Troop Levels in Stability Operations: What We Don't Know
Troop levels in Iraq have been one of the most hotly contested issues in American foreign policy over the past three years, from debates over the initial deployment in 2003 to those surrounding the troop surge in 2007. The Bush administration has faced significant criticism for ignoring the conventional wisdom regarding the number of soldiers required to secure Iraq, and recent attempts to change course in this area are seen by some as too little, too late.
Specifically, the Pentagon's deployment of only 120,000 American troops for the invasion and the decision by Paul Bremer, U.S. Administrator in Iraq, to disband the Iraqi army and police has kept the ratio of security forces to Iraqi civilians well below the 20 per 1,000 seen as the basic ante required to play the high stakes stabilization game. Many supporters of higher troop levels blame these missteps for the emergence of the robust insurgency and the coalition's failure to defeat it.
But where exactly does the 20 per 1,000 figure come from, how strong is the evidence supporting it, and what steps are being taken to assess and improve the conventional wisdom in this area? While the answer to the first part of the question is relatively accessible, the latter are more difficult. They address a daunting problem, but unveil a disconnect between the objectives and methods of policy and social science.
Troops Levels and Iraq
Army Chief of Staff Eric Shinseki set off a firestorm when he told the Senate Armed Services committee before the invasion that "something on the order of several hundred thousand soldiers" would be required to stabilize Iraq, a figure that began to approach 20 troops per 1,000 of the Iraqi population, the ratio that academics conventionally, if not universally, cite as necessary for successful stability operations. Secretary of Defense Donald Rumsfeld called this estimate "far off the mark," as did Deputy Secretary Paul Wolfowitz, who put the figure closer to 100,000 troops total for Iraq. As the war drags on, requiring orders of magnitude greater time, blood, and treasure than estimated by U.S. leadership, troop figures have become the foremost issue debated at all levels.
The debate over troop levels has been constant throughout the war, even among military leaders; a growing number of Democrats and Republicans in Congress are pushing to decrease troop levels over a varying set of timetables. However, President Bush has opted for an increase in troop levels over the coming months as part of one final push to achieve the "victory" he sees as the only acceptable option.
There are currently 169,000 coalition troops (including 152,000 Americans) deployed to Iraq, a ratio of 6.3 per 1,000 if only these forces are counted. If the Iraqi army is added into the mix, then the figure becomes 11.3 per 1,000, and 18.4 per 1,000 if Iraqi police forces are also included. The addition of approximately 20,000 U.S. troops would push those ratios to 7.1, 12.1, and 19.1 per thousand, respectively. These figures include non-combat support troops as well as all Iraqi army and police units that are "in the fight" according to CENTCOM, regardless of readiness. When "tooth-to-tail" considerations are included-the number of combat troops to logistical support troops-the number of U.S. combat troops in country drops to about 60,000, and coalition and Iraqi force figures face similar reductions. Therefore, only if these best case figures are used with all support troops and all Iraqi police included does the current figure even begin to approach the 20 per 1,000 believed to be needed for success.
Glass Half Empty?
Scholarship has less to offer than it should for an issue -- the stabilization of a country by security forces -- that has formed a key part of every military intervention abroad as well as the actions of every state within its own borders on a daily basis. The most well-known and methodologically rigorous work on the topic remains James Quinlivan's "Force Requirements in Stability Operations," now more than ten years old. Quinlivan's article represents an initial attempt to apply the methods of social science to the policy-relevant issue of troop levels. By examining a number of historical cases of stability operations, Quinlivan is able to calculate the troop levels employed relative to the populations they were attempting to control, offering a basic method for calculating the ratio required for successful stabilization, which he defined as "[creating] an environment orderly enough that most routine civil functions could be carried out." In a revision of the counterinsurgency (COIN) literature, which previously included ratios relative to the number of insurgents or a desired threshold of 10 troops per 1,000 of the population, Quinlivan argues for a ratio of 20 troops per 1,000 of the population to achieve successful stabilization, a figure that has remained largely unchallenged.
The theory behind these changes from insurgent to population ratios and from 10 to 20 per 1,000 stems from the idea that the success of COIN operations is based mainly on two factors: the ability of the COIN forces to gather intelligence and to separate the populace from the insurgents, thus negating the insurgents' two main advantages. Troop ratios are therefore calculated relative to the population the COIN force is attempting to control and protect, rather than the insurgents that they are trying to defeat. The ratio is relatively high because the best intelligence-gathering instruments in such a campaign remain the eyes and ears of COIN forces, despite advances in signal and imagery intelligence.
Quinlivan's article sits at the nexus of policy and science, and so carries some of the strengths and weaknesses of each. Policymakers need answers to clear, relevant questions and they need them now. A 50 percent reliable answer today is often far more valuable than a 95 percent reliable answer two years (or even two weeks) from now. Quinlivan's piece provides a clear answer to the question: How many troops do we need to seriously consider undertaking a given stability operation? The 20 per 1,000 ratio isn't perfect, but it has proven a decent ballpark figure.
Social scientists want reliable answers that are part of a cumulative research program that adheres to time-tested, often stringent methodological guidelines. The relevance of the question to policymakers is considered a second-order condition (or worse) by some, and academics would generally prefer a 95 percent reliable answer in two years to a 50 percent reliable answer today. Quinlivan employed a basic method of social science to his work-comparative case analysis-and extrapolated his results to weigh in on a key policy issue for the post-Cold War world. From a social science perspective, however, his study reveals significant flaws.
Quinlivan's lack of methodological clarity concerning key terms and hypotheses inhibit rigorous testing of his claims. For instance, he is inconsistent regarding the inclusion of police in the "security forces" component of his force ratios. He offers only cursory glances at his cases, neglecting to provide either process tracing to identify causal pathways or investigation of outliers to suggest and assess alternate hypotheses. Even the cases he examines pose problems for his argument, since only two represent stabilization successes with ratios of 20 troops per 1,000 (Malaysia and Northern Ireland) while others achieved stability with ratios in the single digits (Germany following the Second World War, India in the Punjab in the mid-1990s, and the U.S. in the Dominican Republic in 1965). Further examination reveals cases with intervention forces yielding troop ratios above 20 per 1,000 that were unable to maintain stability, such as the French in Algeria.
Unfortunately, little rigorous scholarly development has proceeded from Quinlivan's article. It remains widely cited by top academics on intervention, including some of those who criticized the Bush administration for not deploying a large enough force to Iraq in the first place. In fact, the Army's new COIN manual, written in large part by General David Petraeus, includes a summary of Quinlivan in its discussion of troop levels:
No predetermined, fixed ratio of friendly troops to enemy combatants ensures success in COIN. The conditions of the operational environment and the approaches insurgents use vary too widely. A better force requirement gauge is troop density, the ratio of security forces (including the host nation's military and police forces as well as foreign counterinsurgents) to inhabitants. Most density recommendations fall within a range of 20 to 25 counterinsurgents for every 1000 residents in an AO. Twenty counterinsurgents per 1000 residents is often considered the minimum troop density required for effective COIN operations; however as with any fixed ratio, such calculations remain very dependent upon the situation.The Way Forward
Clearly, the 20 per 1,000 ratio and the research behind it could use further examination, if not significant revision. Initial steps that could quickly improve our knowledge of this issue include assembling a larger group of cases of stability operations, followed by an examination of troop levels in cases with similarities to Iraq on some of its key features. Data on a wide range of cases could be collected in short order if one employed the relatively low standard of detail used by Quinlivan, and cases like Somalia in the early 1990s, Lebanon in the 1980s after the intervention of Israel and Syria, and Congo in the early 1960s could yield policy-relevant insights as examples of civil war occurring alongside insurgencies against foreign interveners. More advanced research goals include in-depth case studies using process tracing to explain causes, as well as analysis of other key variables alongside troop levels-such as intervener objectives, strategy, cultural affinity, and geography-to provide more detail and perhaps create typologies of stability operations and their requirements for different troop levels. Further advances could come from research done at the subnational level, which would increase the number of cases and allow for natural experiments where variables like location, force size, strategy, and percentage of foreign vs. domestic troops could be held constant or varied to examine the power of each.
Clearly, troop levels can only have a significant impact on stability operations in concert with their objectives, strategy, and overall quality. The outcome of the U.S. troop surge, among other factors, depends significantly on the ability to "clear and hold" territory and Prime Minister Nouri Al-Maliki's ability to deliver on his promise to send Iraqi troops to confront the militias. Nonetheless, policymakers and academics will continue to consider troop levels in isolation, the former to determine how many soldiers are needed for a given operation and the latter to determine the power of this explanatory variable concerning success in COIN and peacekeeping operations.
Whether American leadership ignores the conventional wisdom in this area (as the Bush administration did originally) or heeds it (as the Bush administration now seems more willing to do), the assessment and adjustment of troop levels in stability operations can and should earn more detailed attention from scholars and practitioners. Improved understanding and performance of stability operations now and in the future are at stake.