In this paper, we analyse the network of international major conventional weapons (MCW) transfers from 1950 to 2016, based on data from the Stockholm International Peace Research Institute (SIPRI). The dataset consists of yearly bilateral arms transfers between pairs of countries, which allows us to conceive of the individual relationships as part of an overall trade network. For the analysis, we extend the separable temporal exponential random graph model (STERGM) to account for time‐varying effects on both the network level (trade network) and the actor level (country effects). Our investigation enables the identification of potentially differing driving forces that influence the formation of new trade relationships versus the persistence of existing ones. In accordance with political economy models, we expect security‐ and network‐related covariates to be most important for the formation of transfers, whereas repeated transfers should prevalently be determined by the importers’ market size and military spending. Our proposed modelling approach corroborates the hypothesis and quantifies the corresponding effects. Additionally, we subject the time‐varying heterogeneity effects to a functional principal component analysis. This analysis serves as an exploratory tool and allows us to identify countries with exceptional increases or decreases in their tendency to import and export weapons.

中文翻译：

---

具有角色异质性的动态可分离网络模型：在全球武器转让中的应用*

在本文中，我们根据斯德哥尔摩国际和平研究所（SIPRI）的数据分析了1950年至2016年的国际主要常规武器（MCW）转移网络。该数据集由成对的国家之间的年度双边武器转让组成，这使我们能够将个体关系构想为整个贸易网络的一部分。为了进行分析，我们扩展了可分离的时间指数随机图模型（STERGM），以考虑对网络级别（贸易网络）和参与者级别（国家影响）的时变影响。我们的研究使人们能够识别出潜在的不同驱动力，这些驱动力会影响新贸易关系的形成与现有贸易关系的持久性。根据政治经济学模型，我们预计与安全和网络相关的协变量对于形成转移最为重要，而重复转移应主要由进口商的市场规模和军事支出决定。我们提出的建模方法证实了这一假设并量化了相应的影响。此外，我们对时变异质性影响进行功能主成分分析。这种分析是一种探索性工具，使我们能够确定进出口武器趋势有异常增加或减少的国家。我们提出的建模方法证实了这一假设并量化了相应的影响。此外，我们对时变异质性影响进行功能主成分分析。这种分析是一种探索性工具，使我们能够确定进出口武器趋势有异常增加或减少的国家。我们提出的建模方法证实了这一假设并量化了相应的影响。此外，我们对时变异质性影响进行功能主成分分析。这种分析是一种探索性工具，使我们能够确定进出口武器趋势有异常增加或减少的国家。