The Limits of TERGM for Analyzing Regional Alliance Networks: A Case Study of Post-Soviet Alliances

The article examines methodological aspects of applying dynamic network analysis in contemporary international political science. The study is motivated by the limitations of standard statistical methods, particularly logistic regression, which ignores network autocorrelation and endogenous interdependence of observations, leading to biased estimates in the analysis of international processes. The paper aims to demonstrate the analytical potential and limits of applicability of Temporal Exponential Random Graph Models (TERGM) for studying the evolution of regional security subsystems.

Using the example of modeling the dynamics of military alliance networks in the post-Soviet space, a methodology for constructing and specifying TERGM models is presented. The procedure for incorporating both endogenous structural configurations (inertia, triadic closure) and exogenous covariates (trade flows, sanctions status, influence of external powers) into the analysis is described. The analysis was conducted in the R programming environment using the «btergm» package.

The practical challenges of interpreting stochastic modeling results in small-sample conditions are demonstrated. In particular, it is shown that the inclusion of deterministic covariates can lead to statistical distortions requiring specific interpretation. The limitations of the method in assessing the influence of node attributes, specifically sanctions status, are identified in cases where these attributes are present in only a small number of network participants. It is concluded that TERGM is an effective tool for measuring structural effects; however, when analyzing small regional network structures, its application requires rigorous preliminary data diagnostics and combination with alternative verification methods.

1. Alyev T., Stoyanova E., Chimiris E. Integration Strategies of the Post-Soviet Countries: Analysis Using the Social Graph Approach. Vestnik mezhdunarodnyh organizacij [International Organisations Research Journal]. 2023; 18(1):127–150 [In Russian]. https://doi.org/10.17323/1996-7845-2023-01-05.

2. Vakarchuk D.O. The Features of the USA and the PRC Digital Diplomacy in the Mediterranean Region. Vestnik SPbGU. Mezhdunarodnye otnosheniya [Vestnik of Saint Petersburg University. International Relations]. 2024; 17(4):411–430 [In Russian]. https://doi.org/10.21638/spbu06.2024.402.

3. Vakarchuk D.O., Borovinskiy A.V. Cohesion of Developing Countries in the UN General Assembly: A Quantitative Analysis of Co-Sponsorship Networks and Voting on Resolutions. Rossiya i sovremennyj mir [Russia and the Contemporary World]. 2024; 1(122):195–205 [In Russian]. https://doi.org/10.31249/rsm/2024.01.11.

4. Degterev D.A. Network Analysis of International Relations. Vestnik Sankt-Peterburgskogo universiteta. Ser. 6 [Vestnik of Saint Petersburg University. Series 6]. 2015; 4:119–138 [In Russian].

5. Desmarais B.A., Cranmer S.J. Statistical Mechanics of Networks: Estimation and Uncertainty. Physica A: Statistical Mechanics and its Applications. 2012; 391(4):1865–1876 [In English]. https:/doi.org/10.1016/j.physa.2011.10.018.

6. Hafner-Burton E.M., Kahler M., Montgomery A.H. Network Analysis for International Relations. International Organization. 2009; 63(3):559–592 [In English].

7. Hanneke S., Xing E.P. Discrete Temporal Models of Social Networks. Electronic Journal of Statistics. 2006; 4:585–605 [In English]. Available from: https://projecteuclid.org/journals/electronic-journalof-statistics/volume-4/issue-none/Discrete-Temporal-Models-of-Social-Networks/10.1214/10-EJS548.full.

8. Lee K.H., Xue L. Model Based Clustering of Semiparametric Temporal Exponential Family Random Graph Models. Stat. 2022; 11(1):e443 [In English]. https://doi.org/10.1002/sta4.443.

9. Leifeld P., Cranmer S.J. A Theoretical and Empirical Comparison of the Temporal Exponential Random Graph Model and the Stochastic Actor-Oriented Model. Network Science. 2015; 7(1):20–51 [In English]. https://doi.org/10.1017/nws.2018.26.

10. Leifeld P., Cranmer S.J. The Stochastic Actor-Oriented Model is a Theory as Much as It is a Method and Must be Subject to Theory Tests. Network Science. 2022; 10(1):19–24 [In English]. https://doi.org/10.1017/nws.2022.9.

11. Leifeld, P., Cranmer S.J., Desmarais B.A. Temporal Exponential Random Graph Models with btergm: Estimation and Bootstrap Confidence Intervals. Journal of Statistical Software. 2018; 83(6):1–36 [In English]. https://doi.org/10.18637/jss.v083.i06.

12. Mehrl M., Thurner P.W. Sharing Rivals, Sending Weapons: Rivalry and Cooperation in the International Arms Trade, 1920–1939. The Review of International Organizations. 2023; 18(1):1–28 [In English]. https://doi.org/10.1007/s11558-022-09468-w.

13. Whetsell T.A. Democratic Governance and Global Science: A Longitudinal Analysis of the International Research Collaboration Network. PLOS ONE. 2022; 17(5):e0267437 [In English]. https://doi.org/10.1371/journal.pone.0267437.

14. Wu G., Dan J. Do Self-Organization and Relational Embeddedness Influence Free Trade Agreements Network Formation? Evidence from an Exponential Random Graph Model. The Journal of International Trade & Economic Development. 2020; 29(7):816–837 [In English]. https://doi.org/10.1080/09638199.2020.1755146.