The State's Struggle to Master Asymmetrical Warfare

Purpose and Overview

The aircraft hovered over an old abandoned home in the deserts of rural Afghanistan. Inside, a group of insurgents braced themselves. Hours before, a convoy of trucks had been sprayed by gunfire. The insurgents responsible had fled to a small village miles away and sought refuge. They could hear the motor and knew what was to come. A split second later, the house exploded. Ironically, the year was not 2012, but 1918 as the British desperately held onto their crumbling empire. Nor was the attack conducted by an American drone, but with what the British considered a piece of untested technology that would revolutionize warfare – the airplane. The similarities between the American drone campaign against terrorism and the British Empire's "air control" doctrine are striking. They serve as a reminder that the clash of terrorists and insurgents against the state is an age-old problem that troubled the ancient city-state of Athens in its war against Melos as much as it does America today. Such wars are better known as asymmetrical wars. Lowther's defines asymmetrical warfare as follows: "It involves [a state] attempting to counter a style of warfare that is alien and unfamiliar to them conducted by an opponent that would not be able to compete if it fought in the same way that the [the state] preferred" (2007). By using Lowther's work as a foundation, this paper seeks to uncover the relationship between the length of asymmetrical warfare and its effect on the outcome of the war. Buried underneath these forgotten wars are clues to how the state can master this form of war. This paper attempts to discover some of these clues.

This study's hypothesis was that as asymmetrical wars get longer, the state's chances of victory dramatically decrease. The study's dependent variable was the outcome of the war, and the independent variable was the length of the war. I controlled for the type of war (colonial or imperial), who initiated the war, and whether or not another state intervened. The dataset used was the "extra-state warfare" set from the "Correlates of War" project. It contains almost two hundred asymmetrical wars going back all the way to 1817. Started by David Singer, an internationally recognized scholar in international politics, in 1963, the project has gone on to influence hundreds of political scientists. The creation of the project alone dramatically increased the quality of the University of Michigan's world politics program, and is considered one of the most reliable sources on the subject (2009).

When I ran a crosstabulation and chi-square tests on my variables, I discovered that the hypothesis was correct. As the war got longer, the state's chances of victory decreased. What I did not expect was how rapidly this transition would occur. There was virtually no difference in chances of victory for any war shorter than three years. However, if a war were to last longer than three years the state's chances of victory fell by 31%. It was almost as if a "three year switch" had been activated. I discovered that these findings worked in tangent with other scholars studies of asymmetrical warfare, explaining how insurgent groups around the world are evolving to effectively combat the state.

Literature Review

A Flawed View of Asymmetrical Warfare

As a line of British infantry men slowly marched forward on an open field, a group of American militiamen did something radical for the time period. Rather than marching in formation and fighting the world's strongest military head-on they utilized the forest as cover. Slowly, they picked off the British infantry until the enemy force collapsed. During our own revolutionary war, American militia-men engaged in what today is better known as "asymmetrical warfare." Ironically, these very tactics are being used against our forces in Afghanistan. The purpose of this literature review is to the guide the reader through the historical, theoretical, and empirical perspectives of the study of this unique form of war. In constructing this literature review, I discovered a gap in the data and research. Few studies have looked at the history of asymmetrical war as a whole, and none have focused solely on testing the relationship between the length of war and the state's chances of winning over hundreds of asymmetrical wars.

This study, and the theory it is trying to investigate, stems from a wealth of past research on what has been colloquially termed the "mathematics of war." Numerous studies have shown that there is a direct relation between insurgencies in asymmetrical war and patterns in the data that emerges from every suicide bombing, to each individual firefight between terrorists and the state (Gourley et al. 2009; Cederman 2002; Johnson et al. 2005; Johnston 2009). Throughout these studies, each team cited Lewis Richardson's trendsetting essay which first discovered war's relationship with power laws (1948). If the major wars of the 20th century were to be placed into a power law graph, the line would be diagonal, plotting major wars with less frequency and those of low intensity to be much more frequent. Richardson's study was just the beginning; soon physicists, military officers, and economists, banded together to develop models to explain the order that was appearing in the data. Considering that Richardson's conclusion states that low-intensity war or asymmetrical warfare will be more common, it is important to deploy statistical analysis to discover whether the state's attempts to master asymmetrical warfare are succeeding.

Power Laws and War

The empirical study of asymmetrical warfare embraces a little-used concept in political science – power laws. Power laws, according to Arthur Spirling, suggest a direct relationship between the frequency of an event and its size (2006). In other words, political scientists can look at the data of any war and accurately summarize its outcome, or even theorize and attempt to predict future events. War is one of the few subjects that have seen power laws used frequently. Armed with such knowledge, numerous studies have utilized power laws to discover that insurgents1 across the world operate in extremely similar ways.

Using the frequency of attacks and the number of causalities inflicted, Johnson et al. discovered that the counterinsurgency wars being waged in Colombia and Iraq showed the same power-law behavior (2005). In other words, the insurgents in both wars were executing the same model of war. They had evolved to effectively fight the state. In the study, both the insurgents in Colombia and Iraq had a very similar frequency of attacks. There were very few large attacks, but thousands of light "hit and run" attacks that slowly weakened the state (2005). For Johnson's team, this was a shocking development because the wars in Colombia and Iraq were started for different reasons, in geographically distinct places.

Taking this further, Sean Gourley et al. discovered the same phenomena in their study, except that they collected data from a much larger dataset and applied the power law. After reaffirming the study by Johnson's team, they continued to form a model for terrorist behavior. They theorized that insurgents who survive the onslaught of military power by the state have developed organizational structures that are united enough to effectively fight the state, but fragmented enough that they can never be easily targeted (2009). When they analyzed the power law relationship in the global war on terror, they discovered that Al Qaeda had developed such an organization model (2009).

Cederman details a practical analogy for power laws. In his study, he found that traditional ways of looking at war see world politics as billiard balls, when one ball slams into another it causes a chain effect. Power laws are much more like moving sand piles. Individually grains of sand shift in a pile, barely changing the status quo. Over time, however, these individual grains can create an avalanche that translates into significant political ramifications (2003). His analogy explains how a series of well timed, but small terrorist attacks can create massive power shifts. For instance, during the initial invasion of Afghanistan, the Taliban and Al Qaeda insurgencies were initially wiped out. However, over time, their collective hit and run tactics completely changed the tide of the war.

Based on the empirical studies listed above, the general consensus is that insurgents show universal patterns of behavior in asymmetrical warfare. In order to survive fighting against the state, insurgencies become fragmented enough so that they are hard to target and united enough that they can mount efficient counterattacks.

The Rich History of Asymmetrical Warfare

The historical perspective of asymmetrical warfare destroys the popular notion that the wars facing America today have no comparison and reveals why victory can be so difficult to achieve for the state. This notion emerges from America's historic short-term memory. Corum and Johnson compiled a detailed history of this forgotten past in Airpower in Small Wars (2003). The work points to America's experience in pursing Mexican rebel forces in 1916 by the U.S Army, the U.S. Marines fighting Dominican army irregulars in 1916, or the long asymmetrical war in Nicaragua that lasted for over a decade with over nine interventions (2007).

By 1940, the U.S. Marines had so much experience in asymmetrical warfare that they published the Small Wars Manual, which directly warned future military officers that the political and economic spheres were just as important, if not more so, than military victories in order to win in asymmetrical warfare (1940). Some chapters in the manual read as if it was written in the twenty-first century, until one turns the page and discovers a chapter on how to care for military horses.

The Tet Offensive highlighted this disparity between military might and political will. On one hand, the attack completely surprised the American forces, but was quickly contained, inflicting massive casualties on the communist North Vietnamese army. However, from a political perspective, the offensive turned the American people against the war. This historical example suggests that because the state actor is already seen as a more legitimate force, the standards of victory on both the military, economic, and political fronts are much higher than those of an insurgency.

In Sarah Sewall's introduction to the U.S. Army and Marine Corps' Counterinsurgency Field Manual, she directly calls into question the definition of victory for a state actor in asymmetrical warfare. Some analysts, particularly neoconservatives, believe in unleashing the full power of the United States military to completely destroy insurgencies and establish unquestionable victory. It is a black and white view on a complex type of warfare. Sewall counters that this "rejects the central truth that counterinsurgency is largely a political exercise. It also demands that Americans abandon their core values. To save ourselves, we would destroy our souls" (2007). This is one reason why the standards of victory are much higher for state; they are the legitimate actor with a set of core values; they are the ones who must answer to the global community if they commit horrific war crimes.

Bruce Hoffman, the Director of the Center for Security Studies at Georgetown University, further explains this higher standard of victory. According to Hoffman and Moshe Ya'alon, the former Chief of Staff of the Israel Defense Forces, insurgents and terrorists see states within the context of the "Spider-web theory." Like the state, spider webs appear to be strong and stable, but a simple motion of one's hand can cause the entire web to collapse. In other words, insurgents see citizens in the Western world as "self-satisfied, comfortable, and pampered to the point where they had gone soft" (2006, 64). When the Spider-web theory is combined with the democratic Western world, it gives insurgents a further advantage of eroding the public's support, making it easier to manipulate the standards of victory in their favor.

Is Asymmetrical War "New" or a Repeat of the Past?

Finally, the theoretical perspective of asymmetric warfare further erodes the many misconceptions about insurgents. John Robb illustrated one such misconception. Robb believes we are entering a "new" era of war. He argues that asymmetrical warfare has become empowered by globalization, adding to the misrepresentation of this "new" form of war. Insurgents utilized what is called "system disruption" to exploit the world's connectivity (2007). For instance, destroying an oil refinery has ripple effects not just in the target country, but also across the world. In the 1920s, America fought insurgents in Nicaragua who used similar tactics (Corum and Johnson, 2003).

Going a step further, the U.S. Army and Marin Corps Counterinsurgency Field Manual (2007) directly explains the theoretical paradoxes of counterinsurgency campaigns. For instance, the manual states that "some of the best weapons for counterinsurgents do not shoot" or "sometimes, the more force is used, the less effective it is" (2007, 49). There is no doubt that globalization has heightened the lethality of insurgents, but the U.S. military disagrees with Robb's bold hypothesis. By understanding that these wars are not "new" but "different," the military had begun unraveling the puzzle, but many of these new theories came after years of intense fighting in Iraq and Afghanistan. Sadly, the counterinsurgency manual does not cover whether or not the length of war will affect its outcome. The work focuses on the theory of counterinsurgency and the practical application of such ideas in the field.

After reviewing the current literature on asymmetrical warfare, it is suggested that my study design is unique, for few studies take into account the asymmetrical wars that go back centuries. Many studies describe complex models for how terrorists operate, but these models do not go into much detail about how the terrorists reach that point. In addition, the historical evidence suggests that there will be a direct relationship between the length of a war and its outcome. There is room within the literature for a study on the length of war to add to the discussion.

Study Design

Theory and Expectations

The following study design focuses on the following hypothesis: in comparing asymmetrical wars, those that are shorter in duration will be more likely to favor victory by the state than conflicts that are longer in duration. The null hypothesis of this research is: in comparing asymmetrical wars, conflicts that are shorter in duration will not have an impact on the state's chances of victory.

Consistently, each study in the literature review developed models that described insurgencies as self-organized groups, which constantly evolve and adapt to the state's tactics on the battlefield in order to survive (Gourley et al. 2009). This study design theorizes that such an insurgent group is designed from the ground up to trap the state in a long lasting war of attrition. This is especially effective against democracies, where long, bloody, and costly asymmetrical wars can turn the public against their leaders (Johnston 2009).

Since wars of attrition are unattractive to most militaries, many states attempt to set limited objectives and focus solely on counterterrorism rather than nation building. Such tactics follow the recommendations listed in Airpower in Small Wars, which analyzed the use of air forces to fight insurgencies throughout the 20th century (Corum and Johnson 2007). The study design seeks to show that by following such steps, states can limit their role in asymmetrical warfare and employ a "get in and get out" attitude, and thus significantly boost their chances of victory. This hypothesis is based on the assumption that the state tries to avoid a long-term war, as they have very limited resources, with a public that apply pressure regardless of government type. Based on these recommendations, and the notion that insurgencies purposefully organize themselves so that they can slowly bleed out the state, this study seeks to investigate just how effective shortening asymmetrical warfare is for the state.

Operationalization and Measurement of Concepts

This study will be based on the "Correlates of War" Project and utilize the "extra-state war" data set. This dataset tracks asymmetrical warfare from 1817 until 2007 (Sarkees and Wayman 2011). The dependent variable will be the "outcome" of the war. Outcome is coded into seven different possibilities: victory for the state, victory for the non-state actor, compromise, the war changes into another type of warfare, the war is ongoing as of 2007, stalemate, and the conflict has lessen in intensity so much so that it fails to be classified as a war. In the study, the researchers determined if the state won based on "common knowledge" at the time. Such information was taken from newspapers, military reports, actual battlefield reports, and other sources. Because outcome has seven different elements it is a nominal-level variable.

Our independent variable will be the length of time, in days, that the war lasted. The "Correlates of War" study breaks down the duration of the war based on the following variables: startday1, startmonth¹, startyear1, along with the corresponding end for the first period of fighting. Since asymmetrical warfare tends to see cycles of intense fighting, the study broke these variables into groups of start and end periods. Thus, other variables included startday2, startmonth2, and so on, which represented a return to fighting after a brief period of peace. The "Correlates of War" considered the "start" of a war based on when actual physical fighting occurred. As for the final "end," the project determined the ending based on the signing of peace agreement, retreat, or destruction of all enemy forces. No project, however, gives perfectly impartial data. Considering the vast amount of information, it is quite possible that the start and end dates may be inaccurate. However, considering the reputation of the "Correlates of War" project, and its existence for almost half a century, there is no better source to turn to.

This study wishes to take this information and condense it into one single variable: duration (in days) for every war in the study. Thus, on each war, I simply added together the total amount of days of the war. From there, I recoded the duration into an ordinal variable with three categories: "short wars" which will range from one to one hundred days, "medium wars" which will range from one hundred days to three years, and "long wars" which will last longer than three years. The longest war in the study was the Angolan-Portuguese War from 1961 to 1974, for a total of thirteen years.

Research Design

The "Correlates of War" project provides one of the most extensive data collections on war. The particular dataset being used contains almost two hundred extra-state wars from almost two centuries of warfare (Sarkees and Wayman 2011). The research design revolves around a chi-square test to evaluate the relationship between the two nominal variables while controlling for whether the war was "colonial" or "imperial" in nature, whether the war was started by the state (the initiator of the war), and finally whether or not other states intervened.

"Colonial" wars in the study involve a state engaging in warfare for the sake of defending territory in the empire. A good example would be the British tactic of "air control" – the use of air power to bombard a population into submission – against insurgents in what was then its territory in Iraq in the 1920s. "Imperial" wars on the other hand, involve a state invading a territory for the sake of conquering it militarily, culturally, and/or economically. For instance, the dataset considers the American war in Iraq to be an "imperial" war. The study design seeks to see if states have historically performed better in colonial wars when placed in a more defensive role, rather than an offensive role in imperial wars (Sarkees and Wayman 2011).

This study design will utilize the chi-square test and run crosstabulations. This test will showcase whether the null hypothesis has any standing and assists in discovering the goodness-of-fit between what is observed in the data and what the study design expects. First, I will run a chi-square on my main independent and dependent variable. Following this, I will employ the same tests on my controls to see if there is any statistically significant relationship. The chi-square test is one of the best possible methods to discover the probability of state victory given certain variables. In addition, if the null hypothesis is correct, it would be possible to convert other variables into nominal ones and run the test again to attempt to uncover what exactly assists the state in winning asymmetrical wars.

Analysis and Conclusion

Analysis

In the previous sections, I predicted that the longer the state engages in asymmetrical warfare the less likely they will achieve victory. The notion was that as states expand their objectives in an asymmetrical war, such as state building and spreading democracy, they leave themselves vulnerable to becoming trapped in their own goals as a war of attrition sets in. Using the extra-state Correlates of War dataset, I ran a crosstabulation on the relationship of the duration of the war compared to the outcome of the war.

As stated before, the dataset splits outcome into seven groups. "Side A Wins" represents a victory for the state, and "Side B Wins" represents a victory for a non-state actor. I decided against viewing the outcome of the war in black and white. Due to the research done by the Counterinsurgency Field Manual, Bruce Hoffman, combined with the practical experience of Moshe Ya'alon, and the "Spider-web theory" discussed in the literature review, it is highly supported that the standards of state victory are much higher than those of an illegitimate insurgency.

By applying this research to the "Correlates of War," I considered the following values as losses for the state: the insurgents earn a clear victory, a compromise, having the war transfer into another type of war, the war is ongoing as of 2007, and a stalemate. If the state must compromise with a non-state actor, the legitimacy of the government is potentially ruined, resulting in a political defeat. If the war is transformed into another type of war, it means the insurgency has amassed so much support that they can engage the state in conventional battle. Thus, the state has failed to contain their opponents, and this once again represents a form of defeat. The final value is wars that are ongoing as of 2007. These wars are the recently ended war in Iraq, and the current war in Afghanistan. The war in Iraq recently ended with the United States pulling out of the country with a list of objectives not accomplished, leaving behind a fragile government. Afghanistan's future is in jeopardy. With a full-scale withdraw planned for 2014, Afghanistan's military still barely functions and our own soldiers are being shot by Afghan military-trainees in horrific "green-on-blue" situations. Thus, both wars fall under the categories of defeat for the state or stalemate, and I considered these as a victory for the non-state actor. Finally, I considered stalemate as a victory for the non-state actor since the state's inability to end a war against insurgents also ruins its legitimacy.

At first, it may appear that I am stacking the deck in favor of my hypothesis by considering six out of the seven outcomes in the dataset as victory for the insurgency. However, considering the mass differences between the well-armed, well-funded, and politically legitimate state, and the typical underfunded armed forces of an insurgency, I stand by my assessment. Furthermore, the state is far more vulnerable to defeat because it has so much more to lose than an insurgency, which has everything to gain by winning, at any cost. This was seen in the works of Sewell and the spider-web theory. Thus, the only way for a state to achieve victory over an insurgency is to completely defeat them. For the insurgents, however, the threshold of victory is much smaller since they have little to lose.

After running a chi-square test, I can reject the null hypothesis that there is no relationship between the length of a war and the states chances of victory. With a p-value of .000 it is clear that a relationship exists between my independent and dependent variables. My Pearson Chi-square value totaled at 50.515, a very high rate that reinforces the p-value. The R square PRE measure of association was .075, giving the relationship a "weak" strength. The chi-square test suggests that the chances of state success after three years of war drops dramatically. Out of fifty long asymmetrical wars, only nineteen nations have achieved victory since 1817.

My results, however, held a few surprises. Ironically, proportionally speaking, states have fared equally well in both short and medium length asymmetrical wars. In asymmetrical wars lasting less than hundred days (termed as "short") the state won 72% of the 96 wars in this group. For wars lasting between hundred days and 3 years (termed as "medium") the state won about 69% of the 51 wars in this group. Before I ran my test, I had assumed that short wars would show an overwhelming percentage of state victories, whereas medium length wars would show a decrease in state victories. This did indeed occur, but by a minor three percent. In war a switch seems to be triggered after three years of warfare. The question is what is this switch that sees a dramatic drop to only a 38% chance of victory?

To further test these results, I selected three control variables: the type of war, whether another state intervened in the war, and whether the state started the war. The most dramatic conclusions emerged from the type of war. The dataset considered wars like the U.S. invasion of Iraq as imperial and other wars such as the British Empire's air control wars as colonial. Imperial wars show a significant increase in the state's chances of victory during medium wars. This may occur because the insurgency has not been able to effectively rally the support of the local population. Colonial wars follow my original hypothesis, but the "three-year switch" effect is not as dramatic. This may be related to the fact that in a colonial war the state has already established a framework or working relationship with the local population. The other two controls show supporting evidence for my original hypothesis. The only control that was not significant was the not having another state intervene in an asymmetrical war.

The Three Year Switch

Carl von Clausewitz once said that "in war even the simple things are difficult" (in Corum and Johnson, 2003, 425). The "simple" task of putting down an insurgency, according to Corum and Johnson in their historical and theoretical analysis of asymmetrical war starts with a comprehensive strategy. By establishing strategy for not just for the military sphere, but social and economic spheres of asymmetrical war, the state increases their chances of victory (2003, 425-427). Time and time again, the state assumes that asymmetrical warfare can be won in the military sphere alone, for instance Vietnam. Almost exactly three years into the war, the Vietcong insurgency launched the Tet Offensive. As stated previously, the attack itself was a massive military victory for the United States. However, back home in the United States it enraged the public, and in Vietnam itself the local populace began to show signs of turning against the Americans. By failing to remember the other sides of asymmetrical warfare, it can be argued that American had set itself up for failure.

Johnston (2009) analyzes the effectiveness of killing the leadership of an insurgency. His research suggests that if the state focuses on decapitation of an insurgency's leadership, this will increase the chances of state victory by 46% (2009). The removal of leaders in an insurgency can decrease the length of the war as well. By surgically targeting the enemy, limiting the objectives, and adopting a "get in and get out" strategy, the state can increase the likelihood of victory.

These results also fit into the models created by many scholars that showcase insurgents as organic decentralized groups that find a delicate balance between unity and fragmentation (Gourley et al. 2009). By being united but fragmented, insurgencies can strike their targets and remain obscure. As Gourley and his team summarize their model as seeing insurgents as an "ecology of dynamically evolving, self-organized groups following common decision-making processes" (2009, 1).

It is this model of organic organization that makes insurgents so deadly. The question of what triggers the so-called "three year switch" remains empirically unanswered and requires additional avenues of research. But, if I were to make another hypothesis, it would be the suggestion that the "three year switch" and the work of scholars like Gourley are directly related. These insurgent groups reach this deadly combination of fragmentation and unity after three years of continuous warfare. The R-square measure of my study showed a weak relationship between the duration of war and the outcome of it, suggesting that the length of war activates other variables, such as the organization model, that decrease the state's chances of winning. If an insurgency can last this long in battle against the state, it has the funding, weaponry, and most importantly, human capital to pull off a war of attrition. In other words, insurgents can evolve to effectively fight against the state and win. Once this stage is activated, the state's political will or resources for war will deplete. This organically organized model could also explain why there is relatively no difference for the state's chances of success between "short" and "medium" length wars, since it seems that insurgent groups reach this level of organization after three years.

In the state's quest to master asymmetrical warfare, it would be suggested that the state adopt strategies to prevent insurgents from finding the "sweet spot" between unity and fragmentation. By pursuing a limited agenda that embraces a "get in and get out" mentality, with no nation building plans or occupation of the targeted country, the state can increase their chances of victory within three years. By providing empirical evidence that suggests a so-called "three year switch," this study advances our understanding of asymmetrical war by a slight extent, and strengthens the models created by other scholars.

References

Bard e. O'Neill, Insurgency and Terrorism: Inside Modern Revolutionary Warfare Washington, DC: Brassey's inc, (1990): 13.

Cederman, Lars-Erik. "Modeling the Size of Wars: From Billiard Balls to Sandpiles." American Political Science Review. 97. no. 1 (2003): 135-150. http://journals.cambridge.org/action/displayFulltext?type=1&fid=142726&jid=PSR&volumeId=97&issueId=01&aid=142725 (accessed February 17, 2012).

Clauset, Aaron. "On the Frequency of Severe Terrorist Events." Journal of Confiict Resolution. 51. no. 1 (2007): 58-88. http://arxiv.org/pdf/physics/0606007v3.pdf (accessed February 17, 2012).

Corum, James, & Johnson, Wray. Airpower in Small Wars: Fighting Insurgents and Terrorists. University Press of Kansas, 2003.

Gourley, Sean, Bohorquez, Juan, Alexander Dixon, Michael Spagat, and Neil Johnson. "Common Ecology Quantifies Human Insurgency." Nature, December 17, 2009. http://www.nature com/nature/journal/v462/n7275/pdf/nature08631.pdf (accessed February 17, 2012).

Hoffman, B. (2006). The logic of suicide terrorism. In R. Howard, J. Forest & J. Moore (Eds.), Homeland Security and Terrorism New York : McGraw-Hill.

Johnson, N., Spagat, M., Restrepo, J., Becerra, O., Camilo, B., Suarez, N., Restrepo, E., & Zarama, R. Universal patterns underlying ongoing wars and terrorism, 2005.

Johnston, Patrick. "The Effectiveness of Leadership Decapitation in Counterinsurgency." International Security and Social Science Seminar. 2009: 1-53. http://iis-db.stanford.edu/evnts/5724/Johnston_-_Decapitation_(CISAC).pdf (accessed February 17, 2012).

Lowther, Adam. Americans and Asymmetrical Conflict: Lebanon Somalia, and Afghanistan. Westport, CT: Praeger Security International, 2007.

Richardson, Lewis F. 1948. "Variation of the Frequency of Fatal Quarrels with Magnitude." American Statistical Association 43: 523-546.

Robb, John. Brave New War: The Next Stage of Terrorism and the End of Globalization. Hoboken, NJ: John Wiley & Sons, Inc, 2007.

Spirling , Arthur. 2009. "The Next Big Thing: Scale Invariance in Political Science." Retrieved from http://www.mendeley.com/research/next-big-thing-scale-invariance-political-science/

University of Michigan. (2009). David singer. Retrieved from http://sitemaker.umich.edu/jdsinger/home.

U.S. Army, Marine Corps. Counterinsurgency Field Manual. Chicago, IL: The University of Chicago Press, 2007.

Endnotes

1. This paper accepts Bard O'Neil's definition on the relationship between insurgency and terrorism. O'Neil defines insurgency as "a struggle between a non-ruling group and the ruling authorities in which the non-ruling group consciously uses political resources and violence to destroy, reformulate, or sustain the basis of legitimacy of one or more aspects of politics" (1990). One such tactic of insurgents is terrorism. In other words, in today's conflict against terrorism the lines between terrorists and insurgents can be hard to distinguish. As a result, "terrorist" and "insurgent" are used interchangeably in this paper.

Appendix

Tables and Figures