Post-Work Requires Power: Saved Hours, Human Activity and the Allocation of Future Electricity

Automation is usually framed as either an electricity cost or an efficiency gain. This working paper argues that both framings omit a critical planning term: the electricity and broader energy-system access required to support the human activity made possible when compulsory work is reduced. It identifies a dual expansion mechanism in which cheaper machine intelligence may increase machine-side electricity demand while released human time may increase human-side activity and energy use. The paper introduces Human Energy Allocation as a planning lens and uses Japan as its first field edition because the country’s FY2026–FY2035 electricity forecast explicitly identifies data-centre and semiconductor demand. The historical analysis finds no reliable fixed electricity–freedom breakpoint across the preferred cross-sections. The paper therefore reaches a narrow normative conclusion: current constrained demand should not be treated as either a planning target or a moral ceiling.