May 1, 2008  

Lies, damned lies and counterinsurgency

Not all insurgencies have been protracted affairs

It has become a matter of conventional wisdom that insurgencies last an average of 10 years and that the insurgents win about 40 percent of the time. These statistics have appeared in USA Today, PBS, Pentagon media briefings and on National Public Radio. The insight these numbers are meant to convey is that counterinsurgencies are inherently long and difficult struggles against wily and resilient foes, so it is unrealistic to expect rapid, quantifiable progress in the near term. Fortunately, these statistics are misleading and the associated analysis is wrong.

The source of this mistaken conventional wisdom is the prestigious Dupuy Institute, which has been providing rigorous quantitative analysis to the military for more than 40 years. In May 2007, Dupuy researchers published the preliminary results of a study in which they examined 63 modern insurgencies for a variety of factors, including the longevity and the success rate of the conflicts. Given their analytical talent and track record of precision, their statistical computations are undoubtedly accurate. The problem, however, isn’t with their math; it’s with the initial selection of cases.

According to the FM 3-24 Counterinsurgency doctrine manual, an insurgency is “an organized movement aimed at the overthrow of a constituted government through the use of subversion and armed conflict.” In the past 100 years, there have been considerably more than 63 movements that would fit that definition, so to create a manageable data set, only the most violent and intense conflicts are likely to be included. However, because casualty counts are often a function of time, this method naturally trends toward long wars and excludes cases in which government forces crushed the nascent insurgency before it even got off the ground.


Suppose you decided to evaluate the average impact of house fires. Knowing you’d never be able to count all the little kitchen and grill fires that are immediately extinguished, you chose instead to hang out at the fire station and only track conflagrations that resulted in a 911 call. Every day you rolled out of the station, you’d find a mix of cases: some small fires, but also some really big ones. However, your count would be skewed, because you were seeing all of the really damaging infernos but only a limited portion of smaller blazes.

After you came back and averaged your results, you’d probably conclude that house fires are a more destructive force than they really are because your list of cases included a litany of charred buildings and hospitalizations, but excluded minor flare-ups. But, of course, you wouldn’t really be talking about the phenomenon of things catching on fire inside a house, you’d only be talking about the fires you’d bothered to count.

To extend the analogy a bit, because you’d weighted your sample toward the big fires, you might be seriously off base in your policy recommendations. Rather than focusing on education, smoke detectors and kitchen fire extinguishers, you would probably want more fire hydrants, bigger fire stations and a fleet of shiny new trucks. You would never talk about how to prevent little fires or how to stop them from spreading once they started, because they’d never really entered into your analysis in the first place. You would take for granted that fires are cataclysmic, and the initial choices that limited your sample would be forgotten in the pronouncements about the scourge of burning homes threatening our nation.

The big fires of 20th century insurgency generally fall into two broad categories (with some notable exceptions): post-World War II national liberation movements and superpower proxy wars. These conflicts tended to last a comparatively long time, inflict substantial casualties and frequently resulted in an insurgent victory. As a result, they have a disproportionate impact on a survey of insurgencies that selects cases by length and intensity. But in some ways, these are two unique categories that contain exceptionally intense wars for very specific historical reasons.

World War II sounded the death knell of the global colonial regime. For centuries, European powers sat atop a system of racist exploitation in which white settlers and colonial administrators extracted the labor and natural resources of the rest of the world. While the specifics differed from place to place, the general outline was the same: Violence and bureaucracy was used to circumscribe the opportunities available to the native population; force them into dirty, dangerous and difficult work; and ensure the privileges of their European masters. After 1945, however, two important conditions changed: the willingness of European nations to tolerate genocide in their name and their capacity to project power abroad.

Before the war, most rebellions in the colonies were dealt with through indiscriminate oppression, collective punishment and a cornucopia of other crimes against humanity with nary a peep out of the Old World metropolitan centers. After the war, similar behaviors, such as the French torture regime in Algiers, elicited widespread protest instead of general disinterest. This public outcry over human rights abuses was, in many cases, a key factor in the eventual withdrawal of colonial forces.


Moreover, the enormous damage wrought by five years of war on the European continent left its members with a limited capacity to undertake new foreign adventures. Indigenous rebels witnessed the limits of colonial power, saw their masters’ regimes collapse during the 1940s and were in no mood to see them restored once the Axis was defeated. Anti-colonial insurgencies spread like wildfire, and the only truly effective method to prevent the insurgents from seizing power violently seemed to be guaranteeing a transition to local sovereignty (this was the British technique in Malaya and Kenya). The alternative was an unaffordable and endless expenditure of national resources for what had become an unprofitable and unpopular system. In fact, the attempt to sustain the costs of the old colonial regime led to the collapse of governments in France and Portugal. The system that had lasted for centuries was largely swept away in a mere three decades, and not a single new colonial possession has been established since.

However, as the colonial era was coming to a close, there was a global wave of proxy wars as the superpowers struggled to place friendly indigenous clients at the head of the newly created post-colonial states. While the concept of pitting indigenous forces against one another is quite old, the scale and type of armaments poured into to these Cold War hotspots defied all historical precedent. Occasionally, as in Afghanistan and Vietnam, local proxies faced the forces of the opposing superpower directly. But often involvement was limited to a seemingly endless supply of advisers, funds and arms. As a result, conflicts could grind on for decades, such as in Angola or Mozambique, or could survive the total absence of any local support, as with the Nicaraguan contras. The only limitations were superpower public opinion, regional basing rights and the population of poor, desperate young men with no futures.

Obviously, not all these conflicts lasted until the end of the Cold War. As Che Guevara discovered in the Congo, rebel organizations were highly vulnerable to collapse due to infighting, lack of commitment and overwhelming government force. Nonetheless, the ability of insurgents to hide out in border havens, draw on an external source of supplies and reconstitute depleted formations with monetary incentives generally exacerbated the length and intensity of these wars.

What is important to note about colonial revolutions and superpower proxy wars is the fact that they last a long time and the insurgents have fairly good odds is a function of their global political environment. It is not that “insurgency” is an overwhelmingly effective strategy that can topple states willy-nilly. Rather, in the specific contexts described above, local armed opposition to the government has a rather unusual set of advantages. These are big fires because external forces fanned the flames.

By contrast, including the little fires brightens the picture for counterinsurgents considerably. In “Guerrillas and Revolution in Latin America,” Timothy P. Wickham-Crowley considers 25 revolutionary movements that occurred in Central and South American from 1956 to 1990. Selecting a fixed geographic area and timeframe may impose its own form of bias, but it has the advantage of including large and small insurgencies. The results of this more inclusive technique are striking. Of these movements, only two succeeded in replacing one system of governance with another. Moreover, their lengths have a bimodal pattern, which means that there are really two sets of averages. One set, the protracted conflicts, which include Peru, Colombia, El Salvador and Nicaragua, last for decades. But in the other set, which includes Uruguay, Brazil, Argentina (three times), Honduras, Bolivia and the Dominican Republic, the results are very different. In this larger group, the state crushes armed opposition in about 3 years — these wars are often over in a matter of months. And this analysis excludes coup attempts, which tend to either succeed or collapse within days.

The extremely high failure rate of insurgency is intuitive. Heading off into the woods with a handful of fighters and some small arms to take on the repressive apparatus of a modern state is a dicey proposition. Very specific factors have to coalesce for the endeavor to succeed. Archetypical revolutionary figures such as Fidel Castro or the Ayatollah Khomeni led multiple failed movements before the Cuban and Iranian governments became so brittle and alienated so many of their elites that they became vulnerable to collapse. As guerrillas across South America throughout the 1960s and 1970s discovered, even a talented and experienced insurgent is a relatively easy target for government forces. Most times and in most places the government deters or defeats armed challenges to its authority and does so relatively quickly and easily.

But that simple fact is obscured by the rhetoric of the 10-year average and the 40 percent success rate. And therein lies the truly dangerous effect of this flawed conventional wisdom: It prevents a detailed discussion about the insurgencies we face in Iraq and Afghanistan, their causes and consequences, and what steps can be taken to move them from protracted conflicts to stable governments.

There are, of course, many reasons that the wars in Iraq and Afghanistan continue on, and the pernicious effect of the statistical sleight-of-hand we’re discussing is that we really never publicly analyze those factors. There are a lot of things that seem to affect the length of an insurgency — government capacity, political context, popular will, external forces, resource availability, geography, etc. Simply asserting that insurgencies naturally take a long time, because it is in their very essence to be protracted and difficult, precludes a more detailed discussion. According to conventional wisdom, we will be in Iraq and Afghanistan for at least 10 years because they are insurgencies, and that’s how long insurgencies take — full stop.


This is an abdication of the responsibilities of strategic leadership. The American public is owed more of an explanation than, “Well, these things take a while.” It is owed a comprehensive strategic vision that clearly explicates the factors that are drawing the conflict out, which are amenable to change, how that change will be implemented, what the expected impact will be, how long it will take, and what the implications of a lack of progress are. Part of the basis for that vision needs to be a more robust model of insurgency and a dedication to speaking precisely when using statistics to explain things.

Adopting this approach would be extremely beneficial to public discourse and policy analysis. By explicating the ways in which contemporary insurgencies are like some historical conflicts and unlike others, we will not only gain insight into how long they will last and the likelihood of an insurgent victory. We will also be able to have a more meaningful discussion about the specific vulnerabilities of the insurgents in Iraq and Afghanistan, the most appropriate methods for exploiting those weaknesses, and the best way to achieve a stable post-conflict system of governance. Rather than a “one-size-fits-all” approach to counterinsurgency, we can build solutions that are responsive to local conditions and also rooted in historical best practices. But the first step is rejecting unhelpful, misleading statistical crutches that take the place of a more rigorous analysis.

It is simply untrue that counterinsurgencies are inherently long-term struggles. History is replete with insurgencies that were resolved quickly and seemingly intractable struggles that were brought to a rapid conclusion once the political will was mustered. If Iraq and Afghanistan share features with historical protracted insurgencies, then a clear understanding of those factors and their import is a necessary step for prosecuting an effective counterinsurgency. Moreover, understanding why we expect them to last so long may illuminate possibilities for shortening the conflicts and increasing the likelihood of success. But to achieve any of this, strategic thinkers must assiduously avoid the tempting shortcuts of lies, damned lies, and statistics.

ARMY CAPT. ROBERT M. CHAMBERLAIN has completed two tours in Iraq, most recently on an adviser team in Tal Afar. The views expressed here are the author’s own and do not necessarily reflect those of the Army or Defense Department.