Developing Bouquet Structure as a Network Analytics Measure

The second speaker in this ACSPRI 2024 conference session is Eve Cheng, whose interest is in party structures in parliamentary networks – party structures here means personal and professional backgrounds, including military and civilian careers, party memberships, educational track records, etc.

The assumption here is that such backgrounds might determine party structures, predict electoral success, and affect policy-making. Key metrics here were average maximal flow (assessing the global network) and transitivity (focussing on local structures), and a comparison between these two networks is especially interesting.

In 1960s Australian Labor, for instance, there was one large trade unionist cluster, with one parliamentarian connecting this to the rest of the party room (which was unified especially by its military background), and a structure like this can be assessed through a ‘bouquet structure’ measure which identifies why someone would be an important connector in the network. This can be done by computationally running through a number of possible permutations of bouquet size.

Working out such bouquet structures, then, it becomes possible to identify party structures, and amongst such party structures it becomes interesting to examine whether those individuals who serve as key connectors are also the ones with the greatest electoral success or party influence.