# Verification checks (P4)

• Clean re-run: results.json byte-identical after deleting outputs: PASS
• CI contains beta: PASS
• SEs positive: PASS
• n_analysed=240: PASS
• exclusions sum: PASS
• decision consistent: PASS
• CI width = 21.96se_hac: PASS
• mean identity alpha+beta*mean_g = mean_du: PASS
• magnitude: mean_growth ~0.75 %/q (~3%/yr, not annualised): PASS
• HAC SE > iid SE consistent with positive resid AC1: PASS
• quarters: PASS
• independent recompute beta: PASS