U.S. productivity is being held back not by a lack of AI innovation but by workforce frictions—mid-career barriers, splintered credentials, and employer adoption gaps—that block diffusion and widen wage gaps; fixing these institutional bottlenecks with demand-aligned, scalable training and regional partnerships is the paper's central policy prescription.

Main Finding

The binding constraint on U.S. productivity growth is increasingly workforce and institutional frictions—rather than frontier technology alone. Rapid AI and automation advances are not translating into broad-based productivity gains because mid-career transition barriers, fragmented credentialing, employer adoption gaps, and geographic mismatches prevent timely, scalable, and relevant skill diffusion across firms and regions.

Key Points

• Central argument: Technological capability at the frontier is outpacing the workforce system’s capacity to supply, validate, and deploy the skills needed for widespread adoption and productivity diffusion.
• Principal bottlenecks identified:
  • Mid-career transition barriers (time, cost, credential recognition, employer hiring practices) that block occupational mobility and reskilling at scale.
  • Fragmented credentialing and training markets that limit portability and signal value to employers.
  • Employer adoption gaps where organizational barriers (integration costs, managerial skills, labor relations) slow internal deployment of new technologies even when skills exist.
  • Geographic misalignment between where skills are supplied and where demand is growing, reinforcing regional divergence.
• Consequences mapped:
  • Suppressed technology diffusion across non‑frontier firms and regions.
  • Reinforced wage divergence as frontier adopters capture productivity premiums.
  • Constrained national competitiveness and slower aggregate productivity growth.
• Policy prescription (sequenced, actionable):
  • Scale mid-career pathways (short, modular upskilling with employer buy‑in and financing support).
  • Develop demand‑aligned, portable credentials tied to employer needs and regional industry clusters.
  • Create employer-centered adoption systems (incentives, shared implementation support, managerial training).
  • Build regional labor–industry partnerships to align training pipelines with local industry demand and reduce geographic friction.
• Positioning: Part of the Productivity Frontier Series; complements work on innovation, diffusion, and structural mobility (e.g., PFS-00 and SMTS-00).

Data & Methods

• Mixed-methods approach combining:
  • Conceptual mapping of how workforce frictions interact with technology diffusion.
  • Empirical mapping across firms, regions, and labor-market institutions using administrative and survey data (employment, wages, training participation) and firm/industry indicators to identify patterns of adoption and skill gaps.
  • Case studies and employer surveys/interviews to document organizational adoption barriers and credential signaling issues.
  • Geographic decomposition to show regional misalignment between supply and demand and its association with adoption/diffusion outcomes.
  • Policy sequencing and scenario analysis to illustrate how targeted interventions could change diffusion dynamics (illustrative rather than definitive causal estimates).
• Analytical focus: identifying mechanisms and leverage points for policy rather than presenting a single causal estimate of national productivity gains.

Implications for AI Economics

• Modeling implications:
  • Macro and micro models of AI-driven growth must incorporate frictions in labor reallocation, credential portability, and firm-level adoption costs to avoid overestimating diffusion speed and aggregate gains.
  • Heterogeneous-firm frameworks should account for adoption gaps and managerial complementarity constraints that create persistent divergence between frontier and laggard firms.
• Measurement priorities:
  • Better, linked data on employer demand for AI-related skills, credential outcomes, and mid-career transitions is critical for evaluating diffusion and policy impact.
  • Standardized measures of credential portability and employer valuation will improve forecasts of labor-market responses to AI.
• Policy and evaluation:
  • Interventions should be tested with implementation‑focused evaluations (RCTs, phased rollouts, regional pilots) that measure both uptake of technology and labor-market mobility outcomes.
  • Policies that reduce frictions (portable credentials, employer adoption support, regional partnerships) may raise the marginal returns to AI investments and accelerate inclusive diffusion.
• Equity and distributional dynamics:
  • Without addressing these bottlenecks, AI adoption is likely to widen wage and regional inequalities as frontier firms and places capture productivity rents.
  • Targeted workforce strategies can help make AI-driven productivity growth more inclusive by enabling mid-career workers and non‑metro regions to participate in diffusion.
• Practical guidance for stakeholders:
  • Firms: invest in managerial and change-management capacity, partner with credential providers, and share implementation resources across industry consortia.
  • Policymakers: prioritize scalable, demand-aligned training, support credential portability, and fund regional labor–industry intermediaries to coordinate supply and demand.

Overall, the paper reframes the productivity puzzle: unlocking AI’s broad gains requires policy and institutional fixes to workforce systems and firm adoption processes as much as continued technological invention.