Research Directions: Towards a cross-sectoral quantitative approach to societal dynamics

Context and Motivation

Contemporary societies face deeply interconnected challenges: economic, social and political crisis, climate change, ecological breakdown, planetary boundaries, information and technological transformations. These phenomena are difficult to understand in isolation – they mutually influence each other and form complex systems whose dynamics remain underexplored.

Scientific gap: Dominant approaches tend to remain disciplinarily siloed (economics, health, sociology, political science, ecology, etc.) or focused on isolated indicators. There is limited systematic research analyzing the interdependencies between these different dimensions and their impact on the evolution of our societies. While some pioneering institutions like the Santa Fe Institute, MIT Media Lab, and INET Oxford are developing quantitative, cross-sectoral or complexity-based approaches to societal dynamics, such research appears to be taking time to develop more broadly. The Meadows Report (1972) illustrates how models simultaneously integrating multiple dimensions (demography, economy, and environment) enabled new understanding of certain global dynamics.

Opportunity: The accumulation of many decades of multidimensional data now makes it possible to take contemporary societies as objects of study and apply computational methods to characterize societal trajectories, identify resilience or fragility factors, and potentially better inform certain public policies to better address these challenges.

Interactive Causal Relations Model

Contemporary societies show extensive interconnections between economic, political, environmental, and social dimensions. While the nature of each causal relationship remains complex, debated and non-deterministic, these interconnections show the interest of transversal/integrated approaches to understand better societal dynamics.

Axis 1: Characterizing and modeling societal trajectories

Typology and clustering of national development trajectories
  • Automatic classification of countries according to their multidimensional evolution: current position, past trajectory, projections
  • Identification of phase space subsets to characterize development "archetypes"
Early detection of societal vulnerabilities
  • Detection of pre-critical patterns and statistical signals
  • Learning from historical crises (2008, Arab Spring, COVID, etc.)
  • Optimal aggregation of heterogeneous indicators and construction of composite fragility indices
Automatic detection of regimes and transitions
  • Identify regime changes (socio-political, economic, etc.) in time series
  • Characterize periods of stability vs instability
  • Analysis of emergence conditions and sustainability of transitions
Resilience and adaptive capacity analysis
  • Measure how different social systems respond to shocks and disturbances (recovery time after shock: COVID, 2008, extreme climate events, etc.)
  • Identify multidimensional factors associated with resilience
  • Analysis of path dependency and systemic lock-ins (growth, AI, energy, etc.)

Axis 2: Coupling phenomena and causal analysis

Analysis of coupling and decoupling phenomena
  • Identify different couplings/decouplings of certain indicators (GDP, innovation, well-being, energy consumption, carbon footprint, biodiversity, AI, productivity, etc.)
  • Study of transition trajectories and coupling/decoupling conditions
  • Understand institutional / social / structural / territorial mechanisms enabling coupling/decoupling of certain indicators and environmental/economical/political transitions
Exploring advanced modeling approaches: from data-driven to hybrid models
  • Explore purely data-driven approaches (convolutional neural networks on temporal trajectories) even if their causal interpretation is limited
  • Investigate hybrid approaches coupling mechanistic models with machine learning to explore causal reasoning capability
Differential impact of political orientations on societal trajectories
  • Analyze the impact of left/right political alternations (CPDS right-left scale, V-Dem economic right-left scale) on economic, social and environmental indicators while trying to correct for international dynamics and inertia effects
  • Evaluate comparative effectiveness of progressive vs conservative policies across domains (inequality, competitiveness, innovation, environment, social welfare)
  • Identify contextual conditions (development level, institutions, political culture) modulating these effects
Identification of high-impact intervention levers
  • Identify policies with the strongest multiplier effects
  • Multi-dimensional optimization for decision-makers

Why it matters

This type of research program, though highly ambitious, could help improve how we understand and navigate societal challenges:

For Policymakers
  • Better identify high-leverage interventions with positive cross-sectoral effects
  • Better quantify long-term consequences of policy choices
  • Learn more systematically from other countries' successes and failures
  • Attempt to explore more anticipatory rather than reactive governance
For Researchers
  • Support the development of quantitative theories of societal dynamics
  • Create reproducible methodologies for comparative analysis
  • Participate in bridging complex systems science and social sciences
  • Enable testing of hypotheses about societal evolution
For Citizens
  • More transparent, data-driven understanding of societal trends
  • Better informed public debate based on data
  • Ability to hold leaders accountable with objective metrics
  • Better understanding of what has worked elsewhere and why
  • Define more consciously the development direction we wish to pursue