Spatial modeling of dyadic geopolitical interactions between moving actors

Datos Bibliográficos

ID6342153
AutoresSangyeon Kim (0000-0001-9105-8598, Pennsylvania State University, autor de correspondencia), Howard Liu (0000-0002-3479-9864, University of Essex), Bruce Desmarais (0000-0002-3031-8883, Pennsylvania State University)
Año2023
Volumen11
Número3
Páginas633-644
Fecha de publicación2023-07-01
Peer ReviewedSí
Open AccessSí
TipoARTICLE
RevistaPolitical Science Research and Methods (JOURNAL)
Identificadores de la revistaISSN: 2049-8470 • E-ISSN: 2049-8489
EditorialCambridge University Press (CUP) (PUBLISHER)
DOI10.1017/psrm.2022.6
OpenAlexW4220696572
IdiomaEN
Citas recibidas2
Referencias citadas24

Political actors often interact spatially, and move around. However, with a few exceptions, existing political research has analyzed spatial dependence among actors with fixed geographic locations. Focusing on fixated geographic units prevents us from probing dependencies in spatial interaction between spatially dynamic actors, which are common in some areas of political science, such as sub-national conflict studies. In this note, we propose a method to account for spatial dependence in dyadic interactions between moving actors. Our method uses the spatiotemporal histories of dyadic interactions to project locations of future interactions—projected actor locations (PALs). PALs can, in turn, be used to model the likelihood of future dyadic interactions. In a replication and extension of a recent study of subnational conflict, we find that using PALs improves the predictive performance of the model and indicates that there is a clear relationship between actors’ past conflict locations and the likelihood of future conflicts

Economic geography · Extension (predicate logic · Geography · Geopolitics · Political science · Politics · Regional science · Replication (statistics · Sociology · Statistics · Computer Science · Electoral Systems and Political Participation · Mathematics · Policy Transfer and Learning · Qualitative Comparative Analysis Research