Ricardo correctly predicted that industrial-era jobs would be automated. He completely failed to predict that new jobs would replace them, and that prime-age employment in 2026 would be near an all-time high. The lump-of-labor fallacy has fooled experts for two centuries — and may be fooling them again.