On the persistence of the belief that authoritarian states are better at COIN

Sometimes it seems like this little conflict studies subculture of ours exists merely to chew over the same intellectual cud repeatedly. How often do I read a paper or a blog post positing some new driver of violence, some controlling factor in conflict, some recipe for pacification and think "geez, where was I just reading about that?"

Patrick Porter, formerly of Kings of War and now writing at the excellent new-ish blog The Offshore Balancer, is not the sort I'd generally accuse of this failing. For one, the new blog tends to focus on broader strategic issues rather than the COINy day-to-day. But yesterday Patrick took up a question that's been batted around quite a bit lately (and quite sensibly, I might add, in response to the love-n-hugs Surge narrative that's taken hold in some corners); quoting Starbuck (lot of that going around) at Wings Over Iraq, Patrick asks "does the brute force method ever work in counterinsurgency? [And if so] Under what conditions?" While stipulating that brutal campaigns have on certain occasions been effective in the past, the rest of the post explores the whys and wherefores and posits that this is true in spite of brutality, rather than because of it.

He then tacks in a different direction, proposing a number of environmental conditions that may be conducive to the success of an insurgency, one of which is this:

when they fight against liberal democracies: this is a more tentative argument here, but liberalism probably makes it harder [for the counterinsurgent] to gain domestic acceptance when most ‘small wars’ are mightily unpleasant and prolonged affairs with almost inevitable abuses, and when elected leaders care an awful lot about being re-elected.

Now this isn't exactly a novel suggestion. Those who make this case usually do so for a few basic reasons. The argument pretty much goes like this (don't get confused by the blockquote -- I'm just setting it off so it's easier to read, I'm not quoting anyone):

Democratic leaders are held accountable by voters, and are thus less likely to engage in unpopular wars or pursue them for an extended period of time. Casualty-averse publics will throw out a government that gets a lot of people killed, and the free media helps bring those casualties to light. Not only that, but press freedom means it's harder for a liberal government to hide the human rights abuses and other brutal tactics that could prove helpful in prosecuting a counter-guerilla campaign. In short, liberal regimes 1) have their hands tied, and 2) give up easier.

All of those things make a fair bit of sense, perhaps with the exception of the bit about the counterinsurgent's brutality being harder to hide or tolerate in a democracy. After all, if we don't think that brutality is particularly useful, anyway, then why would efforts to restrain such behavior really have such a negative effect on the campaign.

But as I'm reading this, I'm thinking "haven't I seen this debunked in a couple of places lately?" And after going through my email and doing a little Googling, it turns out I have.

First we've got Jason Lyall's recent work on this subject, which is probably the freshest and most influential: "Do Democracies Make Inferior Counterinsurgents? [pdf]: Reassessing Democracy's Impact on War Outcomes and Duration," from the Winter 2010 edition of International Organization. Here's the abstract:

A core proposition from decades of research on internal wars asserts that democracies, with their casualty-averse publics, accountable leaders, and free media, are uniquely prone to losing counterinsurgency (COIN) wars. Yet one should question this finding, for two reasons: First, existing studies overwhelmingly adopt no-variance research designs that only examine democracies, leaving them unable to assess their performance relative to autocracies. Second, these studies do not control for confounding factors that bias causal estimates. Democracies, for example, typically fight wars of choice as external occupiers, while most autocracies face homegrown insurgencies, a function in part of divergent levels of state capacity possessed by democratic and autocratic combatants. This study corrects for both problems using a new dataset of insurgencies (1800–2005) and matching to test whether democracies experience significantly higher rates of defeat and shorter wars. No relationship between democracy and war outcomes or duration is found once regime type is varied and inferential threats are addressed.

(For a less academic summary and explanation of the work, see this very good interview with Lyall from the Rosner's Domain blog on the Jerusalem Post's website.)

I read this paper back when it came out, and found it interesting for the challenge it posed to the conventional wisdom. Lyall also echoed some of the points I'd seen in a paper by Michael Engelhardt in the Summer 1992 issue of Conflict Quarterly entitled "Democracies, Dictatorships and Counterinsurgency [pdf]: Does Regime Type Really Matter?" Turns out Engelhardt found basically the same thing: no, it doesn't. Here's what the journal's "In This Issue" section has to say in summary:

The notion that democracies are "soft" and cannot sustain many casualties in revolutionary wars without a collapse of political will gained the status of an unchallengeable truism in the wake of the Vietnam war. It even resurfaced during the Persian Gulf crisis and war of 1990-91, when Iraq threatened to defeat the coalition forces "in a sea of blood." Saddam Hussein apparently believed that his regime could survive significant battlefield losses and that coalition governments could not. Michael Engelhardt examines this assumption and concludes that, at least in respect of insurgent conflicts, regime type does not matter; both dictatorships and democracies can be defeated. Furthermore, he asserts that the inclination of democracies to liquidate costly commitments can be seen as a strength, an affirmation of their democratic character, rather than a weakness.

Again we see basically the same argument about softness, casualty aversion, responsiveness to popular doubts, etc. Are autocracies really unencumbered by these sorts of concerns? Or building on that, how about what may even be a more important question: even if autocrats have a free hand in the prosecution of COIN wars, might that reality not have unconstructive knock-on effects down the line?

That's the question taken up by Yuri Zhukov (yikes!) in the September 2007 Small Wars and Insurgencies. His paper "Examining the Authoritarian Model of Counter-insurgency: The Soviet Campaign Against the Ukrainian Insurgent Army" is unfortunately behind a paywall, but here's the abstract:

In an effort to better understand the benefits and limitations of an authoritarian approach to counter-insurgency, this article examines the relationship between regime type and military effectiveness in the often neglected case of Soviet counter-insurgency operations in Western Ukraine. This study finds that the advantages authoritarian governments enjoy in designing, planning and implementing counter-insurgency campaigns - related to a lack of restraints and constraints - can all too easily become reversed through the excesses they permit.

This piece speaks more directly to the question of brutality than of regime type, I think, though it definitely links the two together: can we really perceive authoritarianism as an advantage at the same time as we dismiss the effectiveness of brutality, as Patrick Porter does? (To be fair to Patrick, he did caveat the original statement about increased odds of insurgent success against democracies by writing "this is a more tentative argument...") When taken together, these three papers would suggest that -- in spite of what looks like a common-sense justification for exactly that -- we can't.