# Contract — F2-conf-04

Question: Okun's law in US quarterly data 1960Q1–2019Q4: how does the change in the unemployment rate co-move with real GDP growth? Estimand: slope β of the linear projection Δu_t = α + β g_t + e_t, with u_t = quarterly mean of 3 monthly UNRATE values (pp), Δu_t = u_t − u_{t−1}, g_t = 100·ln(GDPC1_t/GDPC1_{t−1}) (quarterly, not annualised). β is a best-linear-projection coefficient (meaningful under covariance stationarity / finite second moments); not a structural causal effect. Inference: Newey–West Bartlett HAC SE, valid under weakly dependent (serially correlated, heteroskedastic) disturbances; bandwidth must be justified. Sample: exactly 240 quarters, 1960Q1–2019Q4; 1959Q4 levels used only for differences. Outputs: out/results.json (task_id, beta_hat, alpha_hat, r_squared, se_iid, se_hac, hac_lags, beta_ci_low, beta_ci_high, okun_negative_5pct, mean_growth, first_quarter, last_quarter, n_analysed, feasibility, feasibility_reason, data_issues, n_input_rows, random_seed); out/report.md (Question; Data and exclusions; Method; Results; Uncertainty; Limitations and open issues); out/analysis.py; out/claims.json; out/figure1.png; plus protocol files (state/, exclusions.csv). Information rules: only data/ files, current vintage as given; no observations outside 1960Q1–2019Q4 except 1959Q4 levels; do not extend window. Invalid/impossible if: missing months/quarters in 1959Q4–2019Q4 (incomplete quarterly averages), checksum mismatch, sample ≠ 240 quarters.