Bethany Lachina Civil War
In this research report, we shall investigate, evaluate and present the work of Bethany Lachina into the explanations for the severity of civil conflicts. Lachina’s inspiration for such an inquiry is founded upon the fact that, considering the publishing of numerous studies into civil conflicts — few, if any, focused exclusively and adequately on why some civil wars are so much deadlier than others. Given that domestic discord is the most common form of warfare, accounting for 90 percent of battle deaths (including combatant and collateral) during the period of 1990 to 2002 (Lachina and Gleditch, 2005), it is critical to understand which chief factor, or what combination of factors, influence conflict severity. Fearon interprets civil war as
…show more content…
"a violent conflict” enclosed by the confines of a territory, contended by “organized groups” that seek to “take power” at the “center" or in a province, or to “change government policies”(Fearon, 2007).
Scholarly debate has routinely focused upon the debate between the opportunity and motivation to engage in armed conflict. We shall focus upon suggested areas which may determine ability and willingness to engage in civil conflict; the competency and economic proficiency of the state; the type of government and administration, and, variations in conflict duration.
The work of Fearon and Collier (Fearon and Collier, 2004) examines the relationship between the duration and severity of civil war. Such analysis necessitates the establishment of clear dates marking the precipitation and resolution of conflict. Hoefler uses the example of the assassination of Rwandan President Juvénal Habyarimana on 6th March 1994 as marking the start of the civil war in a number of studies. Hoeffler argues that such “trigger” events cannot be dated as, in most cases, hostility “escalates” over some period of time before it qualifies to be defined as a “civil war”. Hoefler relies upon the Correlates of War (COW) project to define civil war, that is, a) it requires that there is “organized military action” and that at least “1,000 battle deaths resulted” b) as a necessary condition to differentiate such wars from “genocides, massacres and pogroms” there must be “effective resistance”, at least “five percent of the deaths” should be inflicted by the “weaker party.” Keen emphasises that the “national government” at the time should be “actively involved”. Hoefler stresses that the final requirement is somewhat problematic with regard to wars of liberation from colonialism. Keen highlights Angola (1961-75), Mozambique (1964-75) and Western Sahara (1975-83) as ‘extra-systemic wars’ and not civil conflicts for this reason (Keen, 2000, pp. 20-21).
Finally, Hoefler points to the fact that most wars end with either victory of the recognised armed defence forces, settlement between parties of some form of truce. The work of Sambanis (Sambanis, 2000) suggests that 50 percent of all civil wars end in military defeat, which, Hoefler argues presents a more coherent method and framework for dating conflict resolution than using dates of peace agreements, which, Hoefler underlines, may have not resulted in the cessation of hostilities.
Murshed argues that “greed and grievance” are the main explanatory factors for civil war.
Citing the work of Murdoch and Sandler, (Murdoch and Sandler, 2004), Murshed explains that a civil war can reduce a country’s economic growth potential by 31 percent in the long-term and by 85 percent in the short to medium-term. Therefore, the level of income (per capita GDP) is a “robust” predictor of severity. Collier and Hoefler highlight factors which further precipitate such severity, such as sizeable numbers of “idle, young men”, modulated capital and trade markets, “profuse” and “copious” amounts of seizable natural resources and “penetrable” borders and boundaries. Urdal explains that young males, whom are indicative of the vast majority of combatants in civil wars) are, to a lesser degree, probable to participate in insurgency when they are “getting an education” or “have a stable, secure salary”, and can “reasonably assume” that they will “prosper in the future” (Urdal, …show more content…
2012).
Collier & Sambanis state that a “profuse measure” of “primary commodities” in national exports, such as crude oil or precious minerals, markedly augments the severity of conflict. According to the study, a country at "peak danger", with such commodities representing 32 percent of GDP has a 22% “risk” of plunging into civil war in a timeframe of 60 months, while a country with no (i.e. 0%) primary commodity exports has a 1% “risk” (Collier & Sambanis, 2005, p. 16). This analysis relates directly to the ability of combatants to acquire additional wealth promptly and corroborates the fact that appropriating capital from, for example, a gold mine or oil field, is of a more amenable pursuit than attempting to liquidate capital from, for example, garment manufacturing industry or hospitality service.
Non-domestic military assistance to the government and/or its challengers has been a recurrent issue in the duration and severity of intrastate conflicts. Regan illustrates that, of the 138 intrastate conflicts between the end of 1945 and 2000, 75 percent experienced “some form” of non-domestic abettance; the United States intervening directly or in proxy in 35 of these conflicts (Regan, 2000). Hironaka, examining the relationship between conflict duration, severity and non-domestic collaboration found that a civil war with intervention on only one side is 156 percent longer, while when intervention occurs on both sides the average civil war is longer by an additional percent (Hironaka, 2005, pp. 50-51).
Hypothesis 1: Between states experiencing civil conflicts, more severe conflict (y) will occur as the duration of conflict increases (x). The justification for this is that the higher the opportunity (represented by time) there is to engage in armed conflict, the higher the potentiality for successful operational gains, a consequence of which is higher battle deaths.
Hypothesis 2: Among states experiencing civil conflicts, more severe conflict (y) will occur in the states with the highest populations (x). The justification for this is that a larger military can be drawn from a larger population, and therefore the functioning ability of said military to engage in strategic operations to undermine a weaker party is multiplied, and, consequently, the potentiality of collateral, civilian casualties is also increased.
Hypothesis 3: Among states experiencing civil conflicts, more severe conflict (y) will occur in debilitated and unsound states (x). The justification for this is that a low gross domestic product (GDP) per capita, and, a consequence of this, high unemployment (especially among young men), is a compelling forecast of civil war because it is representative of state incapacity.
Hypothesis 4: Among states experiencing civil conflicts, more severe conflict (y) will occur when there is non-domestic military assistance to the government and/or its opposition. The justification for this is that when non-domestic, external actors support factions which would not otherwise be capable of initiating, or indeed sustaining an armed conflict, the severity of internal wars are likely to be more severe, as capabilities (which are necessary for the effective execution of military campaigns and operations, and, which, consequently lead to battle deaths) are invigorated and bolstered by foreign actors.
Hypothesis 5: Among states experiencing civil conflicts, less severe conflict (y) will occur in states with irregular geographical topography (y). The justification for this is that in rough terrain, combat forces will be deployed in smaller units, reducing the possibility of hostile engagement and also restricts the utilisation of heavy weapons and artillery, the employment of which increases the likelihood of battle deaths and collateral civilian casualties.
Hypothesis 6: Among states experiencing civil conflicts, less severe conflict (y) will occur in states where the political system is representative of a democracy (x). The justification for this is that democratic executives, facing domestic electoral pressures and diplomatic constrains abroad, may be more likely to concede to compromise when evaluating the socioeconomic consequences of a severe revolutionary or terrorist threat.
The null hypothesis of all of the above is that there is no relationship in the population, i.e. the impact of x on y is zero.
As part of this research report, the dataset of which I used as the basis of my work is Lacina_JCR_2006_replication.sav and the method which I utilised to assess my hypotheses was regression analysis.
Ordinary Least Squares (OLS) regression analysis determines the line that describes the relationship between x and y where the sum of the squared errors is the ‘least’. In all cases, the dependant variable needs to be of interval/ratio level, the independent variable can be a dummy or an interval/ratio level variable. The intercept, also known as the constant (a or α) represents the predicted value of the dependent variable (y) when the independent variable (x) is zero. The slope (b or β) represents the average change in dependent variable (y) for a one unit change in independent variable (x). For example, for hypothesis 1, the operational indicators were x = Natural log duration of conflict in years and y = Natural log best estimate battle
deaths.
Significance testing in OLS regression analysis is contingent upon whether an estimate of the intercept and the slope is emblematic of the true population framework. There are three ways to assess whether the independent variable has a statistically significant impact on the dependent variable: the p-value of the slope coefficient; the t-ratio of the slope coefficient and the confidence interval of the slope coefficient. When the p-value is <.05.
Model fit is appraised based upon the R Square and Standard Error of the Estimate (SEE). R-square denotes the distribution of variation in the dependent variable explained by the independent variable(s), and ranges between 0 and 1. When R-square is 1, the independent variable(s) account for all the variation in the dependent variable. When it is 0 no variation in the dependent variable is accounted for. For example, when expressed as a percentage, when R-square = .56, then 56% of the variation in the dependent variable is explained by the independent variable(s). The Standard Error of the Estimate (SEE) is the approximate standard digression of the observed value of the dependent variable (y) and the predicted value of the independent variable.
Table B shows the results of OLS regressions for battle deaths in civil conflicts. The results show there is no statistically significant relationship between death per year and conflict duration, and consequently, a larger populous does not forecast a higher number of deaths. The results also show that military quality, GDP per capita and geographical terrain are all statistically insignificant. Consequently, the results do not support hypotheses 1, 2, 3, 4 or 5 — however, hypothesis 6 is supported by the fact that countries which are of a more democratic nature are correlated with less severe, measured by battle-deaths, civil conflict. The implications of this study are profound and wide-ranging. They emphasise the fact that democratic regimes are more likely to result in less severe conflict. I do, nonetheless, agree with Lachina’s analysis that the classical argument — more specifically, the right set of circumstances, is what conditions “where and when” wars begin (Lacina, 2006, p. 287), is uncorroborated and that “knowing when wars start” may not disclose “when they will be most devastating”. This report concludes that representative and egalitarian principles and “norms”, as well as “institutional adaptability” of such regimes give an explanation of this finding.
"a violent conflict” enclosed by the confines of a territory, contended by “organized groups” that seek to “take power” at the “center" or in a province, or to “change government policies”(Fearon, 2007).
Scholarly debate has routinely focused upon the debate between the opportunity and motivation to engage in armed conflict. We shall focus upon suggested areas which may determine ability and willingness to engage in civil conflict; the competency and economic proficiency of the state; the type of government and administration, and, variations in conflict duration.
The work of Fearon and Collier (Fearon and Collier, 2004) examines the relationship between the duration and severity of civil war. Such analysis necessitates the establishment of clear dates marking the precipitation and resolution of conflict. Hoefler uses the example of the assassination of Rwandan President Juvénal Habyarimana on 6th March 1994 as marking the start of the civil war in a number of studies. Hoeffler argues that such “trigger” events cannot be dated as, in most cases, hostility “escalates” over some period of time before it qualifies to be defined as a “civil war”. Hoefler relies upon the Correlates of War (COW) project to define civil war, that is, a) it requires that there is “organized military action” and that at least “1,000 battle deaths resulted” b) as a necessary condition to differentiate such wars from “genocides, massacres and pogroms” there must be “effective resistance”, at least “five percent of the deaths” should be inflicted by the “weaker party.” Keen emphasises that the “national government” at the time should be “actively involved”. Hoefler stresses that the final requirement is somewhat problematic with regard to wars of liberation from colonialism. Keen highlights Angola (1961-75), Mozambique (1964-75) and Western Sahara (1975-83) as ‘extra-systemic wars’ and not civil conflicts for this reason (Keen, 2000, pp. 20-21).
Finally, Hoefler points to the fact that most wars end with either victory of the recognised armed defence forces, settlement between parties of some form of truce. The work of Sambanis (Sambanis, 2000) suggests that 50 percent of all civil wars end in military defeat, which, Hoefler argues presents a more coherent method and framework for dating conflict resolution than using dates of peace agreements, which, Hoefler underlines, may have not resulted in the cessation of hostilities.
Murshed argues that “greed and grievance” are the main explanatory factors for civil war.
Citing the work of Murdoch and Sandler, (Murdoch and Sandler, 2004), Murshed explains that a civil war can reduce a country’s economic growth potential by 31 percent in the long-term and by 85 percent in the short to medium-term. Therefore, the level of income (per capita GDP) is a “robust” predictor of severity. Collier and Hoefler highlight factors which further precipitate such severity, such as sizeable numbers of “idle, young men”, modulated capital and trade markets, “profuse” and “copious” amounts of seizable natural resources and “penetrable” borders and boundaries. Urdal explains that young males, whom are indicative of the vast majority of combatants in civil wars) are, to a lesser degree, probable to participate in insurgency when they are “getting an education” or “have a stable, secure salary”, and can “reasonably assume” that they will “prosper in the future” (Urdal, …show more content…
2012).
Collier & Sambanis state that a “profuse measure” of “primary commodities” in national exports, such as crude oil or precious minerals, markedly augments the severity of conflict. According to the study, a country at "peak danger", with such commodities representing 32 percent of GDP has a 22% “risk” of plunging into civil war in a timeframe of 60 months, while a country with no (i.e. 0%) primary commodity exports has a 1% “risk” (Collier & Sambanis, 2005, p. 16). This analysis relates directly to the ability of combatants to acquire additional wealth promptly and corroborates the fact that appropriating capital from, for example, a gold mine or oil field, is of a more amenable pursuit than attempting to liquidate capital from, for example, garment manufacturing industry or hospitality service.
Non-domestic military assistance to the government and/or its challengers has been a recurrent issue in the duration and severity of intrastate conflicts. Regan illustrates that, of the 138 intrastate conflicts between the end of 1945 and 2000, 75 percent experienced “some form” of non-domestic abettance; the United States intervening directly or in proxy in 35 of these conflicts (Regan, 2000). Hironaka, examining the relationship between conflict duration, severity and non-domestic collaboration found that a civil war with intervention on only one side is 156 percent longer, while when intervention occurs on both sides the average civil war is longer by an additional percent (Hironaka, 2005, pp. 50-51).
Hypothesis 1: Between states experiencing civil conflicts, more severe conflict (y) will occur as the duration of conflict increases (x). The justification for this is that the higher the opportunity (represented by time) there is to engage in armed conflict, the higher the potentiality for successful operational gains, a consequence of which is higher battle deaths.
Hypothesis 2: Among states experiencing civil conflicts, more severe conflict (y) will occur in the states with the highest populations (x). The justification for this is that a larger military can be drawn from a larger population, and therefore the functioning ability of said military to engage in strategic operations to undermine a weaker party is multiplied, and, consequently, the potentiality of collateral, civilian casualties is also increased.
Hypothesis 3: Among states experiencing civil conflicts, more severe conflict (y) will occur in debilitated and unsound states (x). The justification for this is that a low gross domestic product (GDP) per capita, and, a consequence of this, high unemployment (especially among young men), is a compelling forecast of civil war because it is representative of state incapacity.
Hypothesis 4: Among states experiencing civil conflicts, more severe conflict (y) will occur when there is non-domestic military assistance to the government and/or its opposition. The justification for this is that when non-domestic, external actors support factions which would not otherwise be capable of initiating, or indeed sustaining an armed conflict, the severity of internal wars are likely to be more severe, as capabilities (which are necessary for the effective execution of military campaigns and operations, and, which, consequently lead to battle deaths) are invigorated and bolstered by foreign actors.
Hypothesis 5: Among states experiencing civil conflicts, less severe conflict (y) will occur in states with irregular geographical topography (y). The justification for this is that in rough terrain, combat forces will be deployed in smaller units, reducing the possibility of hostile engagement and also restricts the utilisation of heavy weapons and artillery, the employment of which increases the likelihood of battle deaths and collateral civilian casualties.
Hypothesis 6: Among states experiencing civil conflicts, less severe conflict (y) will occur in states where the political system is representative of a democracy (x). The justification for this is that democratic executives, facing domestic electoral pressures and diplomatic constrains abroad, may be more likely to concede to compromise when evaluating the socioeconomic consequences of a severe revolutionary or terrorist threat.
The null hypothesis of all of the above is that there is no relationship in the population, i.e. the impact of x on y is zero.
As part of this research report, the dataset of which I used as the basis of my work is Lacina_JCR_2006_replication.sav and the method which I utilised to assess my hypotheses was regression analysis.
Ordinary Least Squares (OLS) regression analysis determines the line that describes the relationship between x and y where the sum of the squared errors is the ‘least’. In all cases, the dependant variable needs to be of interval/ratio level, the independent variable can be a dummy or an interval/ratio level variable. The intercept, also known as the constant (a or α) represents the predicted value of the dependent variable (y) when the independent variable (x) is zero. The slope (b or β) represents the average change in dependent variable (y) for a one unit change in independent variable (x). For example, for hypothesis 1, the operational indicators were x = Natural log duration of conflict in years and y = Natural log best estimate battle
deaths.
Significance testing in OLS regression analysis is contingent upon whether an estimate of the intercept and the slope is emblematic of the true population framework. There are three ways to assess whether the independent variable has a statistically significant impact on the dependent variable: the p-value of the slope coefficient; the t-ratio of the slope coefficient and the confidence interval of the slope coefficient. When the p-value is <.05.
Model fit is appraised based upon the R Square and Standard Error of the Estimate (SEE). R-square denotes the distribution of variation in the dependent variable explained by the independent variable(s), and ranges between 0 and 1. When R-square is 1, the independent variable(s) account for all the variation in the dependent variable. When it is 0 no variation in the dependent variable is accounted for. For example, when expressed as a percentage, when R-square = .56, then 56% of the variation in the dependent variable is explained by the independent variable(s). The Standard Error of the Estimate (SEE) is the approximate standard digression of the observed value of the dependent variable (y) and the predicted value of the independent variable.
Table B shows the results of OLS regressions for battle deaths in civil conflicts. The results show there is no statistically significant relationship between death per year and conflict duration, and consequently, a larger populous does not forecast a higher number of deaths. The results also show that military quality, GDP per capita and geographical terrain are all statistically insignificant. Consequently, the results do not support hypotheses 1, 2, 3, 4 or 5 — however, hypothesis 6 is supported by the fact that countries which are of a more democratic nature are correlated with less severe, measured by battle-deaths, civil conflict. The implications of this study are profound and wide-ranging. They emphasise the fact that democratic regimes are more likely to result in less severe conflict. I do, nonetheless, agree with Lachina’s analysis that the classical argument — more specifically, the right set of circumstances, is what conditions “where and when” wars begin (Lacina, 2006, p. 287), is uncorroborated and that “knowing when wars start” may not disclose “when they will be most devastating”. This report concludes that representative and egalitarian principles and “norms”, as well as “institutional adaptability” of such regimes give an explanation of this finding.