International Trade of Arms: A Network Approach

该项目提出了从网络角度对国际武器贸易进行实证分析的建议。确定各自的流动结构，它们的动态，以及它们的原因，将是对和平与冲突的实证科学的一个新的贡献。我们以数量为导向的网络方法是基于对长期（1960-2010年）武器出口、转让和进口的计量。我们将分析主要的常规武器（MCW）和小武器和轻武器（SALW）。我们第一次调查国家之间的多个部门内的关系，以找出这些发展一致或不一致，以及它们如何相互作用。更具体地说，我们将测试流行的理论，根据这些理论，小武器和轻武器的贸易结构在更大程度上是出于政治动机，而小武器和轻武器的贸易结构则预计是以经济为导向的。依靠多种数据来源，我们采用综合衡量方法，使我们能够弥补数据差距，克服数据不足，实施复杂的正式武器贸易模型，并预测未来的转让。我们要检验的主要假设来自新贸易理论及其对不完善国际市场的关注。除其他外，我们将具体说明政府间组织、联盟和军备控制制度的成员资格，以评估其有效性及其对贸易偏转的影响。基础数据源的复杂性要求开发新的推理统计技术，用于对有价值的数据进行建模（即：非二进制）网络-这将构成我们的合作跨学科项目的有条不紊的核心目标。这将使我们第一次考虑到内生网络结构和动态（例如，发送者效应表明累积优势，超二元系统特征，如三元闭包），从而超越了迄今为止研究的假设，完全基于演员的属性。

This project proposes a network perspective for the empirical analysis of the international trade of arms. The identification of respective flow structures, of their dynamics, and of their causes will be a novel contribution to the positive science of peace and conflict. Our quantitatively-oriented network approach is based on measurements of the export, transfer, and import of arms over long time periods (1960-2010). We will analyze major conventional weapons (MCW) and small arms and light weapons (SALW). For the first time we investigate multiple intra-sectoral relations between countries in order to find out whether these evolve congruently or disjointly, and how they interact. More specifically, we will test popular conjectures according to which the structure of trade in MCW follows to a higher degree political motivations whereas SALW is expected to be rather economically-oriented. Relying on multiple data sources we follow an integrative measurement approach allowing us to close data gaps, to overcome data deficiencies, to implement complex formal models of arms trade, and to forecast future transfers. Our major hypotheses to be tested are derived from New Trade Theory and its focus on imperfect international markets. Inter alia, we will specify memberships in intergovernmental organizations, alliances, and arms control regimes in order to estimate their effectiveness and their influence on trade deflections. The complexity of the underlying data sources requires the development of new inferential statistical techniques for the modeling of valued (i.e.: non-binary) networks - which will constitute the methodical core objective of our collaborative interdisciplinary project. This will allow us to take into account for the first time endogenous network structures and dynamics (e.g. sender effects indicating cumulative advantages, hyper-dyadic system features like triadic closures) going thereby beyond to so far investigated hypotheses based exclusively on attributes of the actors.