« Six Tools to Keep China From Making Trouble | Main | The Iraq War is an Arab Opportunity, but Lasting Democracy Needs a Lasting US Commitment »
May 03, 2005
The Paramilitary Dilemma: Should We Recruit Iraqi Thugs to Beat Iraqi Thugs?
Praktike spots a trend in Iraq: whether it's paramilitary commandos employed by the government, or paramilitary militias that "pop up" on their own, the war against the Iraqi insurgents is increasingly being fought by Iraqi paramilitaries. This is good news and bad news. Paramilitaries are Iraq's best hope for quick peace -- and also a force that could spark mass killings or civil war.
Most counterinsurgency successes since World War Two, especially the few quick counterinsurgency successes, have relied on paramilitaries. Paramilitaries have unique advantages against insurgents that regular police and soldiers don't -- they combine the strengths police and soldiers each have against insurgents, and leave behind the weaknesses that insurgents try to exploit. These militias and commandos may be Iraqis' only hope to get peace on the streets in months instead of years.
But paramilitaries in Iraq are also bad news: unless the government handles them very carefully, anti-insurgent militias and commandos tend to become as brutal as the insurgents. And because the paramilitaries have the backing of the government, they have the potential to take over the country and kill their personal and political enemies in numbers the insurgents can only dream of. It's happened before.
The paramilitary solution to insurgency
Successful counterinsurgencies usually involve a network of paramilitary "local self-defense" militias, or else an entire national paramilitary professional organization. Historical examples include the Italian carabinieri, which led the effort against the Communist Red Brigades; Peru's endorsement of local militias called rondas campesinas, which turned around the fighting against the Shining Path guerrillas; or the village commando units set up to fight the Hukbalahap insurrection in the Philippines. Probably the best-documented example of paramilitaries' edge against insurgents is, ironically, the Vietnam War.
During the Vietnam War, there were only three programs that actually succeeded in shutting down communist guerrilla activity in their areas. Those were the CIA-administered (using Special Forces personnel) Civilian Irregular Defense Groups (CIDG), the Marines' Combined Action Platoons (CAP), and late in the war the joint Army-civilian Civil Operations and Revolutionary (later Rural) Development and Support (CORDS) program. All three programs ignored the hundreds of thousands of regular American and Vietnamese troops on battalion-sized maneuvers. CIDG, CAP and CORDS all focused on building up and supporting tiny local village militias. And CIDG, CAP and CORDS all succeeded in shutting down guerrilla activities in their areas of operation.
Insurgents defeat police by being more focused and more violent. Insurgents beat soldiers by building ties with the civilians. They lose those advantages against paramilitaries. Whether we're talking about informal militias or formal commando units, paramilitaries, like police detectives, make their habit to operate among local civilians and persuade them to become helpers and informants. But like soldiers, paramilitaries don't spend their days walking a beat to prevent public disorder; they use their informants to point out the government's enemies so they can capture or kill them.
Insurgents intimidate or persuade people to help them kill government supporters. Paramilitaries intimidate or persuade people to help them kill insurgents. Since the paramilitaries are on the same side as the government, they can talk to more people safely, offer bigger bribes, and move around faster and in larger groups than the insurgents. Insurgents rely on staying "asymmetric" to make up for lack of resources. Paramilitaries are a "symmetric" counter to insurgents, using their own tactics against them with the advantages of government support.
The most recent example of paramilitary success against insurgents is Afghanistan today. While it's easy to find things to complain about in Afghanistan, the Taliban there haven't been able to muster anything like the kind of continual attacks that the Iraqi insurgents have. In Afghanistan, we endorsed the local gunmen instead of bringing in our own. The result was that whenever the Taliban tried to reinfiltrate a village or city, there were plenty of local fighters in a position to find out about them and run them back out of town. The warlords' militias kept Afghanistan secure from the Taliban. Of course, now the warlords' militias are a problem of their own -- and that's part of the dark side of paramilitaries.
When paramilitaries go bad
Because paramilitaries have a license to find and kill the enemies of the government, they have almost absolute power in their local community -- and absolute power corrupts absolutely. It's only too easy for paramilitaries to create excuses to kill off anyone in town who crosses them. From Peru to Vietnam, militias have tended to conveniently discover that their personal enemies were enemies of the state. Worse, paramilitaries can use an insurgency to seize power for themselves and set up their own psuedo-governmental empire. That's what's happened in much of Columbia, where the right-wing paramilitaries long ago became more involved in being the government and in running drugs than in helping the government against rebels. But even replacing the government isn't the worst thing that unbound paramilitaries will do.
Paramilitaries only tend to be needed in weakly governed countries, and weakly governed countries tend to have nasty ethnic or class feuds just below the surface -- especially once an insurgency breaks out. If the government lets the paramilitaries take too much authority, the result tends to be ethnic cleansing -- or genocide. In El Salvador, despite American advisors' complaints the counter-guerrilla forces killed 70,000 supposedly left-wing people out of the country's six million. In Guatemala, where America withdrew in disgust, the paramilitary army slaughtered 200,000, mostly ethnic Maya and almost all civilians, out of a population of only 10 million. In Rwanda, in Bosnia, and in Sudan today, the key step for ethnic slaughter was government endorsement of an ethnically based militia.
In Afghanistan, there have been no mass killings by the militias, because the new central government has been using its own (American-trained) army to pressure the local warlords into good behavior. Governors who had been running their provinces too independently have been replaced. While the militias in Afghanistan are still present in the background, the problems Afghanistan has these days are mostly civilian problems -- poverty, poppy farming, bad roads. In Afghanistan, the government let local militias defang the insurgency, and then the national army defanged the militias.
In Iraq, the new government has not had the deft political touch that Hamid Karzai's government has had in Afghanistan, and the insurgency is far stronger. The odds of militias going out of control are worse. But none of the Iraqi ethnic groups want civil war -- the Shiites would lose the fruits of peace, and the Sunnis would lose, period. So perhaps America can keep the central government honest, and perhaps the central government will be able to push the militias into standing down after the insurgents are gone.
It's too late for Iraq to be as good a success as Afghanistan. But the paramilitary commandos and militias should get the insurgency defeated faster. It probably won't lead to another Bosnia. I do wish I didn't have to put "probably" in that last sentence. It all depends on whether Iraq's government will have the nerve and skill to take power from the paramilitaries -- while the paramilitaries take power from the insurgents.
Endnote: the Vietnam War counter-guerrilla militia successes
How successful were our Vietnam militia efforts? During its CIA period (through 1962), the Civilian Irregular Defense Groups (CIDG) featured 24 12-man Special Forces teams coordinating an astonishing 38,000 irregulars against the guerrillas. By 1962 America's fewer than 300 leader-advisors of CIDG were holding secure an entire province of 300,000 civilians against the Communist guerrillas. Then the Army leadership at MACV took over from the CIA. The Army in 1962 didn't yet believe guerrilla wars should be fought differently from conventional wars. The generals at MACV transferred CIDG to offensive operations, ending the support for local militias. Darlac Province was then reinfiltrated by the Communists.
The Marines' Combined Action Platoons were a more improvised program than CIDG, relying on ordinary Marines rather than the unconventionally trained Special Forces of CIDG. CAP was a mix of modest improvements and decisive successes; results varied from village to village. See Bing West's The Village for a wonderful account of one CAP effort.
Civil Operations and Revolutionary (or Rural) Development, CORDS, the last and largest militia-supporting pacification program, didn't begin serious work until late 1968. In 1968 the Communist guerrillas were estimated to be in charge of 45 percent of South Vietnam, about the same as when regular American infantry battalions had first been sent into action against the Communist guerrillas in 1965. CORDS threw its weight behind the till-then neglected Vietnamese village and province militias, the RFs and PFs or "ruff-puffs" -- and using those militias, CORDS reduced the Communists' guerrilla presence from half of South Vietnam to a toehold of about 5 percent by late 1970.
CORDS' success in shutting down the guerrillas forced North Vietnam to finally dare two straightforward conventional invasions of the south, openly driving southward with tanks and artillery. The conventional invasions succeeded, thanks to the South Vietnamese soldiers being commanded by lousy officers. In 1972 North Vietnam took a quarter of South Vietnam, despite American airpower support for the South Vietnamese troops. In the North's 1975 invasion, when the Americans no longer provided the South with airpower support, the Communists did even better, seizing the remaining three-quarters of the country.
Ironically, although we think of Vietnam as a guerrilla war and an American defeat, in fact by the end of 1970 America and South Vietnam had won the guerrilla war. We lost Vietnam just the same. The North Vietnamese were capable of conventional as well as guerrilla war. And America and South Vietnam had failed in their mission of bringing skill and professionalism to South Vietnamese army officers. But in the bitter last years, the American military had finally rediscovered how to succeed against guerrillas.
[E2] Mudville and Beltway OPL.
Posted by danielstarr at May 3, 2005 06:45 AM
Trackback Pings
TrackBack URL for this entry:
http://www.danielstarr.com/movabletype/mt-tb.cgi/31
Comments
Question for you: How do you define a paramilitary? There are no official military advisors (foreign or domestic) advising the Colombian paramilitaries, which are tolerated and worse by the government. But in Iraq it seems that we're talking about something that the government is seeking to control, even if it only came in after they were created. In my experience in El Salvador, I simply don't recall anyone using the term "paramilitary," except perhaps in reference to death squads.
And isn't the problem in highlighting CORDS' success in Vietnam is that it involved the particularly brutal Phoenix operation?
Posted by: David Holiday at May 3, 2005 02:24 PM
"Paramilitary" is almost as flexible a word as "insurgent," but here I use it to mean units who train to carry out military missions in civilian environments -- units that (like police) see having civilians around as normal and even desirable, but (like soldiers) see killing their targets as an acceptable and ordinary if not ideal outcome.
I would not have called the El Salvadoran army paramilitarized, although the death squads were. But then, I also would not have called El Salvador more than a modest success -- it took too many years and lives, and while it was better than what Westmoreland did in Vietnam, I wouldn't hold it up as an example of "best practices".
El Salvador is something the US Army correctly points to as showing it had learned to avoid the worst traps of Vietnam. That doesn't mean we should copy it overall as a model today. The one solid "model" lesson to imitate from El Salvador is that the United States does better by being pushy advisers and trainers in counterinsurgency than by doing it ourselves.
It's worth noting that insurgencies vary in their degree of intimacy with civilians. In Iraq we've got an urban insurgency, which makes the insurgents very intimate indeed with the civilians, and pulls everyone toward the "police tactics" end of the scale (informants, raids) as opposed to the "light infantry tactics" (long-range patrols, ambushes) often used in more rural insurgencies.
In Iraq, there are all levels of control: government-organized, government-cooperative and fundamentally independent militias. You're right that you get a different mix of strengths and problems depending on which we're talking about.
CORDS had Phung Hoang (Phoenix) as a contributor, but not actually a big contributor. On the evidence, I have to say Phoenix was in many ways a failed experiment. The North Vietnamese did blame it for thousands of Communist cadre deaths. But those had more to do with Phoenix intelligence-gathering than the Phoenix hunter-killer teams (Provincial Reconnaissance Units or PRUs). And even Phoenix intelligence-gathering was just a modest boost to the real heart of CORDS: supporting the Regional Forces / Popular Forces militias. Even before CORDS, the RFs/PFs were taking a disproportionate share of the guerrilla burden; the whole history of Vietnam is that no one in Saigon took the militias seriously, but whenever they did the guerrillas started falling apart.
With or without intelligence-gathering programs like Phoenix, villages with sound militias dried up as a source of Communist support unless the Communists were able to bring in larger units to clear out the militias -- and that tended to bring in the regular Vietnamese or American troops.
So it's quite true that Phoenix hunter-killer units (the PRUs) seem to have killed several civilians for each actual guerrilla. It's also true that Phoenix intelligence was a useful contribution. But the intelligence effort didn't require the hunter-killer teams, and the overall CORDS effort was only moderately boosted by the intelligence effort.
CORDS' success in Vietnam was primarily just a matter of being committed to the local militias. Secondarily, it helped a lot that some real land reform ("land to the tiller") finally took place, which took away the Communists' best political argument (compare with holding of real elections in El Salvador, or with the use of honest elections plus land reform in the Philippines by Magsaysay to shut down the Huk rebellion). (The militias could have dominated locally in any case, but land reform made everyone more willing and the outcome less bloody.) On balance the evidence suggests that other factors, like Phoenix intelligence, were useful but not especially necessary.
(Formal intelligence-gathering would have been more important if the work really was all going to be done by outsiders like the PRUs. But the RFs/PFs tended to acquire informants naturally: they lived there.)
I should add, on the subject of both Phoenix's PRUs and Iraq today, that locally raised militias seem to get things done with much less brutality than outsider units like the PRUs or some of Iraq's government commando units. Of course, there's a dilemma there, because the more the government leans on locally raised units to defeat the insurgents, the more trouble it faces later to push aside those same local units. We'd hate to see Iraq avoid "Vietnam 1965" only to become "Lebanon 1982," let alone "Bosnia 1993."
Posted by: Daniel Starr at May 4, 2005 12:03 AM
Thanks. This inspires me to go back and do some reading on Vietnam. Any suggestions?
And, if you don't mind my asking, who are you?
Posted by: dholiday ![[TypeKey Profile Page]](http://www.danielstarr.com/nav-commenters.gif)
at May 4, 2005 05:59 AM
Allow me to reiterate David's question. I came over here after seeing your comments on Liberals Against Terrorism, and am impressed by your level of knowledge on several matters.
Again, no rudeness intended, and if you don't mind divulging, what is your background? I only ask because of the high quality of posts here.
Posted by: Eric Martin at May 4, 2005 06:41 AM
The most readable general text, if that's what you want, is Karnow's Vietnam. But if you want to dig in to the counterinsurgency efforts, I recommend Krepinevich's "The Army and Vietnam" for the big picture, Bing West's "The Village" for the little picture, Sorley's "A Better War" for the optimist's view of the late years, Willbanks' "Abandoning Vietnam" for a more cold look at the late years, and Komer's "Bureaucracy at War" or "Bureaucracy Does Its Thing" for a good look at the upper-level organizational issues that delayed work like CORDS for so long. Summers' "Vietnam War Almanac" is a nice overall reference. Hackworth's "About Face" and "Steel My Soldiers' Hearts" have some good illustrations of the kinds of frustrations the better American officers ran in to. You may also enjoy Herrington's book on his Phoenix advisor experience -- although the trouble with Phoenix is that, like CAP, the implementation varied widely from place to place, and the good analysis tends to come out in in bits and pieces in research papers rather than unified books.
Posted by: Daniel Starr at May 4, 2005 06:52 AM
I'm just a denizen of the think-tank universe. Really nothing more to say about it than that.
Posted by: Daniel Starr at May 4, 2005 06:54 AM
Hmmm, the mystery mostly continues....
(details sought - which think tanks per se, area of expertise etc,)
Posted by: Eric Martin at May 4, 2005 07:38 AM