to drop if they continue the support.

    What was clear that Trump won with voters that wanted a slowdown of movement of goods and people across borders.

    • collin says:

      The 2016 election was probably the sharpest division of Urban City Voting vs Small Town Voting since McKinley v Brennan 1896 election. (With Parties reversed.)

  3. asdf says:

    A very good summary across the board.

    “For a would-be natural monopolist, the difference between success and failure is so dramatic that the savings that might come from setting up in a low-cost area seem trivial. Better to locate in one of the high-growth cities where the best talent can be found.”

    This is true for Google executives, but not for Google workers, even software engineers making a well above average salary. Most Google workers are the engineers and not the executives.

    “Talented, achievement-oriented people move to the big cities”

    I think being a talented achievement oriented person in a big city is like entering a lottery. A few manage to hit it big, but most don’t. Those that don’t hit it big gamble away getting married and raising a family because the rent is too high and the hours too long, amongst other factors.

    The big city talent lottery system relies on a constant flow of high IQ people from the wider world, then it shreds close to half their genes a generation. This is a non-renewable resource being chewed up so that a few lucky start up executives and current owners of big city real estate can get very rich.

  4. Dave Smith says:

    Statistical illusion? Could you look at any time and make the claim “The top five cities/states/areas/whatever accounted for a disproportion of job growth?

  5. Bryan Willman says:

    How does this explain or account for Seattle? Or San Francisco and the Valley?

    What “Natural Monopolies” of any import have arisen in Miami?

    Also, in my experience, the talented don’t go to cities, they go to jobs. Then they leave those jobs to found new ventures – very often near where they happen to be because they took a job there.

    Finally, having grown up in Iowa, it’s no surprize to me that HRC would have been very unpopular there. Don’t ever consider Trump in vacuum, always remember he won running against the most disliked democratic candidate in a very very long time.

  6. Will says:

    In the same way that the industrial revolution led to an economic re-organization around cities (from countryside), should we be surprised when the information revolution leads to an economic re-organization around ‘supercities’ (from ‘cities’)?

  7. Handle says:

    I think there are two problems with the “little incentive to move, and need to be close to recruit talent” explanation of centralization.

    The real estate, tax, regulatory, and cost of living burdens of a place like the bay area are substantial enough that even big tech companies could probably save lots of money both in operating expenses and in terms of compensation needed to attract that talent. And we know those companies care about compensation because of the recent discovery of and lawsuits relayed to those salary collusion agreements.

    Also, how hard is it to recruit from a city that has a lot of top talent while simultaneously locating your HQ 50 miles away where things are a lot cheaper? And yet there seems to be a significant premium on central locations.

    And also, these companies tend to both outsource a lot to save money (so sometimes the incentive matters and they respond) and import a lot of cheaper H1-B visa holders, which both demonstrates a sufficient money-saving incentive and also shows that sometimes the person needs to be close, otherwise it would be even cheaper to pay them to work in their native country.

    In combination, this leads me to conclude that location-based money-saving incentives are significant instead of trivial, and what we’re seeing is being driven by a selection mechanism.

    Specifically, if a company can move some operations even 50 miles away from a high price place, then why not move them to the cheapest feasible place? There’s little reason for half measures, even if one is like a monopoly and isn’t facing intense competitive pressures.

    That outsourcing means that remaining jobs are selected to be the ones that, by their nature, are particularly hard to move away from a particular location.

    In some instances that might be due to some idiosyncratic feature of the geography, e.g. oil fields, refineries, pipelines, and ports nearby could anchor a local petroleum industry, or the law could arbitrarily place the capital between Virginia and Maryland.

    In other cases economies of scale and agglomeration could encourage the centralization of a sector which also draws in the talent pool related to that sector.

    But also many jobs simply must be performed close to enough other human beings that everyone is connected to everyone else through a web of distance-limiting relationships.

    If a job doesn’t have such distance-limitations regarding interactions with other humans, it will immediately be outsourced from high wage counties to the cheapest place.

    That means all the jobs left have increasingly tight ‘counterpart proximity requirements’. That is what generates the force driving ever greater centralization (and by implication also explains why median productivity is stagnant since these are necessarily increasingly ‘Baumol service jobs’ that can only be done by humans at human speed).

    Another way to say it is that the very geography of these sectoral or regional agglomeration hubs is now the major source of productive capital which enables workers to transform labor into value not by manipulating physical machines but by dint of the opportunities that can only exist in a few, increasingly centralized webs of distance-limited human dependencies.

    • foobarista says:

      The problem is the talent picks the employer, not the other way around. If, say, Netflix decides to move to North Dakota, it’ll have a heck of a time recruiting anyone, and very few of its employees would move there, as moving to a single-employer town is a massive professional and personal risk.

      • asdf says:

        North Dakota is an extreme example though. Why don’t these companies move to the suburbs and exurbs of existing big cities? They would probably reduce the commute of many of their workers who don’t live in the city, which is most of them.

        I live in Baltimore. It’s a shit city with a shit commute. Most professionals I know live in the suburbs. If you located an office in Ellicot City or White Marsh or Columbia it would be easier for most people to get to. So why downtown Baltimore?

        My best guess is that if your an executive living in Baltimore isn’t that bad. They have expensive luxury condos right on the waterfront. You can dock your boat there. The commute to a downtown skyscraper with your own parking spot is five minutes. You are about as far removed from the muck of Baltimore as anyone can get. There are five figure a year private schools you can send your kids to. Nobody making less then executive money could afford it, but that doesn’t matter if your an executive.

        So if your an executive locating the office in downtown is convenient for you. As Handle pointed out it may even help you in your personal networking with other rich folks. Getting a house out in the suburbs would hurt your networking, and there would be fewer services to spend your money on.

        I think this is driven by convenience and incentive at the top end. Most people simply move where the jobs are. Even future executives have to accept jobs where they are when they are too young to be making those decisions themselves.

  8. Ted Craig says:

    There was a piece about this in a recent issue of Governing. It pointed out that there’s little the left-behind cities can do. They all have their own Little Brooklyns, but people and jobs still move to the real Brooklyn instead.

    By the way, it’s not just talent, but capital that is in greater abunduance in these cities. VCs want their investments close enough that they can check in on them.

    One final thought: with the exception of New York, those are all warm weather cities. Michigan beats Miami when it comes to talent, but not average temperature.

  9. Octavian says:

    I don’t know if big city vs. small town is the most important dichotomy. I’m inclined to agree with Kevin Erdman’s emphasis on open access vs. closed access cities, e.g. San Francisco and New York (high living costs and restrictive housing policy account for a huge fraction of the income their, leading to their prosperity being overstated) vs. Houston or Atlanta, which have freer housing markets and lower living costs and, when costs of of living are factored in, are possibly wealthier than NYC and Boston.

    Many of the fastest growing (economically and demographically) cities are small or mid sized southern, southwestern, or mudwestern cities.

    Many tech companies are opening offices outside of Silicon Valleu and more and more people work from home remotely in more and more industries. If anything, I’m inclined to say the days of the Bay Area or any city continuing to monopolize an industry are numbered, as information disseminates and the supply of people who know how to code beyond the Valley skyrockets.

    I suspect a great deal of NYC/SoCal’s continued demand among skillled workers isn’t due to economics, but culture, or the fact that a lot more people think it’s cool to live in NYC than Dallas or Phoenix. And California also has good weather going for it.

  10. R Richard Schweitzer says:

    “locate in one of the high-growth cities where the **best** talent can be found.”

    Why is that the locus; sheer density?

  11. Charles W. Abbott says:

    Thanks for letting us play here, Professor Arnold!

    Aaron C. Renn had this blog post about Caterpillar moving its headquarters to Chicago.


    I grew up in Rochester NY when it was still basically Kodak town, with Xerox, Bausch and Lomb, Taylor Instruments, Gleason. Much of that has ebbed away, although not every last bit, and there are attempts to spin off new startups.
    Rochester can be still be a a nice place to live, but apparently Brooklyn is better. “Smugtown no more” could be Rochester’s motto currently. And the racial / class isolation of the innter city vs. the suburbs is perhaps comparable to Hartford, CT’s.

    You can learn a lot from listening to company managers who lament the struggles of trying to get talented executives and even engineers to live in Corning New York, Moline Illinois, Ottumwa Iowa.

    I’m sorry that this tends to be a digressive post. The moving of the leadership from Peoria to Chicago says something.

    Edward Glaeser’s book Triumph of the City makes some good points.

    Many cities would be nicer to live in if they could solve the poor schools / bad students issue and the crime / quality of life issue.

    You can read some of his columns at city-journal

    Steve Sailer (dont be afraid to read him!) says that migrants often don’t acculturate to the prevailing middle class norms when they move into a city. His Wisconsin article probably needs careful fact checking, but it’s alarming nonetheless.


    = – = – = – =

    Even if Sailer is not correct, you could say that there are multiple migration streams, only some of which are driven by labor markets.

    Everyone knows about the retirement migrant flows, e.g., New York State to Florida. .

    The upper middle class moves for jobs for strivers.

    Some households (especially if benefit-eligible) in the bottom two quintiles appear to move for benefits (cheap subsidized housing, short waiting lists for housing).

    College students move for school, and after finishing school for cool urban environments.

    Urban minorities often move just to get away from their dangerous neighborhoods.

    Refugee movement seem to be driven in part by sponsoring agencies. It’s hard to argue that Somalis have aggregated in the Twin Cities because of jobs. The Hmong seem to regroup in a few places. Based on my own research (never published) urban middle class Arabic speaking professionals from Sudan (not refugees from the South) were moving to Iowa City as of 4 or 5 years ago when I last checked. They were mostly asylees or green card lottery winners, not refugees at all

  12. Charles W. Abbott says:

    It would be nice to see marginal rates of substituion between being a striving google techie in New York City or California, compared to being an actuary in Des Moines or a mechanical engineer in Moline or Peoria, IL. It’s cheap to buy a house in the Quad Cities or Des Moines

  13. Lord says:

    I always find it ironic how ‘business friendly’ states and locales implement such regressive policies that all their talent wants nothing more than to leave, and yet never reconsider what business friendly actually means. Maybe right to work, the uninsured, and bathroom laws aren’t really business friendly.

    • asdf says:

      People do lots of business in the UAE and I’m pretty sure they have bathroom laws.

      Patterns of trade today mostly have to do with institutional arrangements that already existed before most modern economic trends. Cities that already had size, influence, and commanding heights institutions simply expanded on what they already had.

  14. Sure says:

    I often wonder how much of this clustering is driven by economics and how much of it is more social. For instance, say you are a very bright, libertarian Bohemian who wants to be friends with other bright, libertarian Bohemians; perhaps you even hope to form a long term marriage with such a person.

    This means that you will gravitate to places where you can find such people. In the “market” for friends or mates then there will be scale effects. Particularly for marriage I have seen many, many times where the need for a spouse to have a high paying, highly respected job highly limits the ability of the other partner to take a best cost of living arrangement.

    Consider the Google buses to downtown San Francisco, prices in the city are much higher and the commute is longer than from places like Gilroy. Yet Google has a multitude of buses headed into San Francisco and very few heading south or west. Clearly either Google or its employees have chosen to locate their operations an hour commute away from many employees and to do so in a geographically distinct fashion. Much more likely, I think, is the fact that San Francisco is the type of place where people could see themselves finding friends, significant others, and the like while the interior of CA is not. You could make an argument about amenities, but for the price differential in housing and taxes you can buy an obscene amount of amenities.

    Certainly I have known an obscene number of telecommuters who choose to live in massively overpriced cities (e.g. working for a London firm while living in NYC or DC) because it is where they are comfortable rather than what is affordable. I likewise know a large number of people who are chained to the major metropolitan areas because their family needs both a robust legal job market (for the lawyer) and a robust tech economy (for the software engineer). I suspect that the friend market has similar issues, people rarely have friendships with people terribly dissimilar these days and it is just hard to get enough jobs for all of them unless you are in a major metropolis with a robust set of jobs.

    • charles w abbott says:

      #1: Comment the First

      In _The black swan_, Nassim Nicholas Taleb discusses the expected value (in a statistical sense) of semi-random encounters with people it can be exceptionally useful to know. At the extreme, some encounters could be “positive black swans.” You are more likely to have them in New York City, he argues, than in more provincial urban areas located farther down the urban hierarchy.

      More generally, in life there is a “lottery of personal connections.”

      #2: Comment the Second

      another useful idea, different but pointing in the same direction, is what Paul Graham says about cities and ambition.

      He’s probably on to something.


      #3: Comment the Third

      Paul Graham also wrote something about why startups condense in America. A few of the variables he lists here might vary a lot between US cities. It’s an empirical question.

      I’m not sure that Graham’s essays are profound. Some of them stick in my memory because of his chatty, accessible tone.


  15. Max says:

    I see the same trend here in Germany. In my region around Stuttgart we have some of the biggest car companies in the world and all good engineering talent is gravitating to it. Of course, many companies follow leading to an infrastructure break-down. With just a Million people here, we have traffic jams the size comparable to Los Angeles (we only have probably a quarter of the road capacity).

    This has prompted amongst many people a return to rural lands, which means that in a decade or two, the region of stuttgart will cover an area twice from now.

    I dont see this to be sustainable, except for everyone becoming poorer. My question is, how do you reverse such changes. The North East of Germany is practically empty and could be a good ground for plant-surface dependent companies and yet I believe none will go there…

Comments are closed.