CIA’s Phoenix Program Flies Again

Mention of the Vietnam-era “Phoenix” program typically conjures images of rogue CIA-backed assassination teams roaming the Vietnamese countryside executing Viet Cong agents and innocent civilians alike. Now, a RAND research team says the controversial Phoenix program demands reexamination and may provide useful lessons for the ongoing war in Afghanistan.

According to the RAND study, the Phoenix program has been mischaracterized as an assassination program. It was not. Phoenix was an intelligence operation that aimed to lift the shrouds of the “invisible Vietcong empire” that operated in rural villages and hamlets. Phoenix did have an “action arm,” made up mostly of South Vietnamese, along with U.S. advisers, that targeted “subterranean” high-value targets: the communist cadres who collected taxes, gathered information, spread propaganda, and recruited new members. Although sometimes unavoidable, killing Viet Cong suspects was not the program’s intent. According to an American adviser quoted by RAND, capturing and interrogating suspected insurgents was the aim so that the shadow government could be unmasked: “prisoner snatches were key. You can’t get information out of a dead man.”

It was counterinsurgency savvy Robert Komer, director of the CORDS rural pacification program, who started Phoenix in 1968, originally named the Intelligence Coordination and Exploitation (ICEX) program. Komer knew that no U.S. counterinsurgency effort could be successful until the Viet Cong shadow government was eliminated; trying to graft a U.S. backed local level government on top of it would only end up getting U.S. allies killed.

The first task Komer tried to solve was the intelligence coordination mess that existed in Vietnam caused by the alphabet soup of rival agencies, both civilian and military, the fractious state of the South Vietnamese government, which had not one but three national level police forces and the age-old challenge of coordinating, prioritizing and disseminating “actionable” intelligence. Phoenix established intelligence coordination centers at the province and district level, but the problem of intelligence “sharing” between agencies and governments was never really solved.

Phoenix’s action arm was the CIA trained, equipped and managed Provincial Reconnaissance Units. Never numbering more than 5,000, RAND describes the PRUs as an “intelligence driven police force-better trained, equipped, and paid than the South Vietnamese National Police, and with a highly specialized mission, to be sure, but a police force nonetheless.”

The intelligence provided by the PRUs was Phoenix’s real value added. Jealously guarded by their CIA handlers, the PRU teams, operated in small, distributed units, “served in their native provinces, giving them a depth of knowledge about local conditions unmatched by any other South Vietnamese government (let alone U.S.) forces.” While supposed to be fed intelligence from higher levels, the PRUs operated as their own intelligence collection and exploitation arm; according to one U.S. adviser, 75 percent of the time the PRUs did their own targeting. U.S. advisers operated alongside the PRUs as they conducted “snatch-and-grab operations against the Viet Cong, providing at once a check on their targeting but also providing medevac and air support if needed. American adviser numbers were always small, at the height of the program in 1970 just 102 U.S. military and five civilians advised the PRUs.

The RAND researchers say the PRUs were highly motivated, vitally important when waging a covert campaign against the equally motivated communist People’s Army. Team members were drawn from elite South Vietnamese Army units and were well paid thanks to CIA “largesse.” Although it would be wrong to characterize them as mercenaries, RAND says, quoting a U.S. adviser: “Most were professional soldiers, they liked soldiering, and they were nationalistic. And they had scores to settle with the communists.”

Ultimately, while neither as wildly successful nor as roguishly merciless as often portrayed, RAND says Phoenix’s integration of local level human intelligence gathering with a responsive action arm proved effective in dismantling the Viet Cong shadow government, but then only when it had adequate forces; the CIA was always stretched thin in Vietnam. Some Phoenix critics charge the program only neutralized low level insurgents. Yet, as RAND says, after the war, North Vietnamese officials said Phoenix proved “extremely destructive” to the village and hamlet level shadow government.

Phoenix has particularly apt lessons for the U.S. led effort in Afghanistan, RAND says, where insurgents groups are too often regarded as “impenetrable black boxes,” whose inner workings, recruitment practices, sustainment activities, leadership, and decision making process, are largely a mystery. The most important lesson from Phoenix: the value in unearthing the insurgent’s subterranean “ecosystems,” the largely invisible structures that sustain it.

What’s the best way to pry open those Taliban and insurgent black boxes? RAND’s researchers looked at the two major “models” used by the U.S.: one based primarily on technical intelligence, the other based on working relationships at the local level. More effective is that second model, RAND says, best exemplified by Iraq’s Sunni “Awakenings” movement, a “real heir” to the Phoenix program, combining as it did close operational ties forged between U.S. commanders and Sunni tribes and power brokers to leverage local intelligence to attack insurgent networks.

Which raises the question of whether something along the lines of Phoenix, specifically the PRU teams, can be replicated in Afghanistan? One of the very weak links in Afghanistan is the national police, a corrupt, poorly resourced and hugely ineffective force. Having spent some time in Afghanistan, observing the Afghan police, I think the U.S. would be hard pressed to create an elite intelligence gathering and action force at the local level along the lines of Vietnam’s PRUs in Afghanistan. The basic building blocks of such a force, a highly motivated, well trained cadre of soldiers loyal to the Afghan government or even the concept of an Afghan state, simply might not exist.

Taking another look at the Phoenix program does raise the question of whether U.S. efforts in Afghanistan are weighted too heavily in trying to build an Afghan national army versus the Afghan police, since counterinsurgency is largely a police function. Rebuilding the Afghan police has, until recently, been left to the Europeans, which has so far proven a disaster. It may be time for the U.S. to devote far more resources to developing a truly effective Afghan police force.