Jonah Schulhofer-Wohl
“Dynamics of Civil Wars: the Causes and Consequences of Subsidies to Armed Groups”
Chapter 1: Introduction
The extent of variation in the duration of civil wars and the lack of comprehensive explanations for it are explored. This chapter reviews the main, existing explanations and lays out their problems, emphasizing the lack of mechanisms and faulty logic where mechanisms are proposed. It also shows the disconnect between the theories and the reality of long wars through comparative anecdotal evidence about civil wars.
Chapter 2: A Theory of External Assistance as a Subsidy to Armed Groups in Civil Wars
This chapter develops a theory of civil war duration which examines the strategic interaction between foreign states and the internal warring parties. The central argument is that when external assistance is available to armed actors, it interacts with the internal strategic environment -- characterized by the cost of escalation of the fighting and the stakes of conflict -- to expand the conditions under which these actors choose to continue to fight. In effect, then, external assistance subsidizes ongoing war. Contrary to the notion that without significant interests hanging in the balance foreign states will abstain from interfering in civil wars, the theory predicts that a low threshold of external interests is sufficient to generate external assistance, even though these states may also be unable to realize the gains they seek through providing assistance. The theory also predicts, counter-intuitively, that civil wars are less likely to end the higher the cost of escalation and the lower the stakes. In addition to its principal focus on explaining war duration, the theory has implications for two sets of wartime behaviors of armed actors which are infrequently studied and not well understood in the literature: the extent to which armed actors pursue military operations without the goal of taking or holding territory (fighting without warfare) and the extent to which outright conflict occurs between armed actors that are meaningfully allied together along the war’s central cleavage (“on-side” armed groups).
Chapter 3: Mechanisms of Continuing Conflict: The Lebanese Civil War, 1975-1990 (available upon request)
I examine the plausibility of the mechanisms underpinning the theory of external assistance as a subsidy to armed groups as an explanation of civil war duration. This sets the stage for later chapters testing the theory’s predictions. Here, I show that the theory’s mechanisms can realistically account for the behavior of armed actors in the Lebanese Civil War, 1975-1990. In particular, the dynamics of external support and the cost of escalating fighting to engage in territorial operations explain armed actors’ seemingly opaque choices at three turning points in the conict that led to continuing war. The chapter draws on 120 hours of interviews conducted in Arabic with former combatants across multiple armed groups which fought in the war. While the chapter does not offer a comprehensive history of the war in Lebanon, its analysis of turning points can provide insight into the causal processes behind the war's duration.
Chapter 4: The Cost of Escalation, Stakes, Foreign Strategic Interests, and the Length of Civil Wars: A Cross-Country Test of the Theory of External Assistance as a Subsidy
Using a survival analysis of civil war duration data covering the period 1944-2006, this chapter tests the predictions of the theory of external assistance as a subsidy to armed groups in civil wars. Interests of external powers are proxied by a set of measures of the relationship between a civil war country and great powers: These include trade statistics and proximity of voting records in the UN General Assembly, and geostrategic features of the country that increase its intrinsic value to great powers, such as distance, pre-war hydrocarbon discovery and production, and an indicator of whether the country borders a world shipping choke point. The cost of escalating the level of fighting and the stakes are proxied by measures of population density, terrain, and ethnic polarization. Higher population density and difficult terrain facilitate the organization of violence against the government and therefore reduces the value of victory in the current war because of the prospect of a future one started by the losing side; they also make fighting to seize territory more costly. An additional variable proxying for the stakes of the conflict is an indicator of whether a civil war occurred in the country in the previous twenty years. Consistent with the theory, the empirical analysis strongly supports the predictions that civil wars become more likely to continue as the cost of escalation increases and as the stakes decrease. I also find a clear threshold effect of foreign strategic interests: war is more likely to continue after these interests pass a low threshold, and as they increase. A series of robustness checks confirms these results.
Chapter 5: Comparative Analysis: Civil Wars in Chad (1965-1994) and Yemen (1994)
In a second cut at testing the theory, this chapter examines two civil wars that differ systematically from the Lebanese Civil War. The analysis provides richer information than possible in the cross-national analysis about the origins of, and rationale for external assistance to armed groups, and the link between this assistance and the duration of war. It helps to verify the theory’s mechanisms and its ability to serve as a general explanation for civil war duration. The first case covered is the lengthy civil war in Chad from 1965 to 1990. Chad saw multiple foreign interventions but in a different geopolitical context and under different domestic circumstances than Lebanon. The second case is the short-lived civil war in Yemen in 1994. Although conducted with ferocity, the war concluded swiftly, and did not see external interference, despite being similar to Lebanon along several internal dimensions of interest. Additionally, although previous wars in Yemen also tended to be fairly short (in the Yemen Arab Republic, 1948; in the People’s Democratic Republic of Yemen, 1986-7), a war in the 1960s in the Yemen Arab Republic saw extensive intervention on the part of Egypt (with approximately 18,000 troops), the British, the Jordanians, and the Saudis, and lasted approximately eight years.
Chapter 6: Conclusion
Summary and discussion of the policy implications of the dissertation’s findings. For international actors involved in conflict resolution, the dissertation’s findings highlight the undesirable consequences of compensating potential spoilers in order to broker an agreement given a context of ongoing conflict. The indirect consequences of external support also suggest that foreign government officials deciding whether or how to support armed actors in a civil war should be less sanguine about the effects of assistance. Conventional wisdom -- which sees gains from pouring resources into a local ally’s coffers -- may obscure the most effective strategy for ending conflict quickly. In addition, the research implies that development policy should explicitly consider the ways in which aid can alter the strategic calculations of armed actors. Rather than only designing aid so that it does not fall into the hands of combatants, more work is likely needed to understand the pathways by which external assistance affects groups’ decision-making, even if through indirect or structural channels.
The extent of variation in the duration of civil wars and the lack of comprehensive explanations for it are explored. This chapter reviews the main, existing explanations and lays out their problems, emphasizing the lack of mechanisms and faulty logic where mechanisms are proposed. It also shows the disconnect between the theories and the reality of long wars through comparative anecdotal evidence about civil wars.
Chapter 2: A Theory of External Assistance as a Subsidy to Armed Groups in Civil Wars
This chapter develops a theory of civil war duration which examines the strategic interaction between foreign states and the internal warring parties. The central argument is that when external assistance is available to armed actors, it interacts with the internal strategic environment -- characterized by the cost of escalation of the fighting and the stakes of conflict -- to expand the conditions under which these actors choose to continue to fight. In effect, then, external assistance subsidizes ongoing war. Contrary to the notion that without significant interests hanging in the balance foreign states will abstain from interfering in civil wars, the theory predicts that a low threshold of external interests is sufficient to generate external assistance, even though these states may also be unable to realize the gains they seek through providing assistance. The theory also predicts, counter-intuitively, that civil wars are less likely to end the higher the cost of escalation and the lower the stakes. In addition to its principal focus on explaining war duration, the theory has implications for two sets of wartime behaviors of armed actors which are infrequently studied and not well understood in the literature: the extent to which armed actors pursue military operations without the goal of taking or holding territory (fighting without warfare) and the extent to which outright conflict occurs between armed actors that are meaningfully allied together along the war’s central cleavage (“on-side” armed groups).
Chapter 3: Mechanisms of Continuing Conflict: The Lebanese Civil War, 1975-1990 (available upon request)
I examine the plausibility of the mechanisms underpinning the theory of external assistance as a subsidy to armed groups as an explanation of civil war duration. This sets the stage for later chapters testing the theory’s predictions. Here, I show that the theory’s mechanisms can realistically account for the behavior of armed actors in the Lebanese Civil War, 1975-1990. In particular, the dynamics of external support and the cost of escalating fighting to engage in territorial operations explain armed actors’ seemingly opaque choices at three turning points in the conict that led to continuing war. The chapter draws on 120 hours of interviews conducted in Arabic with former combatants across multiple armed groups which fought in the war. While the chapter does not offer a comprehensive history of the war in Lebanon, its analysis of turning points can provide insight into the causal processes behind the war's duration.
Chapter 4: The Cost of Escalation, Stakes, Foreign Strategic Interests, and the Length of Civil Wars: A Cross-Country Test of the Theory of External Assistance as a Subsidy
Using a survival analysis of civil war duration data covering the period 1944-2006, this chapter tests the predictions of the theory of external assistance as a subsidy to armed groups in civil wars. Interests of external powers are proxied by a set of measures of the relationship between a civil war country and great powers: These include trade statistics and proximity of voting records in the UN General Assembly, and geostrategic features of the country that increase its intrinsic value to great powers, such as distance, pre-war hydrocarbon discovery and production, and an indicator of whether the country borders a world shipping choke point. The cost of escalating the level of fighting and the stakes are proxied by measures of population density, terrain, and ethnic polarization. Higher population density and difficult terrain facilitate the organization of violence against the government and therefore reduces the value of victory in the current war because of the prospect of a future one started by the losing side; they also make fighting to seize territory more costly. An additional variable proxying for the stakes of the conflict is an indicator of whether a civil war occurred in the country in the previous twenty years. Consistent with the theory, the empirical analysis strongly supports the predictions that civil wars become more likely to continue as the cost of escalation increases and as the stakes decrease. I also find a clear threshold effect of foreign strategic interests: war is more likely to continue after these interests pass a low threshold, and as they increase. A series of robustness checks confirms these results.
Chapter 5: Comparative Analysis: Civil Wars in Chad (1965-1994) and Yemen (1994)
In a second cut at testing the theory, this chapter examines two civil wars that differ systematically from the Lebanese Civil War. The analysis provides richer information than possible in the cross-national analysis about the origins of, and rationale for external assistance to armed groups, and the link between this assistance and the duration of war. It helps to verify the theory’s mechanisms and its ability to serve as a general explanation for civil war duration. The first case covered is the lengthy civil war in Chad from 1965 to 1990. Chad saw multiple foreign interventions but in a different geopolitical context and under different domestic circumstances than Lebanon. The second case is the short-lived civil war in Yemen in 1994. Although conducted with ferocity, the war concluded swiftly, and did not see external interference, despite being similar to Lebanon along several internal dimensions of interest. Additionally, although previous wars in Yemen also tended to be fairly short (in the Yemen Arab Republic, 1948; in the People’s Democratic Republic of Yemen, 1986-7), a war in the 1960s in the Yemen Arab Republic saw extensive intervention on the part of Egypt (with approximately 18,000 troops), the British, the Jordanians, and the Saudis, and lasted approximately eight years.
Chapter 6: Conclusion
Summary and discussion of the policy implications of the dissertation’s findings. For international actors involved in conflict resolution, the dissertation’s findings highlight the undesirable consequences of compensating potential spoilers in order to broker an agreement given a context of ongoing conflict. The indirect consequences of external support also suggest that foreign government officials deciding whether or how to support armed actors in a civil war should be less sanguine about the effects of assistance. Conventional wisdom -- which sees gains from pouring resources into a local ally’s coffers -- may obscure the most effective strategy for ending conflict quickly. In addition, the research implies that development policy should explicitly consider the ways in which aid can alter the strategic calculations of armed actors. Rather than only designing aid so that it does not fall into the hands of combatants, more work is likely needed to understand the pathways by which external assistance affects groups’ decision-making, even if through indirect or structural channels.