Spatial and Network Interdependence in Politics

This proposal envisages a workshop where political methodologists who work in the areas of spatial econometrics and network statistics exchange ideas on the future of the statistical analysis of network interdependence in political science. The Arab Spring showed how popular protest can spread like wildfire across a region of countries; the “systemic risk” of banks has become a key political issue; when Estonia adopted a flat tax, it was soon followed by its Baltic neighbours, Romania, Slovakia and other countries. These three very different examples all point to the same issue: the increased recognition that many political processes and policy-making decisions cannot be seen in isolation. Policy-makers and other actors learn from neighbours, they feel the competitive pressure from neighbours, they can be forced by neighbours, or they react to the same stimuli that neighbours react to. The data structure in these studies is one of interdependence among neighbouring units – with proximity either defined in a geographical sense or as a social network. The statistical methodology related to this interdependence is that of spatial econometrics or network statistics. The statistical analysis of spatial econometrics and of social networks are closely related, whereby the former is typically concerned with explaining variation in the units, taking account of the network interdependence, while the latter is typically concerned with explaining the variation in the connectivity. Whereas so far most applications in political science have been empirical applications of existing techniques from the field of spatial econometrics and social network statistics, in recent years there has been more focus on developing new methodologies specifically of use to applications typical in political science. The aim of this workshop is to bring together applied researchers and methodologists who work at the cutting edge of spatial and network statistical research applied to political science.