Natural and Artificial States, and Firms

19 May 2006 at 4:28 pm Peter G. Klein 4 comments

| Peter Klein |

Among the last published papers of the libertarian polymath Murray N. Rothbard — one of my intellectual heroes — is his 1994 article “Nations by Consent: Decomposing the Nation-State.” Here Rothbard distinguishes sharply between the state, as a political entity, and the nation, a “complex and varying constellation of different forms of communities, languages, ethnic groups, or religions.” He goes on to develop a theory of appropriate national boundaries, based on the principle of volunary association and the empirical claim that people tend to associate with particular familial, linguistic, cultural, and religious groups. “One goal for libertarians should be to transform existing nation-states into national entities whose boundaries could be called just, in the same sense that private property boundaries are just; that is, to decompose existing coercive nation-states into genuine nations, or nations by consent.”

A March 2006 working paper by Alberto Alesina, William Easterly, and Janina Matuszeski, “Artificial States,” proposes several measures of the degree to which state boundaries are “natural” — corresponding roughly to Rothbard’s nations — or “artificial.” One measure identifies state borders that split ethnic groups into separate states, while another uses fractal geometry to characterize borders as straight or squiggly, assuming that straight borders are more likely to be articifially drawn and not corresponding to natural geographic or ethnic boundaries. The authors show that their measures are closely correlated with the usual measures of national economic performance (the more natural, the better).

What does all this have to do with organizations? The capabilities literature distinguishes between firm boundaries that are “natural,” or organic, and those that are artificially constructed. Natural boundaries would tend to exclude unrelated activities, would tend to change gradually, and so on. Natural firms presumably perform better than artificial ones. Indeed, one way to interpret the capabilities approach to diversification is that a conglomerate is like an artificial state (e.g., Sudan, to use an example in the Alesina et al. paper), while a related diversifier is a more natural form of organization (e.g., France).

The Alesina et al. paper uses an index of ethno-linguistic fractionalization to identify state borders that cut across “natural” groups. I don’t know the details of how this measure is constructed, but it would certainly be interesting to have analogous measures for organizations. While capabilities — that rather amorphous, latent construct already the subject of controversy on this this blog — are of course not measurable directly, it is at least conceivable that some observable employee characteristics might correspond to some kinds of capabilities. It would also be interesting to construct spatial measures of a firm’s activities analogous to the straight-versus-squiggly distinction proposed by Alesina et al.

Entry filed under: - Klein -, Classical Liberalism, Cultural Conservatism, Management Theory, Recommended Reading, Theory of the Firm.

Multi-Culturality and Economic Organization Passion, Blogging, and Manuel Castells

4 Comments Add your own

1. Jung-Chin Shen | 21 May 2006 at 7:04 pm

The conclusion of the paper is very different from one case which I know in details—Taiwan and China. Some Taiwanese economist estimate that Taiwan’s GDP per capita would be similar to the coast provinces of China had they not separated over the past 100 years (except for 1945-1949). According to my understanding of the two countries’ situations, I personally believe the result is plausible. Yes, one case cannot overrule the econometric results. But it shows one potential problem with the measure of squiggliness: it fails to distinguish between land and ocean separation. Charles Darwin seems to know this subject better. :)
2. David Emanuel Andersson | 22 May 2006 at 2:36 am

The idea that ethnic or cultural homogeneity is more productive seems like a rather simplistic application of the transaction cost approach. Rural areas/jurisdictions tend to be far more homogeneous than large cities, yet their productivity and – especially – innovative performance is almost always inferior to the large and more multicultural (fragmented) cities. As Jane Jacobs has argued quite convincingly, innovation and creativity is often the result of combination of diverse ideas that tend to be more easily perceived in heterogeneous cities, giving rise to a type of agglomeration activities known as “urbanization economies.” Indeed, many of the most succesful jurisdictions in the history of the world have been more-heterogeneous-than-average city states such as Venice, Bruges, Amsterdam, Homg Kong, Singapore as well as unusually diverse cities in relatively diverse countries such as London and New York for the last 200 years or cities like early 20th century Vienna or Shanghai.

Jung-Chin: I think your comment regarding Taiwan is somewhat beside the point. If coastal China had not been subjected to central planning, it is likely that a free flow of factors between the cities of Taiwan and the cities of eastern and southern mainland China would have accelerated the development of all regional economies, although – of course – the relative performance as opposede to the absolute performance of Taiwan may have lagged behind the Pearl River Delta and Shanghai regions.
3. Peter Klein | 22 May 2006 at 8:37 am

David: No disagreement here. In fact, I've blogged below on Jacobs externalities and how they're generally underappreciated in the clustering literature. One important distinction, though, is that the spillovers discussed by Jacobs apply to cases of voluntary diversity. Whether the same spillover benefits accrue under coercively enforced diversity, as in the case of colonial boundaries, is an open question. Another question is the extent to which Jacobs's within-city economies of diversity apply across cities, within larger territorial areas such as countries (the unit of analysis for Rothbard and Alesina et al.).
4. David Emanuel Andersson | 23 May 2006 at 9:25 am

Peter: I kind of agree, but I think we need a two-dimensional approach which takes the degree of central planning in any jurisdiction into account. My “hypothesis” would be that if we have no or or very little central planning, then the boundaries and ethnic rfagmentation don’t matter, while in jurisdictions with lots of planning and little economic freedom, conflicts are more likely generally, and also politicians may appeal to ethnic loyalties etc…..although maybe also sub-ethnic loyalties like class, religion etc.