Harry Binswanger ended an article recently by off-handedly rehashing the “argument from Somalia” anarchists so often hear. Libertarians have responded to this argument already, both academically (see Notten 2003; Coyne 2006; Leeson 2007; Powell et al. 2008; and Leeson et al. 2009) and popularly (see Maccallum 1998; Davidson 2001; Bigwood 2002; MacCallum 2007; Powell 2009; Knight III 2009; Carson 2010; Barker 2011; Grimmett 2013; etc.), but since folks seem intent on continuing to use it I wanted to respond myself, so I can send anyone I talk with about it here.
I will begin by stating my respectful disagreement with Yumi Kim and any similar treatments people may point me too, which in my opinion erroneously romanticize Somalia. I will take a negative rather than a positive position, i.e. merely explaining why I reject the argument that Somalia demonstrates the “unworkability” or “impossibility” of anarchist ideas, rather than using it as a positive example of a well functioning stateless society.
The argument takes the form: “In Somalia statelessness led to extreme violence and poverty. Therefore, in every possible case, statelessness will lead to extreme violence and poverty.” Since this invokes inductive, not deductive, logic, it technically makes no attempt to “prove” the “impossibility” of a peaceful stateless society. It merely tries to provide evidence for the extremely high improbability of such a society.
Perhaps I expect too much, but if my reader cares to visit their nearest university and ask someone in the anthropology department whether or not “stateless societies” have ever existed, I would expect them to answer that, of course, many such societies have existed, some violent, some not. David Graeber has stated that thousands of stateless societies have existed, and that “for every case of Somalia there are probably twenty where state authority breaks down and people do not start killing each other”. Steven Pinker argues, in contrast, that stateless societies have historically had far more violence than societies with well-functioning states, (see Pinker 2013 and “Lord Keynes” 2013,) and Mark Cooney has argued that statelessness and “tyranny” both tend towards violence, (while pointing out a couple of exceptions like the Semai where people lived in peaceful, autonomous communities,) and that relatively weak government has done the best job of keeping violence at bay. Perhaps statelessness tends towards violence, perhaps not, perhaps this bears on the moral dispute involved, perhaps not. Regardless, stateless societies have existed throughout human history; a good inductive argument needs to examine more than a single example, in my opinion.
Now, what of the premise? We can’t merely provide statistics on Somalia’s impoverishment under statelessness, this would give us data but no correlation across time; to establish a causal connection, (which we need to do to argue that future attempts at creating stateless societies will end in chaos,) we must compare Somalia under government to Somalia under statelessness. Peter Leeson gives these statistics comparing Somalia under Siad Barre’s dictatorship to Somalia after the collapse of his regime, when it fell into a period of statelessness.
I shall refer readers wondering about the ‘?’ for GDP to Leeson’s paper for explanation. In short, he argues that, for multiple reasons, the data likely overstate pre-collapse GDP and understate post-collapse GDP, to degrees difficult to calculate. In regard to the two measures which worsen, Adult Literacy Rate and Combined School Enrollment, Leeson argues that these reflect the withdrawal of foreign aid financing education pre-1991 more than the shifting of educational burdens from state to community. For those curious to know more about how education fared in Somalia post-collapse, Samatar 2001 and Abdinoor 2008 provide good information.
One other measure stayed the same, and the rest, the majority, improved, giving us statistical indication of a positive correlation between the introduction of statelessness and improved living conditions. But how come? I can think of a few responses.
1) Leeson, (and others,) used wildly inaccurate statistics. If so I’d imagine we’d have to rely on qualitative accounts (preferably first-hand), of goings on in Somalia to determine whether their lives improved or deteriorated. But some sources portray Somalia as having existed in a “bitter 20-year civil war”, with attacks on foreign doctors and reporters, and widespread internal conflict, while others describe it more as a torn region struggling, with some success, to recover, while providing no clear conclusion as to the eventual necessity of a new government, (see, e.g. Maass 2001; Cockburn 2002; Bhalla 2004; The Economist 2005; Donnelly 2006; and Rogers 2013) and a few like Jamil A. Mubarak, Jim Davidson, Michael van Notten, and Peter Little say that parts of Somalia grew peaceful under statelessness and significantly improved economically. Since these portrayals contradict a bit, I don’t think we can use them by themselves to establish a negative correlation between statelessness and standards of living in Somalia.
2) Of course Somalia’s living conditions improved, the U.S. and U.N. sent monetary aid, and ground troops to help stabilize the region. However:
A) The main area seeing international intervention, Mogadishu, has evidently seen more violence than other areas of Somalia, providing some question as to whether intervention had a stabilizing effect, rather than perhaps making things worse. (see, among others, Davidson 2001; Leeson 2007; and Powell et al. 2008)
B) Anthropological research by Ioan M. Lewis and Ahmed Y. Farah in the relatively peaceful northern region of Somaliland indicates that local, relatively decentralized, “grass-roots” peacemaking efforts, and not U.N. intervention, brought about the level of peace achieved in that area. Their research seems to me to agree with the thesis put forward by Coyne and MacCallum that international intervention had no net positive, and may have had a net negative, effect on Somalia’s stability. (By providing a financial incentive for “warlords” to continue fighting over aid money.)
C) The actions of U.S. and U.N. peacekeepers over the years have apparently included torturing Somali teenagers, injuring and killing civilians in air strikes, funding “warlords” and running secret prisons. None of these sound like particularly effective peace building measures.
3) Ian’s research just shows that Somaliland and Puntland achieved a degree of peace by establishing governments. Problems:
A) As Leeson points out, these “governments” only controlled small parts of the territories they claimed, and didn’t seem to have the power to force people to pay taxes, or establish a monopoly over the use of force. Andre le Sage’s research on justice in Somalia indicates that they depended on a plurality of institutions for dispute resolution, and security provision, and according to most authors they relied most heavily on customary law (what they called Xeer), not centrally provided courts. (see, e.g., Lombard, Gladitz, and Friedman)
B) The Awdal region seems to have had relative peace and security even without falling under the nominal protection of Somaliland or Puntland. The Awdal “capital,” Borama, may have gained its security from a local city government, but again, we would need to establish that they had some power to enforce taxation, and a monopoly over the use of force.
C) Leeson points out that the distinction between “governments” and non-government institutions may constitute more of a spectrum than a black and white dichotomy. I do not regard Somalia as purely anarchistic, (having non-hierarchical social organization, not necessarily equivalent to “stateless”,) and while according to I.M. Lewis Somalia’s traditional stateless system of justice lacked hierarchy, others have criticized the portrayal of Somali customary law as egalitarian or non-hierarchical. Various anarchists have also argued that Somalia never qualified as anarchic. In any case, an in-depth comparison of Somalia to the sorts of societies different anarchists advocate probably escapes this article’s scope.
4) Dani Rodrik argues that while Somalia may have improved under statelessness, without the creation of an effective government they will remain impoverished. While I find this plausible, it differs from the argument I wish to respond to presently.
Conclusion: If I can not find a way to demonstrate…
A) a negative correlation between statelessness and standards of living in Somalia…
B) that a positive correlation resulted from a negative effect from statelessness combined with a significantly stronger positive effect from international intervention or…
C) that the positive effect resulted primarily from the creation of proto-state institutions, rather than trade and customary law…
…then I have not yet thought of a way to establish a causal link between statelessness and violence or poverty in Somalia. I hope I have helped readers understand why I do not accept the claim that Somalia demonstrates the impossibility of a peaceful, prosperous anarchist society. I invite those who’ve made the argument I criticize, especially fellow skeptics and freethinkers, to respond.
References: (Mouse over for full citation)
(1981: Robarchek) (1993: Everett) (1994: Kusow; Farnsworth) (1997: Cooney; Mubarak; Lewis et al.) (1998: Maccallum) (2001: Samatar; Maass; Davidson) (2002: Bigwood; Cockburn) (2003: Notten; Little) (2004: Hargis; Bhalla; Graeber) (2005: Le Sage; Nenova et al.; Hagmann; Graeber; The Economist) (2006: Kim; Donnelly; Coyne) (2007: Menkhaus; Leeson; MacCallum; Rodrik) (2008: Abdinoor; Powell et al.; Schaeffer; Lefkow) (2009: Leeson et al.; Powell; Knight III) (2010: Long; Carson) (2011: Drum; Francavilla; Scahill; Barker; Atkins) (2012: Reporters Without Borders; Leeson; Halliday) (2013: Pflanz; Pinker; “Lord Keynes”; Rogers; Pflanz; Grimmett; Arman) (2014: Abdihamid; Binswanger; Various) (Year Unknown: Lombard; Rorison; Gladitz; Friedman)