Interregional trade is growing much more slowly than international trade

by on October 7, 2013 at 4:40 pm in Economics | Permalink

Paul Krugman writes:

…we can measure the growth of each flow from 1997 to 2011, which covers much though not all of the era of “hyperglobalization.” And here’s what I get for percentage changes from 1997-2011:

Exports: 46.5
Imports: 108.7
Total trade (exports plus imports): 81.2
Domestic shipments: 25.6
Real GDP: 36.6

I think this makes sense: the forces behind hyperglobalization — reduced transportation and communication costs leading to vertical disintegration of production — are encouraging mainly long-range trade to save a few percent on labor costs, not shipping stuff between U.S. cities. Interregional trade seems even to be lagging GDP, possibly because our cities are becoming less specialized than they used to be. (What does Atlanta do for a living, exactly?)

Axa October 7, 2013 at 4:47 pm

Atlanta airport is the place to eat your first burrito when coming back from Europe.

msgkings October 7, 2013 at 5:21 pm

LOL! Completely accurate.

Ray Lopez October 8, 2013 at 1:56 am

They have burritos in Europe and even in Thailand and elsewhere. It turns out the cheapest most profitable food is Mexican. Low cost, popular to eat (high calories and fat tastes good), and with high profit margins.

Chris H October 8, 2013 at 2:46 am

Maybe I just missed them, but the few Mexican places I found in Europe were utter crap (this isn’t to say the US doesn’t have crap mexican, just see Taco Bell, but I can find a lot of good ones just about anywhere here). This was back in 2010 mind you so perhaps it has changed, but even Atlanta’s sub-par burritos are better than anything I had overseas.

lonely ranger October 8, 2013 at 5:52 am

decent tex-mex in london these days and even a couple of OK real mexican joints. everywhere else in europe is pretty miserable

Roy October 8, 2013 at 6:45 am

I would kill for a taco bell in europe or asia, its not really Mexican food but it is better thsn what the locals think Mexican food is. Taco Bell is better than what Europeans ans Asians think Mexican Food is. But then I am from Texas with family in NM and lots of time in Mexico, so maybe I am less appalled at Taco Bell than the typical SF style burrito place.

Steve Sailer October 7, 2013 at 4:55 pm

Krugman says:

“possibly because our cities are becoming less specialized than they used to be. (What does Atlanta do for a living, exactly?)”

Is this, on the whole, true?

For example, what Atlanta primarily does for a living seems really obvious to me. It is a major airport hub, so it attracts frequent flyers who travel on corporate business, much like Dallas-Fort Worth and Chicago do.

My impression would be that industries are increasingly clustering in single metropolitan areas and, on the flip side, that cities are specializing. For example, when I was a kid in the San Fernando Valley, there were two huge local industries: entertainment and aerospace. Now, aerospace is largely gone and entertainment reigns unchallenged.

Or, back in the 1980s, it was only moderately anomalous that the San Fernando Valley was home to the most dynamic figure in global finance, Michael Milken. In retrospect, it seems weird.

Someone from the other side October 8, 2013 at 1:28 pm

A sketchy finance guy in the mid of porn Eldorado? Not that surprising, really

bdbd October 7, 2013 at 5:12 pm

Imperfect measures, certainly, but the Atlanta metro GDP is on the order of $250 billion while the economic impact of the airport is $30-40 billion. There’s something else going on in Atlanta besides the airport. Hartsfield is the busiest airport in the world, but 65-70% of its passengers are connecting passengers who contribute marginally to the local economy

Steve Sailer October 7, 2013 at 5:26 pm

Wow, $30-40 billion just from the airport out of $250 billion is a huge fraction. Thanks.

Handle October 7, 2013 at 5:57 pm

Whoa … this can’t be right.

La Wik: “Hartsfield–Jackson held its ranking as the world’s busiest airport in 2012, both in passengers and number of flights, by accommodating 95 million passengers (more than 260,000 passengers daily)”

Is bdbd telling us that the Atlanta Metro somehow benefits $300-$400 per passenger trip? That seem ludicrous.

If we’re counting the total over a generation, or using the low discount rate to get a present value of all future GDP contributions, then, ok, that’s more plausible. But then the actual annual GDP impact is under 1%.

bdbd October 7, 2013 at 6:18 pm

A couple of points. (The numbers reported are annual numbers, by the way, but understanding the methodology behind such studies is important)

a) Airport economic impact studies are a bit unique in that they include the spending in the metro region by visitors arriving using the airport — that amounts to nearly $14 billion of the total for ATL. ( page 11).

b) Economic impacts represent transactions or economic activity, which can be much more than GDP. Also, the volume of transactions shouldn’t be regarded as a “benefit” for the region, in the way a more difficult to get at measure like “value added” could be.

Handle October 7, 2013 at 6:25 pm

Isn’t ‘impact’ vs ‘benefit’ apples and oranges then? One doesn’t count transactions to get to GDP – one adds up the values-added.

To do a real comparison, we either need,

“The Economic Impact of all the transactions which occur in Atlanta” – which would be in the trillions. Or,

“The net contribution to Atlanta GDP of the Airport” Which I’m guessing is a dozen or so bucks a passenger trip, and maybe some extra for cargo. Total Hunch: $3 Billion = 1%.

Either way, the Airport does not account for 15% of Atlanta’s GDP.

bdbd October 7, 2013 at 6:42 pm

That’s correct, which is why I prefaced my first comment with “Imperfect measures…” and concluded that there must be more going on in Atlanta than ATL.

I expect it’s more than $12 a passenger though. A rule of thumb might be to use the personal income generated in Atlanta by airport activities ($16 billion, page 2), which is 6% or so of regional GDP. Even after adjustments, ATL really is a huge operation

mulp October 8, 2013 at 1:56 pm

Passengers do most of the sorting work for free and thus do not add to the local GDP.

On the other hand, packages and cargo flowing through Atlanta require paid labor to be sorted.

Atlanta has become less economic for cargo sorting since 2005 as that volume has fallen by 25%. Logistics is spreading out into all regions to reduce interregional trade (shipping cost).

Steve Sailer October 7, 2013 at 6:41 pm

The main advantage of having a major hub airport is that it attracts businesses and workers who prefer direct flights to as much of the rest of the country as possible. Frequent flyers are, on average, affluent.

Cities with grand old business traditions that have, unfortunately, missed out on being major hubs include Milwaukee, Cincinnati, Cleveland, and other places with depressing prospects.

Gabe October 8, 2013 at 12:06 am

The Corporate Head of Southeast Regional Sales for everyone from Dunder Mifflin to Morgan Stanley is in Atlanta. Seriously lots of sales jobs in Atlanta.

Steve Sailer October 8, 2013 at 2:34 pm

Right. And sales jobs, especially the kind where you fly 100+ times per year, can bring in a lot of money.

So, Atlanta is an excellent place to put national or regional corporate headquarter offices. They may be in different industries, but the white collar workers they employ aren’t that different.

Chris H October 8, 2013 at 3:01 am

A few potential contributors.

Coca-Cola headquarters is in Atlanta, so that’s corporate management for a major multinational. The CDC is also here which is a rather large operation. The city has a number of major universities (at least three of which have 10,000+ students) which is also a contributing factor. Lockheed Martin also has significant operations in the city.

So what does this all add up to? Well if Krugman’s point is Atlanta isn’t a very specialized city he nailed that one on the head. There are major corporations and operations in the city and the metro, but it’s a bit schizophrenic. Aircraft manufacturing, corporate management, scientific research, and education all play important factors in the city’s economy. And none of them are really big enough to be what the city is “known for” in contrast to some other majors cities (like New York with finance, San Fransisco/Silicon valley with tech start-ups, Boston with universities, and L.A. with entertainment). The airport is probably the best known single thing, and it’s important, but it’s actual value added impact can’t explain more than a fraction of what’s going on.

John Thacker October 7, 2013 at 5:33 pm

When I think Atlanta I think airport, Delta, Coca-Cola, Waffle House (which is not in the city but close), CNN, SunTrust, and Chik-Fil-A (not HQ in Atlanta anymore but close.) Also Lockheed Martin, thanks to Marietta, Georgia being in the Atlanta MSA and Lockheed having merged with Martin Marietta.

dan October 7, 2013 at 6:09 pm

Also Home Depot, UPS, Napa Auto Parts, and Southern Company. Lots of big corporations based in the Atlanta area.

Steve Sailer October 7, 2013 at 6:19 pm

Atlanta was the railroad hub before the Civil War (that’s why Sherman’s 1864 capture of it doomed the Confederacy), and it always had a more pro-business environment than the rest of the South. (E.g., Atlanta business leaders were much more open to putting Jim Crow — an expensive, stultifying set of limits on maximizing profits — behind them than most other white Southerners.) Atlanta’s huge airport is a continuation of the transport hub / pro-corporate strategy that has worked well for Atlanta since before the Civil War.

Douglas Knight October 7, 2013 at 10:29 pm

Atlanta was founded by railroads.

Roy October 8, 2013 at 6:53 am


It was originally named Terminus, because it was the end of the line for The Atlantic and Western RR, and then they just renamed it after the name of the road. Atlanta was founded by the railroad, it is more like Pocatello or Omaha than other railroad towns like Chicago, St Paul, and Houston, which were at least transshipment points.

John Thacker October 8, 2013 at 2:33 pm

Most East Coast state capitals were located at the Fall Line, which matters for shipping down rivers. It’s basically where US1 goes. Atlanta is newer, so it’s railroad.

In GA, Macon, Columbus, Augusta, all Fall Line.

B October 7, 2013 at 5:43 pm

What do most major cities do? NYC clearly has finance. There’s also Silicon Valley. But are external economies of scale important enough that an industry will come to represent a sizable fraction of a major city?

Steve Sailer October 7, 2013 at 6:14 pm

It’s not necessarily economies of scale. It has a lot to do with network effects (i.e., Not What You Know But Who You Know). For example, if you want to get started in the sit-com business, you have very little chance of making it if you don’t live within commuting distance of Burbank, CA.

It used to be that Route 128 around Boston was a major locus of tech startups, but as Justin Timberlake explains to Mark Zuckerberg in The Social Network, only fools would live in some Podunkville like Boston instead of Silicon Valley.

It’s funny how the Information Superhighway hasn’t geographically decentralized the highest rungs of success at all, at least not at the entry level. (Billionaires can live wherever they feel like, of course.)

Therapsid October 7, 2013 at 7:13 pm

It’s also interesting that Silicon Valley remains the center of the tech industry after all these decades. America isn’t reproducing the California of the mid-20th century as you’ve pointed out repeatedly.

albert magnus October 7, 2013 at 8:38 pm

Houston doesn’t have much oil, but it does have lots of white collar workers in the oil business. Also, the manufacturing of oil equipment and so on. It does a lot of other things, but still that’s the base.

Roy October 8, 2013 at 7:00 am

Actually Houston has plenty of oil, Webster Field is still producing and so is offshore. These fields have been producing since the 30s and are still kind of amazing. It is also home to much of the countries refining capacity and petrochemical industries. I have lived in a lot of oil towns and Houston has as much oil in the immediate vicinity than any other major city in the US

Larry October 7, 2013 at 6:27 pm

The rapidly increasing health care share of GDP and the fact that it is much less regionalized than other industries is one reason that “cities are becoming less specialized”.

Steve Sailer October 7, 2013 at 8:49 pm

Right. But it seems like we could back out health care and a few other non-”Domestic Export” industries like that (e.g., home improvements) to see if metropolises are getting more or less specialized.

nl7 October 7, 2013 at 6:56 pm

Does it make sense for cities to specialize much? Detroit leaned heavy on autos and industry and depopulated; Chicago has a pretty balanced economy and is doing better than most rust belt cities. Not sure if it’s true, but living in Chicago I heard the ‘balanced economy’ thing a lot as a reason why Chicago wasn’t Detroit. It’s still suffering population loss to the suburbs (25% decline versus 1950, despite national growth), but nothing catastrophic like others nearby.

Steve Sailer October 7, 2013 at 8:15 pm

Chicago has been a crucial transportation chokepoint / hub since the 1830s when the canal connecting the Great Lakes watershed and the Mississippi watershed was built.

It would be interesting to study what helps a metropolitan area maintain those advantages.

For example, maybe having an airport convenient to the white collar part of the metropolis helps. For example, Chicago’s old Midway airport was built in the middle of the industrial, polluted, smelly (stockyards and slaughterhouses) southwest side. But then, O’Hare was built out in the northwest. The wind tends to blow from the north in Chicago, so the wealthy long ago moved out of the South Side for the north side. O’Hare is relatively convenient for the upper middle class.

Similarly, LAX is next to the beach, which had the least smog in Los Angeles, so it was relatively convenient for the affluent. (Numerous plans to build a giant airport in the demographically downward trending High Desert fell through.)

In contrast, Cincinnati’s main airport is way out in the sticks in Kentucky. I presume business travelers tend to live on the Ohio side, so they probably don’t find the airport’s location as convenient as they would hope in a small metropolis.

But, I’m not that familiar with airports so I only have a few datapoints.

William October 7, 2013 at 8:29 pm

Atlanta is also home to many branches of federal agencies, which is why it is often called the “capital of the South”. The HQ of the CDC is based in Atlanta. Of course, this allows for the existence of many federal government spinoff industries.

Willitts October 7, 2013 at 9:26 pm

Atlanta is asking what PK does for a living.

In a primarily services economy, most of GDP will be delivered within a geographic region. This intraregional trade becomes more important as manufacturing declines but it doesn’t mean there isn’t economic growth. I also think public goods and externalities are inadequately represented in interregional trade.

bdbd October 8, 2013 at 10:59 am

I think PK’s point, made perhaps too obliquely, was that Atlanta does a bit of everything, and is not specialized.

Yancey Ward October 8, 2013 at 12:15 pm


The Cranky Professor October 7, 2013 at 11:07 pm

I’d have thought that someone in Professor Krugman’s line of work would link “Atlanta” to “Federal Reserve Bank.”

Atlanta is a regional financial and corporate headquarters center – Charlotte tried to overtake it, but I don’t think that worked out too well.

TallDave October 7, 2013 at 11:54 pm

Wouldn’t you expect that anyway? There are fewer local regions to trade with, trade among them would already tend to be maximized, and the farther ones take more time and effort to establish trading relations with.

Rob October 7, 2013 at 11:55 pm

Atlanta also creates traffic. Horrible, soul destroying traffic. Please, make it stop.

Ryan Vann October 8, 2013 at 11:38 am

“Exports: 46.5 Imports: 108.7 Total trade (exports plus imports): 81.2 ”


Exports + Imports = 81.2 (81.2 what exactly?)
Exports = 46.5
Imports = 108.7
46.5+108.7= 81.2?


myb6 October 8, 2013 at 12:46 pm

From ’97-’11, exports grew 46.5%, imports grew 108.7%. If you track the growth of (exports+imports), it falls in the middle, leaning slightly towards imports b/c in 1997 we imported a bit more than we exported.

Cue tie-in w/ skills gap post.

Ryan Vann October 8, 2013 at 3:55 pm

I see, so this are percentages and not net numbers.

Randy October 8, 2013 at 4:12 pm

Not sure what you mean…Charlotte still has a very strong financial industry that is growing. Being a low cost city with that expertise has led many companies to shift jobs from higher salary locations in the northeast and west coast.

Energy has become another big focus, but the region’s industry mix as a whole is certainly less diversified than Atlanta.

Randy October 8, 2013 at 4:12 pm

This was in reply to “The Cranky Professor”

Comments on this entry are closed.

Previous post:

Next post: