{
  "citation": "statsmodels 0.15.0: the reference specification of fujiwara_2015 tab2_pb_valid94_c3 (infer.rdd, figure 'effect') on an independent code path",
  "deviation_evidence": null,
  "doi": null,
  "excerpt": null,
  "independent": true,
  "item_set": null,
  "kind": "software",
  "known_deviation": null,
  "licence": null,
  "locator": "fujiwara_2015: Table II, Panel B (placebo tests), row 'Valid Votes/Turnout (1994 Election)', column (3), bandwidth 5,000",
  "metric": null,
  "n_items": null,
  "page": null,
  "read_copy_sha256": null,
  "retrieved_at": "2026-10-15",
  "se": 0.03278673393864601,
  "se_as_published": null,
  "source_id": "fujiwara_2015.tab2_pb_valid94_c3.software",
  "tolerance_reason": null,
  "units": null,
  "url": "https://pypi.org/project/statsmodels/0.15.0/",
  "value": 0.006370183782644062,
  "value_as_published": null,
  "verified_by": "software_run",
  "verified_note": "sm.OLS(r_util94, [const, treat, d·(1-treat), d·treat]).fit(cov_type='HC1') with d = voters96 - 40500, rows |d| <= 5000, treat = d >= 0 on munic (munic.dta of the file pinned at sha256 5cd5b8796355e025807dc303263b1fbb16f442c3c7551f14ee3557768a310894); n_obs = 4809, n_left = 75, n_right = 41. The tool infer.rdd (magi_infer, numpy) does not call statsmodels. Note: the strict rule (|d| < h, treated when d > 0) selects the same rows and treatment. Written by scripts/paper_repro_reference.py."
}