Resilience is a long-run institutional process, not a short-term rebound: societies that encode lessons from repeated shocks into laws, routines and fiscal architectures build durable adaptive capacity, while selective or politicized forgetting locks others into vulnerability — a dynamic that will shape how countries cope with AI-driven disruption.

Main Finding

Economic resilience should be reconceptualized not as an economy’s short-term response to singular shocks but as a long-run institutional process. Durable resilience emerges from institutional memory and politically mediated, selective learning from recurrent shocks. A dynamic feedback framework—where each shock’s outcomes update institutional memory and shape future responses—explains why societies diverge into resilient vs. vulnerable trajectories over time.

Key Points

• Event-based metrics (speed of recovery, output loss after a single crisis) are limited; they miss cumulative, institutional change.
• Institutional memory is the mechanism through which past crisis experience becomes durable: encoded in laws, administrative routines, fiscal/monetary arrangements, and informal governance norms.
• Learning from shocks is selective and politically mediated:
  • Experience can be retained and translated into adaptive reforms;
  • Experience can be reinterpreted to legitimize existing arrangements;
  • Experience can be forgotten, reproducing vulnerability.
• Institutional responses are path dependent: outcomes of one shock change the structure of institutional memory and thus alter responses to later shocks.
• The framework links institutional economics, historical institutionalism, and economic history to explain long-run divergence in resilience across societies and periods.
• Paper positions the framework as a foundation for comparative historical research (e.g., Asia–Pacific cases facing repeated economic, political, and trade disruptions).

Data & Methods

• Nature of the paper: conceptual and theoretical synthesis rather than new statistical estimation.
• Methods used/proposed:
  • Literature synthesis across institutional economics, historical institutionalism, and economic history to build the framework.
  • Development of a dynamic feedback model (conceptual) in which shocks → outcomes → updates to institutional memory → altered future responses.
  • Suggested empirical approaches for follow-up work include comparative historical case studies, process tracing, archival research, and cross-country historical comparison to track how crisis experience is encoded (or not) into institutions.
• Operationalization guidance (implicit in framework):
  • Study codified rules (laws, regulations), administrative procedures, fiscal/monetary policy architectures, and informal norms as carriers of institutional memory.
  • Examine political mediation channels (actors, coalitions, veto points) to explain selective learning or forgetting.

Implications for AI Economics

• Institutional memory matters for AI-induced shocks: economies repeatedly exposed to automation, algorithmic disruption, and data-driven trade shocks will only be resilient if institutions encode prior lessons (labor protections, retraining systems, data governance).
• Selective, politically mediated learning implies uneven AI policy adoption: some countries will institutionalize adaptive AI governance (e.g., rules for algorithmic accountability, social safety nets for displaced workers), while others may reinterpret AI risks to preserve incumbent interests or fail to learn at all.
• Long-run path dependence warns of lock-in effects from early AI-regulatory choices: early institutional responses can bias future trajectories toward either robust AI adaptation or persistent vulnerability.
• Research applications:
  • Use AI tools (text mining, NLP) to extract institutional memory signals from legal archives, parliamentary debates, and administrative records to measure how societies encode crisis lessons.
  • Build computational models (agent-based or dynamic systems) that simulate shock–institution feedbacks to study counterfactual resilience trajectories under different policy rules.
  • Comparative historical studies of previous technological disruptions (industrialization, ICT waves) can inform likely institutional responses to AI and identify transferable policy designs.
• Policy takeaway: strengthening durable adaptive capacity for AI requires deliberate efforts to record, translate, and institutionalize lessons from each disruption—through codified rules, resilient administrative routines, fiscal buffers, and inclusive politics that prevent selective forgetting.