28 Jul Anarchy and Development
Randy Barnett at Volokh Conspiracy mentioned today a new paper, “Anarchy and Development: An Application of the Theory of the Second Best,” 2 Law and Development 1, article 4 (2009), by Peter T. Leeson and Claudia R. Wilson. Professor Barnett says he hasn’t yet read it, but the abstract looks interesting. It intrigued me enough to download it (it’s a Bepress journal) and read it this afternoon. It’s fascinating and insightful, and if you do development, development finance, law and development, I recommend it. Here is the abstract:
Could anarchy be a constrained optimum for weak and failing states? Although a limited government that protects citizens’ property rights and provides public goods may be the first-best governance arrangement for economic development, among the poorest nations such “ideal political governance” is not an option. LDCs face a more sobering choice: “predatory political governance” or no government at all. Many predatory governments do more to damage their citizens’ welfare than to enhance it. In light of this, we show that conditional on failure to satisfy a key institutional condition required for ideal political governance—constrained politics—citizens’ welfare is maximized by departing from the other conditions required for this form of governance: state-supplied law and courts, state-supplied police, and state-supplied public goods. Since departing from these conditions produces anarchy and fulfilling them when government is unconstrained producers predatory political governance, anarchy is a second best.
This is largely counter to the conventional wisdom because the conventional assumption is that anarchy is the most predatory social ordering – the Hobbesian state of nature and all that. Certainly it has how I have tended to see things, at least with respect to development issues. This paper winds up suggesting that a better way to understand things is that the worst social (dis)ordering is a “predatory” government that has sufficient resources, including sufficient political control, to be able to inflict huge damage on a population, especially in development matters, such as long term investments in many public and private goods – but it is able to use that control in a predatory way. A well ordered state would be great, but, the authors suggest, predatory governance is actually worse than anarchy. Hence the appeal of the second best solution.
I do not say I am entirely persuaded, although the argument has strongly appealing parts. Anarchy in this discussion is a somewhat slippery term, and it has an ability to appeal to different people for different, inconsistent reasons. I think I’ve seen both predatory government up close, as well as places lacking of any real governance structure in which marauding gangs worked their will. It has not always been clear to me which is worse – although those situations were often ones of civil war, which raise questions very different from those of development, even if in real life they are often intertwined. The urgencies of order in the midst of endemic war and armed conflict are different from those of a predatory state diverting and deterring resources needed for long term development and investments, public and private, therein. But this is an outstanding, provocative paper, and I commend it to readers interested in these areas.
(I would also refer to a largely unnoticed book by Frank Fukuyama, written while everyone else, me included, was focused on the war on terror – he was researching and writing very important things on development and governance. His book State Building: Governance and World Order in the 21st Century (2004) offers a schema that bears on the remarks in this paper; government, he suggests, should be seen on a grid with “strength” of governmental power on one axis and “extension” or “reach” of government on the other. The best governments, he says, are those that are (a) limited in their reach, but (b) strong within those limits. He has raised this as a point against some forms of small government theorists who want weak government. On the contrary, he says, weak government is disastrous, just as unlimited government is disastrous. And on that matrix, he observes that a fundamental problem of the Chinese party dictatorship of China is that China’s central government is at once unlimited but weak. That matrix has important bearing on the institutional problems of developing states.)
Response…
We’ve come across a similar dynamic in Peru in terms of CSR, where the centralized strength of the Fujimori regime has given way to the problems of developing effective government infrastructure in rural regions far from the capitol. Peru has adopted a series of laws to address the problem of Decentralization, which in our research has been largely influenced by mining projects. In a nutshell, the mines located in rural areas pay taxes to the central government, which is then unable to redistribute those taxes to appease rural area farmers whose local environments and markets have been impacted by the presence of the mines. The incentive for CSR is that the locals have blocked access to the mines and taken other actions that interfere with extraction, so the companies must take measures to continue to operate. The development issue is how to make satisfactory contributions and infrastructure in the local areas that will support at least the 20+ year lifespan of the mine.
This does not seem likely, unless the constraint is so strict that essentially only anarchy is left. (If you optimise over a domain that contains only anarchy, then anarchy is going to come out as your constrained optimum.)
In a situation where at least a vaguely stable dictatorship is possible (i.e. one where the average dictator survives for at least, say, 5-10 years), I’d imagine Mancur Olson’s optimal dictator model suggests that the dictator would be preferable to anarchy.
(Basically Olson’s idea is that the more stable a dictatorship, the more the dictator takes a long term perspective, the more he has an incentive to optimise total welfare, and so the more the outcome comes close to a non-dictator level of welfare.)
I should add that the paper is a sophisticated model, and I’m far from doing justice to it here. I have always thought – and imagine someone has said in detail in the institutional economics lit – that a problem with Olson on this is that signaling to the dictator turns, at the margin, into Kim Jong Il, or Stalin – the Autumn of the Patriarch, in the end – within its own enclosed shell.
Peru is a good case for the paper’s thesis. I spent time there just ahead of the rise of Sendero, in the late 70s, and again under Fujimori, and afterwards. Fujimori was the predator state par excellence, in the devvelopment sense, but it is also the case that the weakening of the predator state has also led to a bad weakening of the state even in matters and jurisdictions in which it needs to be strong.