The Impact of AI on the Labor Market: Beyond the Anxiety

In this episode of Thoughts on the Market, Seth Carpenter, Morgan Stanley’s global chief economist, cuts through the prevailing "hype and anxiety" surrounding artificial intelligence to examine its actual impact on the labor market. While public discourse is dominated by fears of mass unemployment, Carpenter argues that the economic reality is more nuanced, characterized by a complex interplay between productivity gains and structural adjustments.

The Fallacy of Zero-Sum Labor

Carpenter addresses the common assumption that because AI can perform tasks more efficiently—lowering the cost of production—it must inevitably lead to "millions of people" losing their jobs. He challenges this "same amount of output" logic, noting that it ignores the possibility that the economy could simply produce "a lot more output" using the same amount of labor. The debate, he asserts, hinges on whether AI functions as a labor-substitute or a labor-augmenter.

Early Evidence: Productivity vs. Displacement

According to Morgan Stanley’s research, the data currently offer "cautious optimism." Despite the "rapid advances in AI capability," broad labor market indicators show "remarkably little disruption." Crucially, industries with high AI exposure are experiencing "stronger labor productivity gains" driven by "faster output growth" rather than a reduction in labor hours. This suggests that, at this stage, firms are prioritizing output expansion over workforce contraction.

Physical Constraints and the Opaque Future

Carpenter highlights a critical "physical constraint" on the immediate acceleration of AI: infrastructure. He notes that out of the expected $3 trillion in "data center and related infrastructure CapEx" projected through 2028, only about 25% has been deployed. This bottleneck provides a buffer, though the future remains "opaque."

The Central Risk: Speed of Transition

While historical innovation waves occurred over decades, AI is "compressing the adjustment period." This creates a "central risk" where the rate of "job destruction" might temporarily outpace the rate of "new job creation." Carpenter acknowledges that "some job losses are likely unavoidable" as the economy navigates this rapid technological shift.

The General Equilibrium Perspective

To understand the long-term outcome, Carpenter urges listeners to adopt a "general equilibrium" view. He argues that higher productivity translates into "higher incomes" and "higher wealth," which ultimately drive consumer spending and economic growth. Furthermore, he emphasizes that the labor market is not static; "new tasks and new roles will likely emerge" within corporations to absorb displaced workers.

The Role of Policy Buffers

Should AI-driven shifts put "downward pressure on inflation" or lead to temporary spikes in unemployment, Carpenter is confident that policymakers will not remain on the sidelines. Central banks can "stimulate the economy" to return to full employment, and fiscal policymakers can leverage "automatic stabilizers" like unemployment benefits. These institutional buffers suggest that any potential rise in unemployment will likely be "smaller, shorter and easier to manage" than initial alarmist analyses predict.

Conclusion: Productivity Ultimately Wins

Carpenter concludes that the impact of AI is "not predetermined." The defining factor will be the speed of adoption relative to the economy's "ability to adapt." While history provides comfort that "productivity ultimately wins"—leading to a larger economy and stable employment—he warns against being "blithely optimistic." The transition may not be smooth, and not everyone will benefit equally, but for now, the early evidence suggests a manageable evolution rather than a cataclysmic disruption.