Event or Hazards Analysis in Civil War Duration - Coursework Example

Event/Hazards Analysis in Civil War Duration Inserts His/Her Inserts Inserts Introduction Survival analysis is a statistical analysis method that concerns itself with the analysis of the duration of time that it takes for one or more events to occur. It is also referred to as duration modelling or analysis (economics), event history analysis (sociology), or reliability analysis (engineering). Civil wars have a very big difference from interstate wars. Many studies have been carried out to analysethe outbreak and duration of civil wars based on economic, political and ethnic grievances. This paper looks at five articles that utilize event history/survival analysis 1. Data on Interventions during Periods of Political Instability (Reagan and Meachum, 2014) This article studies pre-conflict intervention strategies and how they help to resolve instances of civil wars.According to Regan and Meachum (2014), many studies that have been carried out with respect to civil wars are only limited by an emphasis of intervention once conflict starts. The authors then seek to develop a dataset that provides an appropriate basis to study pre-conflict interventions. The hypothesis of the study is that interventions before the onset of civil war might influence the ability or willingness of opposition groups to challenge their government by the use of force. The study first looks at previous literature on civil wars with an emphasis on how they start, how interventions help resolve the conflicts and finally how they end. Having done this, they look at the inherent limitation of the majority of the previous literature which they note is that this study fail to study thecore processes involved in pre-conflict intervention due to lack of an appropriate dataset on the subject. Having presented the problem under study, the article then strives to create a new dataset on external interventions to nations that are likely to be at risk of civil wars (Regan and Meachum 128). In order to develop this dataset, the article first describes the data, the sampling and coding methods, and some of the core patterns witnessed in the data. After this, the article shows how the data can be utilized to illuminate the civil war processes and the role of third parties by using the data in an empirical model. In developing the data for pre-war interventions, the authors noted that the greatest difficulty is narrowing down the potential universe of case as all countries not at civil war have a risk of actually engaging in conflict. In order to generate a sample of those nations which have a considerable chance of experiencing a civil war and for which there is a reasonable possibility that an external actor would intervene for either party in the conflict, the authors used risk estimates developed by Goldstone (Regan and Meachum 128).Goldstone developed a model that generates a risk score that reveals the possibility of a country to experience civil war two years in the future. The risk score ranges from 0 to 0.99 with a standard deviation of 0.23 and a mean score of 0.24. In collecting the data, this article established a threshold of 0.3 to indicate those nations that are most susceptible to civil wars. A team of data coders carried out archival searches on 24 different sources to determine the occurrences of civil wars. Data for all countries with a risk factor greater than 0.3 was entered into a web interface and then analysed for consistency. The data was recorded on the date of intervention, the target, the type and amount of the intervention and the identity of the intervener. In the 1443 years-at-risk that the authors identified, they were able to code 449 separate interventions: military interventions (230), diplomatic interventions (110) and economic interventions (109). They also found that 75 out of the 112 at risk countries were targeted for at least one intervention (Regan and Meachum, 129). Apart from the number of interventions, the study also elucidates patterns with regard to the identity of the intervener. To test the hypothesis about the effects of external interventions on civil war onset, the studio collapsed the data to the level of the country - month. The data utilized a series of control variables that have been used in predicting the onset of civil wars and developed a duration model. The study discovered that in countries at risk of civil wars, diplomatic interventions can delay the onset of the war while military interventions increase the likelihood of the war. This was carried out using the dataset that was developed creating a way forward in studying pre-conflict intervention strategies and the lifetime of civil wars Causal Diagram 2. When and how many: The effects of third party joining on casualties and duration in interstate wars (Shirkey 2012) This study seeks to find out the factors that make some wars longer and more severe than others. The study looks at the impact of third party joiners on the duration and casualties civil wars and supports their findings through a Cox analysis, ordinary least squares regression and non-proportional hazards model regression(Shirkey, 322).Modern literature has mostly concentrated on which nation join wars and why they join these wars but however do not explain how joining affects these wars. This study endeavours to fill this void in existing literature. The hypothesis of this study is that late joining by third parties in a waris correlated with increased durationof the war independent of having more participants. They argue that this relationship is caused by joining increased duration rather than longer wars proving the chance for third parties to join. In order to prove their hypothesis, the authors utilize an informational approach(Shirkey 322). In developing their methodology, the authors developed several hypotheses to add to the main hypothesis. First, late joining should increase the duration of the war. Second, two-sided late joining should increase the duration more than one sided late joining. Finally, early joining should not be linked with longer duration, as those countries that join early are likely to have been pre-war allies with the initial belligerents. The theories on war duration were combined with these hypotheses and a set of control variables were added in order to carry out their analysis. The wars that were used in this paper were retrieved from the COW dataset on interstate wars. The dataset includes 79 warsfought between 1816 and 1997(Shirkey 324). Two datasets were then developed, one where each case is a single case and another where each case stands for a war day in a given war. Most of the control variables used in this study were taken from the COW database as well.Dummy variables were used to record whether a greater power was involved in the war or on both sides of the war. To determine is joining affects the duration of a war, Cox non-proportional hazards model was developed. The authors chose the non-proportional hazards model as they note that the Cox proportional model assumes that variables affect the hazard rate in a proportional manner over time, but the Schoenfeld residual procedure reveals that that late joining plus several strategies variables potentially violate the proportional hazards assumption. The authors carried out an analysis of four models that include: wars by war-day: non-dummy variables, war by war-day: dummy variables, wars as a single case: non-dummy variables, and wars as a single case: dummy variables. From the cox non-proportional hazards model for these four models revealed that late joining is significantly correlated with a large increase in duration. The results do not show an obvious relationship since the impacts of late joining on shape parameters and scale run in different directions. By holding all other variables at their medians and calculating the predicted values for when late joining occurs a relationship was derived. In the model, this revealed that late joining increased the duration by 603, 543, 44, and 229 days respectively. Using percentage, the results were 164%, 155%, 8% and 68% respectively(Shirkey, 328). In all the models, it was revealed that early joining is significantly correlated with shorter wars. The study thus concluded that it is not the length of wars that allow late joining to occur, as the majority of those durations arise after joining has occurred. Causal Diagram 3. Information, Commitment, and Intra-War Bargaining: The Effect of Governmental Constraints on Civil War Duration (Thyne, 2012) This article looks at the manner with which variations by governments affect the duration of civil wars. The paper notes that past literature suggests that the termination of wars is possible when competing actors have information on the balance of power and are therefore able to credibly abide to contracts that would end the war (Thyne 307). The main purpose of this study is to clear up the inconsistency about the length of civil wars present in studies that focus mainly on the third party actor, population size or ethnic fractionalisation, by focusing on variations within governments. The paper first looks at barriers to civil war settlements where they find that the main barriers are information problems and commitment problems. Alternative explanations to explain how variability in the government may affect the duration of civil wars include democratic reliability, war-fighting strategies, and government demands dependency on stability and power (Thyne 308).In the research design, the paper observes the duration of civil conflicts from 1946 to 2004 using the Uppsala dataset. The duration of each civil war is analysed by noting whether or not the conflict ended in each month using a hazards model. The basis of the analysis used in this study is that civil wars end after they have occurred for a long time due to war weariness. The first set of independent variables utilised in this study is information variables. The first hypothesis holds that civil conflicts will last longer when institutions limit the ability of the executive to act unilaterally. This hypothesis is tested using executive constraints (XCONST) a component of the polity index (Thyne 313).The second hypothesis holds that civil conflicts will last longer in a situation where the executive shares power with other political actors. This hypothesis was tested using a variable known as political constraints, which is the index of executive electoral competitiveness and is available in the database for political institutions. The third hypothesis predicts that heightened levels of political divergence within the ruling coalition will increase the length of civil conflicts.The maximum level of polarization was drawn from political polarization available from the DPI dataset. The second set of variables is the commitment variety that endeavour to capture the ability of the government to credibly commit to war-ending contracts. The fourth hypothesis predicts that civil conflicts within parliamentary governments should last longer than those in presidential governments. This hypothesis was captured with a dummy variable coded 0 for all presidential governments and 1 for parliamentary governments (Thyne 313). The analysis begins by developing a framework on the general expectations for how variations in government constraints affects the duration of civil wars. After this, some control variables are chosen in order to isolate the impacts of the main independent variables. The primary control variable used in the study is opposition vetoes that explain fragmentation in the opposition side of the conflict. The dummy variable used is coups, derived from Powell and Thyne study carried out in 2011.The data analysis starts by examining indices on the impact of information variables on civil war duration. The positive and significant coefficients used in the first two hazards model provided support for the general expectation of the study for both information (p Read More