Check out these two new twin papers by that take on a core problem in Australian policy: we talk a lot about social cohesion, but we do not measure it well.
The aim across the two papers is straightforward. The first paper, led by Rouven Link, maps how fragmented and inconsistent definitions of social cohesion are across disciplines and policy. The second, led by Matteo Vergani, builds on this and proposes a new framework to define and measure cohesion as a system of relationships across society.
The methodology is rigorous and cumulative. The first paper reviews existing frameworks in Australia and internationally, including major survey-based programs such as the Scanlon Index and comparable initiatives in Germany and Chile. It identifies how these frameworks differ in assumptions, indicators, and policy use. The second paper then develops a network-based model. Social cohesion is defined as the pattern of connections between people, groups, and institutions, assessed along three dimensions: durability, inclusivity, and quality of ties. Crucially, the measurement strategy expands beyond surveys. It integrates survey data with social media data, administrative datasets such as PLIDA and BLADE, and public discourse analysis. The authors also outline concrete pilots, including network mapping using large-scale digital data and monitoring of public narratives.
The main result is a shift in how cohesion is understood and measured. Instead of treating it as an aggregate of individual attitudes, the framework treats cohesion as a dynamic system. This allows researchers and policymakers to analyse how cohesion forms, how it breaks down, and how it responds to events and shocks. It also separates cohesion from its enabling conditions, such as institutions and infrastructure, which improves analytical clarity.
A key contribution here is the work led by the Tackling Hate Lab. Vergani and Giovannetti have been central in pushing a definition of social cohesion that can be measured using multiple data sources, not just surveys. The integration of social media and other behavioural data is a major step towards real-time, local, and policy-relevant measurement.
The policy implications are clear. Governments can move beyond annual survey snapshots and towards continuous monitoring. This enables earlier detection of fragmentation, better targeting of interventions, and more precise evaluation of programs. For research, the framework opens a pathway to combine computational social science with traditional approaches in a coherent way.
It is also a strong collaborative effort, bringing together SoDa Lab’s Simon Angus, ANU’s Nick Biddle, and Scanlon Foundation’s Rouven Link, alongside the Tackling Hate Lab team and government practitioners Hugh Piper, Alex Fischer, and Melanie Rayment. A solid example of interdisciplinary work with direct policy relevance.