By Steven M. GoodeMarch 25, 2010
Over the last eight years, one question has repeatedly come up in regard to the wars in Iraq and Afghanistan: How many soldiers are enough' The question was first raised before the Iraq war started, with highly publicized disagreements between senior military leaders regarding the number of forces needed to secure Iraq after the invasion. The debate reached another peak when the "surge" strategy was announced. It has once again become the subject of national discussion, this time with respect to Afghanistan.
Despite years of debate, our understanding of force requirements for counterinsurgency has advanced little since 1995, when James Quinlivan of RAND published a seminal article on the subject.
The current article describes work done by the Center for Army Analysis (CAA) to better inform the discussion by examining historical data related to counterinsurgencies. The intent is not to make any policy recommendations. Nor should this analysis be interpreted to suggest that force levels alone are the key to victory in counterinsurgency. Having enough military forces is a necessary, but by no means sufficient, condition for success.
The twentieth century is replete with examples of counterinsurgents winning Pyrrhic military victories that resulted in political losses. The French in Algeria and the Portuguese in Mozambique and Guinea-Bissau prevailed militarily but ultimately acceded to the insurgents' demands for independence. The British devoted enormous forces to Cyprus and suffered relatively few losses but nonetheless had to give up control of most of the island.
That said, however, force levels do matter, and history can provide a guideline for force requirements in counterinsurgency. The analysis described in this article shows that there are three major drivers of military requirements. First, as previous studies have argued and current doctrine emphasizes, security forces have to be sized relative to the population. Second, the more intense the insurgency, the more forces are required to reverse increasing insurgent violence. Third, the larger the percentage of personnel that are drawn from the host nation, the fewer forces will be needed overall. Before detailing the exact relationships between these factors and force requirements, or discussing their implications for US policy, it is appropriate to briefly review the current state of the debate.
Current doctrine as contained in an Army and Marine Corps field manual says the following about force levels in counterinsurgency:. . . [N]o predetermined, fixed ratio of friendly troops to enemy combatants ensures success in [counterinsurgency] . . . . 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 1,000 residents in an [area of operations]. Twenty counterinsurgents per 1,000 residents is often considered the minimum troop density required for effective [counterinsurgency] operations; however, as with any fixed ratio, such calculations remain very dependent on the situation.
These ratios appear to be based on Quinlivan's work. He emphasized, though did not originate, the idea of sizing security forces to the population rather than the enemy. Further, the recommendation of 20 to 25 counterinsurgents per 1,000 residents appears to be based at least in part on his observations that British forces in Northern Ireland and Malaya peaked at about 20 per 1,000 residents, and that international forces in Bosnia and Kosovo reached levels between 20 and 25 per 1,000.3 As Quinlivan noted, however, those cases represent only part of the scale. History also includes cases such as the postwar occupation of Germany, where successful stability operations were mounted with only 2.2 security forces per 1,000 residents. In contrast, the counterinsurgents under French command in Algeria peaked at nearly 60 per 1,000 residents, and the Russians committed more than 150 soldiers per 1,000 residents in Chechnya in 2003.
The other major study on this topic was by John McGrath.4 McGrath's major contribution was the observation that counterinsurgent forces need to dedicate a portion of their personnel to police duties. Based on his calculations of police-to-population ratios in several major American cities, McGrath suggested a figure of 4.1 soldiers per 1,000 residents should be devoted to police duties. This figure may not be precisely appropriate for all nations, but McGrath's insight is valuable because it underscores the need to include police functions in any discussion of security force requirements.
Data and Definitions
Past efforts on this subject have been hampered by the relative lack of data. What information was available on past counterinsurgencies has been scattered across a number of sources. The CAA identified three problems caused by this lack of data. First, it has limited the number of case studies available for examination. Quinlivan and McGrath together looked at only 15 cases. This contrasts with the dozens of insurgencies, peacekeeping operations, contingency operations, and the like that have taken place since 1945. The second problem is that while not all conflicts are equally intense, there has been no agreed-upon measure of intensity. If, as this article will show, more intense insurgencies require larger counterinsurgent forces to defeat them, then the intensity of the insurgency becomes a key driver of force requirements. Third, the lack of data has prevented any significant analysis of the relationship between force levels and intensity over time. It can be misleading to use peak military numbers, as force levels often continue to rise for a time after an insurgency is largely defeated. In order to remedy this lack of data, CAA built an Irregular Warfare Database containing year-by-year data on more than 100 conflicts. The analysis described in this article was built on that database.
The analysis of irregular warfare has also been bedeviled by a profusion of terms of overlapping scope. Quinlivan examined "stability operations," while McGrath studied "contingency operations." Each term covers a wide range of conflicts and operations. Of the 15 cases studied by the two authors, three (Germany, Austria, and Japan) were peaceful post-war occupations of defeated powers. Six (the Philippines, Malaya, Northern Ireland, the Punjab, Iraq, and Afghanistan) can be described as counterinsurgencies.
Six (the Dominican Republic, Haiti, Cambodia, Somalia, Bosnia, and Kosovo) were stabilization, humanitarian, or peacekeeping operations. The problem with grouping such disparate situations is that the force requirements for different kinds of operations may be very different. The occupation of Germany after World War II could be conducted with a relatively low force-to-population ratio because the Germans had been soundly defeated and offered essentially no resistance to the occupying forces. Similarly, in many peacekeeping operations, the peacekeepers are present with the consent of the local population and face relatively little armed opposition. In contrast, the insurgencies in the Philippines and Malaya required considerably larger forces...
To continue reading the entire article, please click here to download the PDF.