Building on Dan’s observation this past week, Theo McLauchlin is a PhD student at McGill University offers us some insights on the role of the military in the various Arab revolutions we’re witnessing. He works in the area of military defections and civil wars.

Which Middle Eastern regimes seem liable to fall? That’s a popular question these days, and an important answer, as Dan Nexon points out, is that it depends on each country’s armed forces. But what they are likely to do is something most people don’t seem inclined to speculate about. That caution is warranted, as I’ll argue below. But what can we say? What ideas do we have at our disposal for thinking through what militaries will do?

One important factor might be professionalization. Lucan Way notes that in comparison with the rest of the Middle East, Tunisia and Egypt seemed to have relatively professional, depoliticized armed forces, able and willing therefore to act in concert to say “enough”. The regimes lacked the “coup-proofing” techniques that, according to a quite extensive literature, have helped prop up authoritarian regimes across the Middle East. You promote your friends, marginalize your adversaries, and don’t trust anyone too much. Multiple different internal security agencies, for example, keep an eye both on officers and on each other. If an officer is plotting a coup, he stands an awfully good chance of being reported by someone else eager to curry the dictator’s favour. These techniques really do seem effective at preventing coups.

Despite some confusion about how to think about Egypt–John Barry and Christopher Dickey argue that it was in spite of Mubarak’s coup-proofing that he fell, not because he didn’t do it enough–there’s a lot going for this approach. But it shouldn’t be taken too far. It is not as though heavily politicized armies have a great overall track record of defending regimes. For example, in comparing China in 1989 to Indonesia in 1998, Terrence Lee finds quite the opposite: China’s more professional army stood by its regime while Indonesia’s more heavily politicized armed forces fell apart.

Lee makes the point that coup-proofing isn’t necessarily meant to deal with massive popular uprisings. The threat of punishment for defection depends on the willingness of others to inflict it. It’s rational for an officer do the regime’s bidding as long as, but only as long as, he expects the regime to survive. When officers have good reason to believe that the regime will fall, they have heavy incentives to make sure they’re not backing a losing horse. And popular uprisings can throw a previously stable expectation out the window. As I argued in a paper last year, an external rebellion can provoke a cascade effect within the military. Timur Kuran’s insights about the tipping point in Eastern Europe in 1989 can apply to armed forces too. This is how it’s possible for officers and soldiers to defect from a dictator despite fearsome coup-proofing systems.

The trouble for the analyst is, as Kuran argued, that this implies that prediction is extremely difficult if not impossible. Everyone has a strong incentive to keep their preferences and their likely actions a secret–from the dictator and, necessarily, from us.

But in my paper I also tried to develop some limits to these revolutionary cascades. In particular, some regimes are governed by minority groups that are given heavy preference–especially within the military. Jordan, Syria and Bahrain are important examples. This means that opposition tends to rally among out-groups–and, in consequence, the loyalty of the in-group gets reinforced. Essentially, over time, a stable expectation can develop, associating some groups with the regime and others with the opposition. In rebellions, they have little other choice than to stick with the regime. An East Bank officer in Jordan in 1970 was not likely to do well out of a PLO victory, nor an Alawi officer in Syria in 1982 from a Muslim Brotherhood victory. Palestinian and Sunni officers, respectively, were more likely to defect. While this makes regimes vulnerable to out-group opposition, it can be a perverse strength, because it gives regimes a relatively stable core of support to count on. You get a core of strong support at the cost of encouraging continuous, low-level unrest. It’s loyalty on the cheap.

One interesting consequence of this approach is that the regimes that look the least stable–Jordan in 1970, Syria in the late 1970s–can have a better chance of surviving a rebellion than regimes that look lots more solid, like Iran’s before 1978.

Is all this any help in understanding the Middle East today? It suggests attention not so much to professionalism vs. non-professionalism, but rather to how much of the army is closely identified with the regime. The answer in Egypt, by all accounts, is: not much. This helps explain why basically none of its army was willing to defend Mubarak to the hilt, unlike in Libya. It also helps explain why the army’s command had the flexibility to declare, early on, that they would not fire on unarmed protesters. They were able thereby to keep their own credibility with the opposition. And, given that many soldiers looked likely to disobey such an order, it’s unclear that the regime would have lasted anyway, even if the top brass had tried to crack down.

In Libya, more of the armed forces seems closely identified with the regime. More have therefore stayed with Gaddafi, with bloody consequences. Part of this has to do with Libya’s now-famous tribal divides. According to Hanspeter Mattes , Gaddafi’s own Gadhadhfa tribe, along with the Warfalla and Maqarha, have dominated the armed forces. However, both the Warfalla and the Maqarha have had a chequered history with the regime; Warfalla officers rebelled against Gaddafi in 1993 and picked up the support of some Maqarha officers (see this free but gated article). The Warfalla and Maqarha tribes have defected. I suspect, as do others, that the Gadhadhfa tribe will stick with the regime out of fear of what happens if they don’t. In this context I note that Gaddafi’s resort to bombing civilians might not just be because he is horrible and callous (though of course it has a lot to do with that, too). It may also have something to do with his tribe’s dominance in the air force. It reminds me that Hafez al-Assad used the Alawi-dominated air force and artillery at Hama in 1982 to awful effect. That is not a pleasant thought.

I could well be wrong, and I hope I am: one of the pilots who deliberately crashed his plane rather than bomb civilians was a Gadhadhfa tribe member.

More broadly, communal politics should provide a bulwark to the regimes in Syria, Jordan, and Bahrain. It’s in the latter where the protests are strongest at the moment. And there, the opposition has attempted to cut across the Sunni/Shi’ite divide, but the regime seems to know where its strengths are: it’s been specifically targeting Sunni protest leaders. (I’m not sure this applies to Yemen’s north/south divide; the fact that the south seems to aim at secession rather than a government takeover suggests it’s not so easy to blame for protests in Sana’a.) Every regime tries to delegitimize its opposition, with Gaddafi’s Bin Laden conspiracy theory only the most absurd attempt so far. In Bahrain, the regime has more to work with.

My approach fully accepts the extreme difficulty of making predictions where there aren’t communal divides. I don’t think either Tunisia or Egypt (or Algeria, or Morocco, or Saudi Arabia) was really a foregone conclusion. We can try to understand both in retrospect, but I don’t claim that we could have predicted the fall of either regime terribly easily. It is still just as amazing as it ever was.