Paul Staniland raises important questions in his review of Philip Roessler’s latest book (highly recommended):

I just finished reading Philip Roessler’s excellent book for my graduate Civil War seminar. Already a fan of his 2005 piece on electoral violence, I learned a lot from the new book and highly recommend it. But, just as when reading major work by Will Reno, Reno and Chris Day, Jeremy Weinstein, Paul Collier, Jeffrey Herbst, and others, I had the reaction that “This looks nothing like the places I study.” At least in the stylized world of African politics presented in these projects (I have no idea if this is accurate), Hobbesian insecurity preys on all in the absence of any real institutions, ethnic balancing and calculation dominates any other form of politics, and regimes are held in place by fluid, shifting alignments with “Big Men” rooted in local power bases.

As a result, we get shambolic and weak central regimes prone to either coups or revolts, and rebels easily bought off by patronage or co-optation. Weinstein highlights the inability of ideological rebels to overcome waves of material resources that eliminate discipline or politics, Roessler’s regimes are simply what Skocpol calls an “arena” for political competition between social actors rather than possessing any institutions or interests autonomous from social forces, and Reno’s civil wars (with the exception of “reform rebels”) are simply a grim game of bargaining over patronage between states and insurgents that are more similar than different.

Is Africa that different?

Roessler, indeed, argues that Africa has a “unique institutional structure” in which external conflicts are rare and internal disorder common. If Africa is indeed unique, it is hard to know how arguments rooted in the African context can travel beyond Africa.

Read the whole thing here.

I would argue that there is not a uniquely African civil war story. Weak states everywhere, including in Africa, are gonna weak state.

A more useful analytical delineation is what Staniland suggests:

At minimum, I’m becoming increasingly convinced that research on civil war needs to become at least partially bifurcated into work on its dynamics in very weak states (the representation of African conflicts dominant in the literature, plus Afghanistan and a few others) versus those in medium-capacity states (India, Colombia, Indonesia, Russia, etc) that possess large, centrally controlled conventional an

Think of the Nigerian Civil War between 1967-1970. The Biafra War involved a relatively strong state facing a relatively well-organized and disciplined secessionist army — much in the mold of middle income conflicts. In the same vein, countries like Kenya and Ethiopia have managed to quell rebellions in Mt. Elgon & the south coast, and in the Ogaden, respectively, in ways that would look very familiar to Staniland.

Completely anarchic conflicts involving collapsing states and incoherent hyper-localized rebellions — your stereotypical African conflict, if you will — are a unique historical experience rooted in the states that did really fall apart in the late 1980s to early 1990s (pretty much in the midst of Africa’s continental economic nadir). It is instructive that these states were concentrated in the Mano River region and Central Africa, some of the regions worst affected by the socio-political challenges of Africa’s lost long decade (1980-1995). 

And given recent economic trends in Africa (see image), it is not surprising that conflicts are becoming rarer in Africa (much in line with Fearon and Laitin). I would also expect markedly different kinds of conflicts should they emerge. There is a reason Boko Haram has never posed an existential threat to the Nigerian state, very much in the same way that India’s Maoist rebels are a peripheral matter.

I always remind my students that the Africa they know is more often than not the Africa that existed between 1980 and 1995. We all need to update.