A Gini‑adjusted GDP per capita index reveals that welfare‑adjusted prosperity in the G7 has lagged headline GDP growth since 2010 and the shortfall widened sharply after 2022; the simple, auditable GAGI is offered as a regulatory trigger to spot distributional erosion that GDP alone misses.

Main Finding

The paper introduces the Gini‑Adjusted GDP per Capita Index (GAGI): a simple, auditable welfare monitoring statistic that rescales real GDP per capita by an inequality factor (1 − Gini) and by price level, normalized to a 2010 baseline. Applied to the G7 (2010–2026), GAGI shows that welfare‑adjusted prosperity has systematically diverged from headline GDP per capita growth, with the gap widening sharply after 2022. This divergence is temporally coincident with COVID after‑effects and the rapid deployment of generative AI, suggesting standard GDP monitoring can miss distributional harms that may accompany automation even when aggregate output grows.

Key Points

• Purpose: Provide an operational monitoring trigger — minimal, publicly computable, transparent, and year‑on‑year comparable — that flags changes in lived prosperity overlooked by GDP per capita.
• Definition: GAGI rescales real GDP per capita by (1 − G), where G is the income Gini coefficient, and adjusts for price level; the index is normalized so GAGI(2010) = 100.
• Properties: Reproducible from public data, low modelling burden (uses observed Gini and GDP per capita), auditable across countries and years.
• Empirical result (G7, 2010–2026): Welfare‑adjusted prosperity diverges from headline growth persistently; the divergence increases markedly after 2022.
• Causal stance: The observed post‑2022 widening is temporally coincident with COVID after‑effects and accelerating generative‑AI deployment, but the paper does not claim causal proof linking AI to GAGI changes.
• Policy claim: GAGI should complement—not replace—GDP metrics; relying solely on aggregate output risks missing distributional harms from automation.

Data & Methods

• Core formula (operational form):
  Let GDPpc_t be real GDP per capita (price‑adjusted, constant base year or PPP), and G_t be the income Gini for year t. Define GAGI_t = [(GDPpc_t) × (1 − G_t)] / [(GDPpc_2010) × (1 − G_2010)] × 100. This yields an index with 2010 = 100; alternative equivalent formulations first convert GDPpc to a common price base (real/PPP) before applying the (1 − G) factor.
• Data sources (typical, publicly available): national accounts (World Bank, IMF, OECD) for GDP per capita (constant prices or PPP), household survey or national statistics for annual Gini (World Bank WDI, OECD, Luxembourg Income Study, or WID), and price‑level series for real adjustment (CPI/PPP conversion factors).
• Computation: simple algebraic rescaling performed annually for each country; no parametric welfare function or distributional modelling required.
• Empirical sample: G7 countries, 2010–2026. (2023–2026 years may include provisional GDP/Gini estimates depending on source.)
• Robustness / sensitivity checks: alternate inequality measures (consumption‑based Gini, wealth Gini), alternate price adjustments (constant‑price GDP vs PPP), and alternative baseline years recommended.

Implications for AI Economics

• Monitoring: GAGI provides a compact, operational trigger for regulators and policymakers to detect when aggregate growth masks falling or stagnating welfare for households — a key early‑warning for harms potentially associated with automation and AI diffusion.
• Policy targeting: A falling or diverging GAGI suggests distributional interventions (progressive taxation, transfers, retraining, wage support) may be needed even amid positive GDP growth; it helps prioritize jurisdictions or subperiods for deeper causal investigation.
• Research agenda: Use GAGI as an outcome variable in studies linking firm‑ or sector‑level AI adoption to welfare outcomes; complement with microdata to attribute GAGI changes to displacement, wage polarization, rental/wealth effects, or price shocks.
• Practical adoption: Publish GAGI alongside GDP in official statistics releases to make distributionally‑adjusted welfare transparent to policymakers, media, and the public.
• Limitations relevant to AI studies: GAGI signals distributional shifts but does not identify mechanisms (automation vs policy vs pandemic effects). It depends on the Gini (which has measurement issues and ignores within‑group heterogeneity and wealth), and lags in survey‑based inequality measures can delay detection.
• Next steps: Pair GAGI with higher‑frequency administrative indicators (tax records, wages, unemployment insurance claims), sectoral AI‑intensity measures, and inequality decompositions (by income source/quantile) to build a causal chain from AI deployment to welfare outcomes.

Short recommendation: Treat GAGI as a low‑cost, transparent, annual welfare monitor to complement GDP; when GAGI diverges negatively from GDP, prioritize targeted, causal diagnostics (microdata, firm‑level AI measures) and distributional policy responses.