[
 {"id":"C1","text":"Over 1960Q1-2019Q4 the Okun coefficient is about -0.256 pp per 1% quarterly real GDP growth, and its HAC 95% CI [-0.326,-0.186] lies below zero.","results_keys":["effect_estimate","effect_se","hac_lags","effect_ci_low","effect_ci_high","conclusion"],"status":"supported","caveat":"Reduced-form association, not causal; revised data."},
 {"id":"C2","text":"Output growth explains a substantial part of unemployment changes: residual RMSE 0.250 vs 0.325 pp for the mean-only baseline (R2 0.41).","results_keys":["model_metric","baseline_metric","r_squared"],"status":"supported","caveat":"In-sample fit only."},
 {"id":"C3","text":"The classical Chow test rejects joint equality of (alpha, beta) across 1960-1983 and 1984-2019 (F=6.09, p=0.0026).","results_keys":["chow_f_stat","chow_p_value","stability_conclusion","chow_df"],"status":"supported","caveat":"Classical test assumes equal, homoskedastic, uncorrelated errors; residual variance differs across subperiods and residuals are autocorrelated, so size may be distorted."},
 {"id":"C4","text":"The subperiod slopes are similar (-0.259 vs -0.273); the instability appears to come mainly from a lower intercept after 1984.","results_keys":["beta_sub1","beta_sub2","beta_sub1_hac_se","beta_sub2_hac_se","intercept_sub1","intercept_sub2"],"status":"supported","caveat":"Descriptive; no separate slope-only or intercept-only test was run."},
 {"id":"C5","text":"The HAC standard error is insensitive to the lag choice (0.034-0.037 for lags 2-12), and the sign conclusion holds for all lags tried.","results_keys":["effect_se_lag_sensitivity","ols_classical_se_beta"],"status":"supported","caveat":"Bartlett kernel only."}
]