This paper extends the Labor Flow Network (LFN) framework from a static, time-aggregated representation of worker mobility to an explicitly dynamic account of how inter-firm links evolve and shape macroeconomic outcomes. Using rich matched employer–employee data for Finland and Mexico, we construct rolling-window LFNs and show that they combine a stable backbone of persistent mobility channels with substantial annual link turnover of around 15–25%. We document that firms’ trajectories are tightly linked to their evolving network positions: firms that expand their connections and occupy bridging roles between otherwise weakly connected clusters experience systematically higher employment growth and lower exit risk, whereas firms in shrinking or isolated neighborhoods face substantially elevated failure hazards. Building on an agent-based model calibrated to the data, we then compare the propagation of firm-level and sectoral shocks under static versus dynamic network assumptions. When shocks are simulated on a fixed LFN, unemployment spikes but decays relatively quickly; when the empirically observed network evolution is incorporated, unemployment becomes more persistent and spatially diffuse because the very mobility channels needed for reallocation deteriorate during the downturn. Finally, we propose a network-based measure of regional economic resilience—the inverse area under the post-shock unemployment curve—and show that pre-shock LFN topology is strongly predictive of resilience across Finnish regions. Dense, redundantly connected, modular networks with robust bridging firms recover faster, while over-centralized or fragmented structures are prone to long-lasting distress. The results highlight that labor markets are dynamic networked systems and that policies which preserve and strengthen key mobility channels can play a pivotal role in mitigating the long-run costs of economic shocks.
This paper extends the Labor Flow Network (LFN) framework from a static, time-aggregated representation of worker mobility to an explicitly dynamic account of how inter-firm links evolve and shape macroeconomic outcomes. Using rich matched employer–employee data for Finland and Mexico, we construct rolling-window LFNs and show that they combine a stable backbone of persistent mobility channels with substantial annual link turnover of around 15–25%. We document that firms’ trajectories are tightly linked to their evolving network positions: firms that expand their connections and occupy bridging roles between otherwise weakly connected clusters experience systematically higher employment growth and lower exit risk, whereas firms in shrinking or isolated neighborhoods face substantially elevated failure hazards. Building on an agent-based model calibrated to the data, we then compare the propagation of firm-level and sectoral shocks under static versus dynamic network assumptions. When shocks are simulated on a fixed LFN, unemployment spikes but decays relatively quickly; when the empirically observed network evolution is incorporated, unemployment becomes more persistent and spatially diffuse because the very mobility channels needed for reallocation deteriorate during the downturn. Finally, we propose a network-based measure of regional economic resilience—the inverse area under the post-shock unemployment curve—and show that pre-shock LFN topology is strongly predictive of resilience across Finnish regions. Dense, redundantly connected, modular networks with robust bridging firms recover faster, while over-centralized or fragmented structures are prone to long-lasting distress. The results highlight that labor markets are dynamic networked systems and that policies which preserve and strengthen key mobility channels can play a pivotal role in mitigating the long-run costs of economic shocks.
