Research: Job Displacement—Evidence and Trajectories

Content Summary

Current empirical evidence on AI-driven job displacement (as of late 2025):

Existing Impact:

14% of workers have already experienced job displacement due to automation or AI
Occupations with higher AI exposure experienced larger unemployment rate increases (0.47 correlation coefficient)
Computer and mathematical occupations saw steepest unemployment rises
Unemployment among 20–30-year-olds in tech-exposed occupations rose 2.8 percentage points since early 2025

Public Concern:

~66% of UK public believe AI likely to increase unemployment
~52% of 18–24-year-olds worry about AI impact on careers
~30% of workers across 44 countries concerned about displacement in next 3 years

Projected Impacts:

Innovation-driven AI could displace 6–7% of US workforce if widely adopted
By 2030, potentially 30% of US jobs fully automated; 60% experiencing significant task-level changes
However, if current AI use cases were only moderately expanded, Goldman Sachs estimates only 2.5% at risk

Occupations at Highest Risk:

Computer programmers
Accountants and auditors
Legal and administrative assistants
Customer service representatives
Bank tellers, cashiers
Clerical and administrative roles

Equity Concerns:

79% of employed women work in high-risk roles (vs. 58% of men)
Entry-level opportunities shrinking (particularly concerning for first career rung)

Wage Effects (Complex Picture):

Workers in AI-exposed sectors sometimes see faster wage growth than less-exposed workers (suggesting complementarity, not pure substitution)
But overall concern that productivity gains accrue to tech firms and capital owners, widening inequality and hollowing out middle-income jobs

Current Usage

Chapter 11 mentions job displacement but without the granular data. The manuscript's treatment of work displacement is more narrative (Chantal's experience) than empirical.

Unused Material

Most of the empirical detail is unused:

The 14% Already-Displaced Figure – This is concrete evidence that AI displacement is happening now, not a future risk. Deserves prominent mention.
Gender Disparity – 79% of women vs. 58% of men in high-risk roles is a critical equity issue largely absent from the manuscript.
Wage Complementarity Evidence – The finding that some AI-exposed workers see faster wage growth challenges the pure-substitution narrative and deserves more space.
The Goldman Sachs Projection Variance – The gap between "6–7% displaced if widely adopted" vs. "2.5% at risk under modest expansion" reveals how much displacement depends on how fast and how widely AI is deployed. This is policy-contingent, not technologically inevitable.
Entry-Level Job Collapse as Unique Risk – Unlike previous automation waves, this one threatens early-career job-finding. This is a real and distinct risk that merits specific discussion.

Suggested placements:

Chapter 2 or 3: Establish empirical baseline—AI displacement is happening now (14% figure)
Chapter 5 or 8: Gender disparity as equity concern in UBI discussions
Chapter 11 or 12: Wage complementarity evidence; debate about whether displacement is inevitable or policy-contingent
Chapter 1 or 6: Entry-level collapse as distinct from historical automation fears
Chapter 14: Data on public concern (66% unemployment worry, 52% youth concern) as evidence of real anxiety

Connections

Core to the manuscript's argument that material risks are real and require policy response:

automation-and-displacement – Current trajectory and acceleration
work-disincentive – Fear driving resistance to change
identity-through-work – Career identity at risk, not just income
universal-basic-income – Policy response to displacement

Notes

Strengths: Current, specific data with citation trail. The wage-complementarity finding is particularly valuable as it complicates the simple "job loss" narrative.

Limitations: Some projections (2030 figures) are model-dependent and should be treated with uncertainty. The distinction between "tasks displaced" vs. "jobs displaced" is important—a job can lose 60% of its tasks without the worker being displaced. That nuance is mentioned but could be clearer.

Quality: High. This is concrete empirical evidence with a clear evidentiary basis and transparent limitations.

Recommendation: Use the 14% already-displaced figure prominently. Use wage-complementarity evidence to complicate the pure-substitution narrative. Flag gender disparity as equity concern.