# Contract — F2-dev-01

Question: within-firm, within-year slope β of y on x in a firm–year panel, model y_it = a_i + g_t + βx_it + u_it (firm and year effects unrestricted, may correlate with x). Estimands: β (two-way FE / LSDV OLS slope); its sampling SE clustered by firm (arbitrary within-firm serial dependence, independence across firms). Meaningful under strict exogeneity of x conditional on a_i, g_t, and many clusters (G large) for cluster asymptotics. Baseline: pooled OLS slope with intercept. Outputs: out/results.json (task_id, beta_pooled, beta_fe, se_conventional, se_cluster, beta_ci_low, beta_ci_high, n_firms, n_years, 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. Information rules: only data/firm_panel.csv. Invalidity/impossibility: missing columns, x with no within variation after FE (collinear), too few firms for clustering, duplicate firm-year keys left unhandled.