Are isolated capital cities worse?

From Filipe R. Campante, Quoc-Anh Do, and Bernardo Guimaraes (pdf):

We show empirical evidence that non-democratic countries with [geographically] isolated capital cities display worse quality of governance, as measured across many diff erent dimensions. We provide a framework of endogenous institutional choice that accounts for this stylized fact, based on the idea that autocratic elites are constrained by the threat of rebellion, and that this threat is rendered less eff ective by distance from the seat of political power. Broader power sharing (associated with better governance) means that any rents have to be shared more broadly, hence the elite has less of an incentive to protect its position by isolating the capital city. Conversely, a more isolated capital city allows the elite to appropriate a larger share of output, so the costs of better governance for the elite (the rents that would have to be shared) are larger. In equilibrium, a correlation between isolated capitals and misgovernance emerges as a result. The framework yields additional predictions on the size of the income premium enjoyed by capital city inhabitants and on the level of military spending by ruling elites, which are also supported by the evidence.

The title is “Isolated Capital Cities and Misgovernance: Theory and Evidence.”

As I wrote yesterday, Naypyidaw!  And yet there is more.  Across the fifty American states, isolated capital cities are more corrupt:

We show that isolated capital cities are robustly associated with greater levels of corruption across US states. In particular, this is the case when we use the variation induced by the exogenous location of a state’s centroid to instrument for the concentration of population around the capital city. We then show that different mechanisms for holding state politicians accountable are also affected by the spatial distribution of population: newspapers provide greater coverage of state politics when their audiences are more concentrated around the capital, and voter turnout in state elections is greater in places that are closer to the capital. Consistent with lower accountability, there is also evidence that there is more money in state-level political campaigns in those states with isolated capitals. We find that the role of media accountability helps explain the connection between isolated capitals and corruption. In addition, we provide some evidence that this pattern is also associated with lower levels of public good spending and outcomes.

Trenton Makes the World Takes” — not exactly!


So Illinois would be less corrupt if its capital were Chicago rather than Springfield? Counterintuitive. But possible.

It comes down to whether Illinois is better or worse governed than Chicago. If worse, then the result is completely intuitive: Illinois is governed by Chicagoans behaving more badly than they would at home.

In other news, newspapers in Berlin, Germany, report that the city's new "Berlin Brandenburg International" airport, whose (already once postponed) opening date of June 3, 2012 was postponed again in early May 2012 to March 2013 due to problems with fire protection in particular and "chaotic planning" in general, is unlikely to open before 2014.

There are no reports, however, of reconsidering the decision of not allowing larger passenger jets to land at Eberswalde-Finow airport 55 km from Berlin (as, among others, suggested by ryanair) on the grounds that the airport may be suitable, but its opening to commercial traffic would be incompatible with the goals of the regional airport planning in the Berlin-Brandenburg development plan.

Damn you reunification.

Civilization players figured this out over a decade ago.

+1, The obvious solution is to change your government type to Democracy.

actually Fundamentalism (on Civ 2) was the best form of government


I always enjoyed Joseph L. Bruno stadium.

Talking of which, are the people involved so clueless that they don't get how creepy it is to name stuff after politicians who are not only still alive, but are still in office? I can't imagine how people don't cringe when they see Frank Lautenberg Rail Station, or all that stuff in North Carolina that was named for Strom Thurmond while he was still in the Senate (I guess in that case no-one realized how long they'd have to wait).

(I meant South Carolina)

I will never forgive the USS Carl Vinson. It should be illegal to name any government work after a living person.

The authors have never been to Rhode Island

Oh come on the State House is at least ten miles from Warwick, how remote can you get? Or maybe I should rephrase that, if they moved the capital to Westerly, Providence would actually get worse?

US state capitals, in imitation of DC, are more often in remote burgs, so this is terrible place to look for examples. And in some cases it has to be an improvement, Baton Rouge has to be a better place for a capital than New Orleans. Maybe the real lesson is that planned capitals are always a disaster (looking at you Austin, well maybe not) along with capitals located by bribery by civic boosters (Olympia for example, getting closer) or capitals picked by people who think the "Big City" is a horrible corrupter (Pierre or Sacramento, a lot warmer) are indicative of a horribly dysfunctional political culture.

Also a lot of state capitals were placed in the middle of nowhere as a direct attack on the metropolis of their state, Albany being the ultimate example, but even Pierre is an example of this. In some states that Metropole was able to vanquish its fows abd rectify this, Boise's capture of Idaho's capital from Lewiston is an example, while in other states this has repeatedly failed, Juneau is possibly the most ridiculous state capital in America and their have been repeated efforts to change it. The states that are capable of avoiding such a bad initial choice or correcting it are possibly inherently more politically healthy.

Back in the '70s there was a successful state-wide vote to move the Alaska capital from isolated Juneau with its ultra-terrifying airport to an unbuilt area northwest of Anchorage, Willow, I think. Anyway, after returning form the trapline the following winter I went directly to the airport and asked to buy a ticket to the new capital. This produced much laughter from the station agent. And would even today. Local Juneau politico Bill Ray was able to stymie any attempt to move the center of administration and reduce Juneau to the remote minor league fishing port that would ordinarily have been its destiny. So much for democracy.

Ultra-terrifying airport? I flew into and out of Juneau in 2010 and I do not recall any terror in the process.
I do agree Juneau is a bit of a bore, though it's hardly the only "capital village". Montpelior VT is so small you'll miss it if you blink. Nor are Topeka, Pierre or Dover happening metropolises.

By the way, with the exception of Juneau, is any state capital today truly isolated? It's not exactly arduous to drive from Chicago to Springfield, Philadelphia (or Pittsburgh) to Harrisburg, Louisville to Frankfurt, or Detroit to Lansing. and with electronic communications even Juneau is not cut off from civilization.

The paper on US state corruption is laughable in its simplistic assumptions. Most all state capitals were established when the state entered the union and this took place prior to much of the economic development. Springfield IL was a reasonable location for a capital of a largely and sparsely populated agrarian state. Chicago did not really come alive until a number of decades later as a manufacturing and transportation hub. To compare Carson City NV with Salt Lake City is just plain dumb because NV has legalized gambling and UT does not (and of course UT is largely a theocratic state anyway with strong moral values). I could go on and on but the bottom line is the authors take a lot of data throw in some equations without really questioning assumptions. I could do the same with reams of baseball data and show that Bobby Bonds is the greatest player of all time (neglecting the underlying PED use of course which is not found in the numbers; it's exogenous and important information). Junk science with a capital 'J'.

"Most all state capitals were established when the state entered the union"

Except for Alabama, California, Delaware, Georgia, Illinois, Indiana, Iowa, Louisiana, Maine, Michigan, Mississippi, Missouri, New Hampshire, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Vermont, Virginia, and West Virginia, and arguably Connecticut and Rhode Island (they had mulitple capitols and then chose one).

In your example of Illinois, the first capitol was actually Kaskaskia and then Vandalia before Springfield.

Typo: 'capital', please. Ooops.

Fair critique but given the data set that these authors are working with the original argument still stands.

Well children, the sign has been there a long time and Trenton used to be an industrial city. Besides, I believe it irritates Paul Krugman and that's reason enough for keeping it.

Once thing I might suggest for state capitals, when I think of it it also seems to work for nations, are those states who colocate their capital and chief university and those who don't.

Wisconsin politics are made far more vituperative by the coloaction of capital and University than say Oregon whose are quite separate. This might also pply to countries. France shares a capital with its greatest university, while England and Germany do not. This must have some effect.

I thought Humboldt University in Berlin was, for most of its history, Germany's best university. Is that not so?

Nein. Germany is very pluricentric in such regards.

Göttingen, Heidelberg, Leipzig, Mainz ... there is no such thing as "the best German university", AFAIK.

The mountain is high and the king is away

This study uses statewide federal corruption convictions by all public officials in the state as the benchmark for overall corruption. Here is an example of why the data are skewed. In South Dakota, there are many convictions for corruption on the reservations, and hardly any charges, let alone convictions, for corruption at the state level. Despite this, the report lists that the capital cities (and presumably the state) are corrupt.

I'm surprise that no one have brought up Brasília, the capital city literally carved out of a jungle by the the Brazilian government back in the '50s. It's hundreds of miles from anywhere.

Very interesting. I did not know South Dakota was as corrupt as Illinois. As a near-native of Springfield, Illinois (lived there from age 3-18; parents still do), I fully support the capital moving to Chicago, or alternately, splitting the state up. As a resident of California I'm somewhat surprised at the low measure of corruption, but I suppose such is just one aspect of malgovernance. Anyway, I additionally support moving the capital to LA, or splitting the state up.

Also, states (and polities generally) could be usefully redrawn, centered on metro areas.

Most of these capitals were established when communication was a person-to-person affair. Government officials, elected and appointed had to either meet in person or communicate by the written word, delivered by another human. This is no longer the situation. In the case of the US federal capital, the biggest improvement to the government would be for legislators to remain in their home districts, easily available to their constituents and not concentrated in one target area for lobbyists. Even in present-day DC federal officials communicate via telephone, FAX, email, etc. There's no logical reason for them to have to meet in a single location as was the case in the 18th century. DC could be an administrative center but it should no longer be the meeting place of the US Congress.

I have experience to the contrary. Even heavily-IT'd industry branches like software development benefit enormously from colocation. The informal channels between people are more important than they look.

One of the most important channels is common smoking at the balcony or in front of a building ... as a lifelong nonsmoker, I was always baffled how many interactions actually take place in smoking pauses. A civilization which stamps out smoking may actually pay for that with some efficiency.

I wonder how much this analysis is swayed by NY state with population centroid of NYC.

Comments for this post are closed