Book Review: America’s Modern Wars: Understanding Iraq, Afghanistan, and Vietnam
Shawn Robert Woodford
In the wake of the United States’ major counterinsurgency efforts in Iraq and Afghanistan, nearly every aspect of those interventions remains hotly debated. Despite a plethora of explanatory theories and analyses, observers are no closer to understanding the true nature of these conflicts and the effectiveness of the approaches used to wage them than they were when the wars started. This is partly due to a lack of a detailed and comprehensive understanding of the phenomenon of insurgency. While there are many useful and insightful works that address insurgency and counterinsurgency, most are derived from the examination of a single case, or, at best, the comparative study of a handful of examples. Generalizing from limited data samples has neither provided wholly definitive nor persuasive results.
America’s Modern Wars: Understanding Iraq, Afghanistan, and Vietnam (Casemate, 2015, 376 pages, $32.95) by Christopher A. Lawrence offers a look at the possibilities offered by a broader, more systematic approach to the subject. The book summarizes the findings of several analytical studies on insurgency and counterinsurgency undertaken by The Dupuy Institute (TDI) between 2004 and 2009[i], sponsored by the U.S. Army Center for Army Analysis; Office of the Secretary of Defense, Net Assessment; and the National Intelligence Council of the Central Intelligence Agency.
As Lawrence relates in an informal, Socratic-style narrative, the effort began with TDI’s unsolicited proposal in 2004 to the Center for Army Analysis to estimate the likely number of fatalities U.S. forces would suffer in the Iraq conflict. TDI’s estimate, delivered in December 2004, predicted that the insurgency would last approximately 10 years and U.S. forces would suffer no fewer than 3,000 troops killed by all causes, most likely between 5,000 and 10,000 fatalities. When U.S. forces withdrew from Iraq in 2011, the conflict had persisted for 8.7 years and 4,485 American military personnel had been killed, 3,436 of them in combat. Impressed by the methodology behind the estimate, several U.S. government organizations contracted with TDI for further data collection and follow-on studies.
This resourcing enabled TDI to amass a data set unique in size and detail. Lawrence, TDI’s Executive Director and President, directed a team of a dozen researchers, quantitative analysts, and consultants, who combed a vast array of primary and secondary historical sources to assemble a database dubbed the Modern Insurgency Spread Sheets, or MISS. The Iraq estimate had been derived from data collected on 28 post-World War II insurgencies, interventions, and peacekeeping operations. Further research expanded the MISS to 83 complete cases. (An additional 26 partially completed cases were not used for analysis.) Each case incorporated data for at least 75 variables, including basic quantitative information such as country size, area, border length, population, and population density, as well more complex information like conflict duration, insurgent and counterinsurgent force size, losses, and numbers of violent incidents – much of it broken down by year. Qualitative variables ranged from the indigenous and intervening government types, insurgent motivation, and insurgency structure, to the type of counterinsurgency waged, counterinsurgent rules of engagement, and the nature of the conflict’s end.
Lawrence identifies two variables with particularly high correlations to outcome in the data set: force ratios and insurgent motivation. His analysis suggests that the result of an insurgency closely tracks the force ratio of counterinsurgents to insurgents. This correlation held up whether the data set was filtered for purely domestic insurgencies, third-party interventions, or “colonial wars of liberation.” From this, Lawrence concludes that while overwhelming numbers are not required to defeat an insurgency, force ratios above 10-to-1 pretty much guarantee a counterinsurgent victory. Conversely, relatively fewer counterinsurgent forces do not preclude success, but force ratios below 2-to-1 greatly favor an insurgent victory.
The other variable that demonstrated a strong association was the nature of the motivation for the insurgency. Again, regardless of the way the data set was filtered, counterinsurgents usually succeeded when facing insurgencies based on limited political goals or rooted in ideology. Insurgencies with a broadly appealing rationale, such as nationalism, however, were nearly always successful.
When force ratios and motivation were assessed together, Lawrence found that the former had little impact on the outcome of insurgencies with a limited political or ideological basis. On the other hand, when facing broad-based insurgencies, counterinsurgents lost every time they possessed a force ratio advantage of 5-1 or less, failed half the time with odds between 6-1 and 10-1, but succeeded three-quarters of the time when outnumbering the insurgents by 10-1 or more.
TDI’s research suggested several other variables that might also have significant influence on outcome, but these could not be analyzed in detail before its government patrons stopped funding the project in 2009. These include rules of engagement, terrain, counterinsurgent burden, insurgent strategy, and indigenous government type and elections. Lawrence and his team also had a list of potentially influential factors that they had yet to examine in detail, including insurgency structure, government political reform, outside support, sanctuaries, barrier systems, population resettlement, and staying the course.
One uniquely useful aspect of quantitative analysis is the ability to report a null result, which can be as telling as a correlation. Lawrence tested several counterinsurgency verities that turned out to be questionable in light of historical experience. For example, the assertion of a connection between the outcome of an insurgency and the ratio of counterinsurgents to the general population could not be substantiated. Some quite surprising results turn up as well, such as the fact that France endured much higher killed per capita before conceding defeat in its post-WWII counterinsurgencies than the U.S., Great Britain, or the Soviet Union.
Lawrence used these findings to develop a logit regression model for predicting the outcome of insurgencies based on force ratio and insurgent motivation. Despite its simplicity, the construct provides a powerful tool for analysis, which Lawrence applies to the conflicts in Iraq, Afghanistan, and Vietnam.
In Iraq, Lawrence asserts that the U.S., Coalition, and Iraqi government success was due to the generation of sufficient collective counterinsurgent forces to achieve a favorable overall force ratio against what he deemed a regional insurgency. The key to this, in Lawrence’s estimation, was the Sunni Awakening in 2006-2007, which removed large numbers of insurgents from the conflict while simultaneously boosting the ranks of Iraqi security forces. This quickly and decisively tipped the force ratio in favor of the counterinsurgents. Lawrence contends that this took place largely independently of the 2007 U.S. troop surge and indicates the often-overlooked effectiveness of political policy-based solutions in countering insurgencies.
When applied to Afghanistan, the model indicates that the counterinsurgents should have been successful in bringing the Taliban insurgency under control by 2008. If not by then, then surely the U.S.-led Coalition surge that began in 2009 should have achieved success. Yet, this clearly was not the case. Lawrence’s analysis suggests two explanations: first, that the Coalition was probably underestimating the number of insurgents, perhaps by as much as 60%. The second was that the Coalition likely underestimated the appeal of the insurgency, and that it is far more popular than thought among the Afghan people, particularly among the Pashtun population that is the Taliban’s base of support. Either of these conclusions would indicate that to defeat the insurgency, the Coalition would need to maintain its forces at the 2009-12 surge levels, or even higher, for several more years.
Lawrence’s assessment of the Vietnam War arrives at similar conclusions. Based on the model, the U.S., South Vietnam, and their allies committed troops in sufficient numbers to have brought the Viet Cong insurgency under control by 1968 if the motivation of the insurgents was rooted in a factional or ideological basis. The fact that this did not happen lends support to the contention that the insurgency was based on a broader nationalistic appeal than was appreciated at the time. Simply put, the U.S. did not commit enough resources long enough to bring the insurgency under control. The level of American political commitment needed to mobilize the necessary wherewithal to defeat a nationalist insurgency in Vietnam likely made the war unwinnable in Lawrence’s judgment.
Despite the ample evidence for its conclusions, America’s Modern Wars is unlikely to end the ongoing debates, and it may well prove controversial. Lawrence concedes that his work could be revised based on new data. Though the significant sample size of the MISS and statistical testing of Lawrence’s analyses suggest his conclusions are firmly rooted, quantitative techniques will undoubtedly continue to be viewed with suspicion by traditional historians and military analysts. To be fair, there is more than enough poorly executed quantitative analysis associated with past conflicts to warrant skepticism. However, a widespread belief also exists that warfare is far too complicated a phenomenon to be quantified. This would be unfortunate, as Lawrence has sketched out a variety of promising directions future research could profitably pursue. One of the biggest U.S. government failures the last time a better understanding of insurgency and counterinsurgency warfare was desperately sought, Lawrence concludes, was the decision to abruptly cease support for nascent research efforts during the Vietnam War.
[i] I was employed as a research analyst at TDI from 2004 to 2006 and participated in the initial stages of data collection and research for the presentations and reports described in the book.