Economic Impacts of AI and The Need for Policy Preparedness

Predictions regarding the economic impacts of artificial intelligence (AI) vary significantly. Some analysts predict rapid productivity growth paired with widespread prosperity. Others foresee extensive job displacement, increased inequality, or a significant shift in income from workers to capital owners. The future isn’t certain, but the potential for major disruptions is significant. AI could reshape the labor market and income distribution in ways that necessitate substantial fiscal policy changes. However, creating robust fiscal policies requires time for evidence collection, interpretation, and policy development.

Reflecting on responses to the ‘China shock,’ much action occurred only after embedded economic and social damage was evident. If AI benefits accrue mainly to a few capital owners, their economic and political power would complicate future policy responses. Hence, policymakers should develop ‘fiscal insurance’ against the most significant risks. These preparations are not due to certainty about which policies will be needed but rather to ensure readiness when the need arises.

Risks and Policy Considerations

Two risks demand significant attention:

Worker Displacement

AI may generate jobs but also eliminate existing ones. Workers losing jobs might struggle to find suitable new roles. Extended unemployment or persistent earnings losses could occur, affecting workers’ health, families, and communities. The U.S. has Trade Adjustment Assistance offering training and income support to some trade-displaced workers. However, this narrow and burdensome model isn’t feasible for an AI-influenced economy.

A revamped system of adjustment assistance is necessary. It should be widely accessible to displaced workers without requiring proof of job loss cause. Support options might include temporary income support, training, job-search assistance, and wage insurance. Key questions about program design remain concerning eligibility, generosity, and duration. Mixed evidence from training programs highlights the need for further experimentation to identify effective solutions for displaced workers.

Income Shift from Labor to Capital

If AI increases the capital share of national income, further widening income disparities could result. Capital ownership concentration could boost the political influence of already-wealthy asset owners. Raising taxes on capital income, wealth, or consumption are familiar responses. Increased revenue, especially if AI boosts national income, could fund worker displacement cushioning policies.

Expanding financial asset ownership offers an alternative policy approach. Broader ownership might prove more socially appealing and politically sustainable than tax increases. Implementing broader ownership of capital returns involves challenges, such as establishing a sovereign wealth fund or creating individual equity stakes, each facing unique obstacles and design considerations.

Developing policies to address potential worker displacement and income shifts requires further refinement and testing. Policymakers should proactively prepare these solutions for eventual national implementation, ensuring readiness when responses become necessary.

Douglas Elmendorf, former Congressional Budget Office Director and Harvard Kennedy School Dean, and Louise Sheiner, Policy Director at the Brookings Institution’s Hutchins Center, advocate for early policy groundwork. Their insights underscore the importance of preparations in addressing AI’s economic challenges.