Intervention Strategies and the Outcome of Intrastate Conflicts

本项目将以早期探索外部干预与敌对行动分离之间关系的工作为基础。具体来说，我的研究计划将解决以下问题：1)有和没有外部干预的冲突的预期持续时间；2)检查策略与结果之间的关系，特别是“胡萝卜加大棒”干预方法的有效性，通过检查惩罚和奖励的相互作用，以及渐进式干预与大规模、偶发性干预的相对有效性（前者集中在尼加拉瓜或萨尔瓦多，后者集中在索马里）；(3)国家在试图影响国内冲突进程时取代军事、经济和外交举措的条件。除了这些具体目标之外，该项目还将有助于我们了解各国何时以及如何在一系列潜在的政策选择中做出选择，选择机制在评估结果中的作用，以及更广泛的政策评估。该项目将增加我们对干预有效性的现有知识，并向社会科学界提供关于内部冲突干预的相当精确的数据集。具体来说，我将讨论以下问题：发展和测试对干预策略对冲突结果的作用的更复杂的理解检查干预替代的影响将外交干预作为军事或经济策略的补充或替代检查外部干预对国内冲突持续时间的影响数据改进：根据干预的规模和速度记录干预数据按顺序排列干预策略更新1994年至今的当前干预数据最初的重点是考虑到（或尽管）有外部干预，预计冲突的持续时间。一个有趣的理论和政策问题是，干预会缩短还是延长国内冲突的预期持续时间。我们期望，当第三方进行干预时，他们会预期他们的努力将导致缩短冲突。这一点从未得到充分的检验。分析的第二阶段将检查不同的干预策略对政策成功或失败的影响。如果国家（或国际组织）使用干预策略作为冲突管理的一种形式，那么了解什么最有效是至关重要的信息。目前这方面的知识还相当贫乏。例如，我们对在试图让国内冲突参与者停止战斗时使用外交和胁迫相结合的冲突管理影响知之甚少。我将首先测试第三方选择干预策略的条件。使用Heckman选择模型，我可以测试在潜在条件中是否存在系统性变化，这些条件“导致”参与者选择属于特定类别。这需要一个两阶段的估计过程，其中将指定一个解释干预成功或失败的回归方程，以及一组应该解释干预选择的变量。第一阶段将在给定的一组特定条件下，估计我们观察到干预的可能性；第二阶段是评估干预成功的可能性。这项研究工作将产生两件对社会科学界和政策界具有持续价值的东西。首先，一系列的期刊文章将描述数据并呈现分析结果。这些分析应具有明确的政策相关含义，提出最有可能成功制止国内冲突的战略。我预计这个项目最终会写成一本书长的手稿，它将有一个由政策制定者和学者组成的双重读者。第二，作为劳动密集型数据生成工作的结果，我们将有一个关于国内冲突干预措施的数据集，这将是非常需要和广泛适用的。我将通过我的网站公开这些数据，并/或将其存入ICPSR。

This project will build on earlier work that explores the relationship between outside interventions and the secession of hostilities. Specifically my research proposal will address the following:1) the expected duration of a conflict with and without outside interventions;2) an examination of the relationship between strategy and outcome, specifically the effectiveness of the "carrot and stick" approach to intervening, by examining the interplay of punishments and rewards, and the relative effectiveness of gradual interventions versus large, episodic ones (the former is epitomized by Nicaragua or El Salvador; the latter by Somalia);3) the conditions under which states substitute military, economic, and diplomatic initiatives in their attempts to influence the course of internal conflicts.Beyond these specific objectives the project will contribute to our understanding of when and how states choose across a range of potential policy options, the role of selection mechanisms in evaluating outcomes, and policy evaluation more generally.This project will add to the current state of our knowledge of intervention effectiveness, as well as making available to the social science community a fairly refined data set on interventions in internal conflicts. Specifically I will address the following: Conceptual Refinements: Develop and test a more complex understanding of the role of intervention strategy on conflict outcomes Examine the effects of intervention substitution Incorporate diplomatic interventions as a compliment to or alternative for military or economic strategies Examine the effect of outside interventions on the duration of intrastate conflictsData Refinements:Record intervention data in terms of size and speed of the intervention Sequentially order the unfolding strategy of intervention Update current intervention data from 1994 to the presentThe initial focus would be on the expected duration of the conflict in light of (or in spite of) an outside intervention. The interesting theoretical and policy question is whether interventions shorten or prolong the expected duration of a civil conflict. We would expect that when third parties intervene they anticipate that their effort will lead to shortening the conflict. This has never been sufficiently examined. The second stage of the analysis will examine the effect of different strategies for intervening on the success or failure of the policy. If states (or international organizations) use intervention strategies as a form of conflict management then knowing what works best is critical information. Currently this level of knowledge is rather meager. For example, we know very little about the conflict management implications of using a mixture of diplomacy and coercion in trying to get actors in civil conflicts to stop fighting. I will first test for the conditions under which third parties select into intervention strategies. Using a Heckman selection model I can test whether there is systematic variation in potential conditions that 'cause' an actor to choose to be in a specified category. This requires a two-stage estimation process, where a regression equation that would account for the success or failure of an intervention would be specified, along with a set of variable that should account for selection into the intervention. The first stage would estimate the likelihood that we would observe an intervention, given the set of specified conditions; the second stage would estimate the likelihood that an intervention would be successful. This research effort will produce two things of continuing value to the social science and policy communities. First, a series of journal articles will describe the data and present the results of the analyses. These analyses should have clear policy relevant implications, suggesting strategies for stopping intrastate conflicts that have the greatest chance of success. I anticipate that the project could culminate in a book-length manuscript that would have a dual audience consisting of policymakers and scholars. Second, as a result of the labor-intensive data generation effort we will have a data set on interventions in civil conflicts that will be highly desired and broadly applicable. I will make these data publicly available via my web site and/or depositing them with ICPSR.