{"args":{},"checks":[{"abs_tol":null,"bind":"any","column":null,"failure_class":null,"field":"value","interval":false,"key":null,"kind":"answer_backed","min_lre":null,"note":"every number traced to a tool call or the question","open_interval":false,"rel_tol":null,"required":true,"source_id":null,"table":null,"tool":null,"value":null,"weight":1.0,"where":null},{"abs_tol":null,"bind":"answer_call","column":null,"failure_class":null,"field":"value","interval":true,"key":null,"kind":"answer_figure_close","min_lre":null,"note":"CRB estimate_within_solution_ci: the answer figure inside the gold 95% interval effect \u00b1 1.96\u00b7|SE| of source crb_papers.q081.gold (closed, evaluate_result.py; no --auto-rescale)","open_interval":false,"rel_tol":null,"required":true,"source_id":null,"table":null,"tool":null,"value":[-0.13094526994891065,-0.023037402860295993],"weight":1.0,"where":null},{"abs_tol":2.752751711444251e-08,"bind":"answer_call","column":null,"failure_class":null,"field":"value","interval":false,"key":null,"kind":"answer_figure_close","min_lre":null,"note":"replication cell: the gold effect to rel 1e-6 (abs floor 1e-6\u00b7SE); never required","open_interval":false,"rel_tol":1e-06,"required":false,"source_id":"crb_papers.q081.gold","table":null,"tool":null,"value":-0.07699133640460332,"weight":1.0,"where":null},{"abs_tol":2.752751711444251e-08,"bind":"answer_call","column":null,"failure_class":null,"field":"se","interval":false,"key":null,"kind":"answer_figure_close","min_lre":null,"note":"replication cell: the gold SE to rel 1e-6; never required","open_interval":false,"rel_tol":1e-06,"required":false,"source_id":"crb_papers.q081.gold","table":null,"tool":null,"value":0.027527517114442512,"weight":1.0,"where":null}],"datasets":{"ds_0051":{"columns":null,"description":"| Variable Name | Description |\n|-------------------------|-------------------------------------------------------------------------------------------------------------|\n| X | Original row index or internal identifier for the observation (file artefact). |\n| poll_id_07 | Polling-division identifier for the 2007 provincial election. |\n| year | Election year for the record; all 708 records in this released extract have year=2011, with earlier-election measurements retained in separate fields. |\n| ed_id | Numeric identifier for the provincial electoral district (riding). |\n| ed_name | Name of the provincial electoral district (riding). |\n| poll_id_11 | Polling-division identifier for the 2011 provincial election. |\n| poll_id_03 | Polling-division identifier for the 2003 provincial election. |\n| master_id | Cross-year master identifier linking the same polling division across elections. |\n| area_m2_07 | Precinct land area in square meters (2007). |\n| area_m2_11 | Precinct land area in square meters (2011). |\n| area_m2_03 | Precinct land area in square meters (2003). |\n| redist | Indicator for whether the precinct/district experienced redistricting or boundary changes. |\n| km2_07 | Precinct area in square kilometers (2007). |\n| km2_11 | Precinct area in square kilometers (2011). |\n| km2_03 | Precinct area in square kilometers (2003). |\n| id_count | Count of overlapping census dissemination areas or sub-units matched to the precinct. |\n| electors | Number of registered electors (voters) in the precinct (year-specific). |\n| voter_turnout | Number of ballots cast in the precinct (absolute turnout count). |\n| perc_turnout | Turnout rate in the precinct (proportion of electors who voted). |\n| votes_lib | Number of votes received by the provincial Liberal Party candidate in the precinct. |\n| perc_lib | Share of votes for the provincial Liberal Party in the precinct (proportion). |\n| prop | Indicator (0/1) for presence of at least one proposed wind turbine within the precinct boundary. |\n| op | Indicator (0/1) for presence of at least one operational wind turbine within the precinct boundary. |\n| treat_p | Derived analysis flag indicating the precinct was coded as having a proposed turbine (derived indicator). |\n| treat_o | Derived analysis flag indicating the precinct was coded as having an operational turbine (derived indicator).|\n| op_1km | Indicator for precinct overlapping a 1 km buffer around an operational turbine. |\n| treat_o_1km | Derived analysis indicator for being within 1 km of an operational turbine. |\n| op_2km | Indicator for precinct overlapping a 2 km buffer around an operational turbine. |\n| treat_o_2km | Derived analysis indicator for being within 2 km of an operational turbine. |\n| op_3km | Indicator for precinct overlapping a 3 km buffer around an operational turbine. |\n| treat_o_3km | Derived analysis indicator for being within 3 km of an operational turbine. |\n| op_4km | Indicator for precinct overlapping a 4 km buffer around an operational turbine. |\n| treat_o_4km | Derived analysis indicator for being within 4 km of an operational turbine. |\n| op_5km | Indicator for precinct overlapping a 5 km buffer around an operational turbine. |\n| treat_o_5km | Derived analysis indicator for being within 5 km of an operational turbine. |\n| op_6km | Indicator for precinct overlapping a 6 km buffer around an operational turbine. |\n| treat_o_6km | Derived analysis indicator for being within 6 km of an operational turbine. |\n| prop_1km | Indicator for precinct overlapping a 1 km buffer around a proposed turbine. |\n| treat_p_1km | Derived analysis indicator for being within 1 km of a proposed turbine. |\n| prop_2km | Indicator for precinct overlapping a 2 km buffer around a proposed turbine. |\n| treat_p_2km | Derived analysis indicator for being within 2 km of a proposed turbine. |\n| prop_3km | Indicator for precinct overlapping a 3 km buffer around a proposed turbine. |\n| treat_p_3km | Derived analysis indicator for being within 3 km of a proposed turbine. |\n| prop_4km | Indicator for precinct overlapping a 4 km buffer around a proposed turbine. |\n| treat_p_4km | Derived analysis indicator for being within 4 km of a proposed turbine. |\n| prop_5km | Indicator for precinct overlapping a 5 km buffer around a proposed turbine. |\n| treat_p_5km | Derived analysis indicator for being within 5 km of a proposed turbine. |\n| prop_6km | Indicator for precinct overlapping a 6 km buffer around a proposed turbine. |\n| treat_p_6km | Derived analysis indicator for being within 6 km of a proposed turbine. |\n| avg_pwr | Average wind power resource value for the precinct calculated from spatial wind rasters (native units). |\n| count_pwr | Count of wind-power raster cells or points intersecting the precinct used to compute avg_pwr. |\n| pop | Estimated resident population in the precinct (from census-derived mapping). |\n| pop_denc | Population density for the precinct (people per unit area). |\n| p_uni_degree | Proportion of the adult population with a university degree. |\n| unemploy_rate | Unemployment rate (proportion) in the precinct. |\n| median_inc | Median household or individual income in the precinct (local currency). |\n| p_immigrant | Proportion of the population that are immigrants. |\n| avg_home_val | Average home value or assessed property value in the precinct (currency). |\n| p_housing_own | Proportion of housing units that are owner-occupied. |\n| log_pop | Natural logarithm of the precinct population. |\n| log_pop_denc | Natural logarithm of population density. |\n| log_median_inc | Natural logarithm of median income. |\n| avg_home_val_log | Natural logarithm of average home value (contemporaneous measure). |\n| longitude | Longitude coordinate of the precinct centroid (decimal degrees). |\n| latitude | Latitude coordinate of the precinct centroid (decimal degrees). |\n| mindistlake | Distance from the precinct centroid to the nearest Great Lake (units consistent with dataset, e.g., km). |\n| fed_id | Identifier for the corresponding federal electoral district. |\n| fed_dist | Federal electoral district name or code. |\n| votes_lib_fed | Number of votes for the federal Liberal Party candidate in the precinct (federal election). |\n| turnout_fed | Number of ballots cast in the federal election at the precinct-level. |\n| electors_fed | Number of registered electors for the federal election in the precinct. |\n| perc_lib_fed | Share of votes for the federal Liberal Party in the precinct (proportion). |\n| perc_turnout_fed | Turnout rate in the federal election (proportion). |\n| log_pop_sq | Squared term of log_pop (quadratic covariate). |\n| log_pop_denc_sq | Squared term of log_pop_denc. |\n| log_median_inc_sq | Squared term of log_median_inc. |\n| avg_home_val_log_sq | Squared term of avg_home_val_log. |\n| unemploy_rate_sq | Squared term of unemployment rate. |\n| p_uni_degree_sq | Squared term of p_uni_degree. |\n| p_immigrant_sq | Squared term of p_immigrant. |\n| p_housing_own_sq | Squared term of p_housing_own. |\n| weights_p | Retained observation weight associated with the proposed-turbine sample; equal to 1 for every row of this released extract. |\n| weights_o | Retained observation weight associated with the operational-turbine sample; equal to 1 for every row of this released extract. |\n| chng_lib | Provincial Liberal Party vote share in 2011 minus its 2007 share, expressed as a proportion; negative values denote declining support. |\n| avg_home_val_log_07 | Natural logarithm of average home value measured for 2007. |\n| avg_pwr_log | Natural logarithm of avg_pwr (logged wind power). |\n| mindistlake_log | Natural logarithm of mindistlake. |\n| mindistlake_sq | Squared distance to nearest Great Lake (quadratic term). |\n| long_sq | Squared term of longitude. |\n| lat_sq | Squared term of latitude. |\n| long_lat | Interaction term: longitude multiplied by latitude. |\n\nData Description: This released file is a selected 2011 comparison sample of 708 Ontario polling divisions: 354 with a proposed wind turbine within 3 km (prop_3km=1) and 354 without one within that distance (prop_3km=0). It is not the full multi-year provincial panel. The groups were selected for comparability on recorded pre-existing characteristics; this does not by itself establish that all determinants of turbine proposals or voting are comparable.\n\nThe source data link Ontario provincial election returns across years, corresponding federal returns, geospatial information on onshore wind projects and wind resources, and census-derived socioeconomic indicators. Turbines were geolocated from project lists and spatially joined to precinct polygons to produce presence flags, buffer indicators (1\u20136 km), counts, and average wind-power measures from Land Information Ontario rasters. Census dissemination-area data were geographically weighted into precincts to produce demographic and housing variables. Logged and squared transformations and precomputed weight fields are retained; weights_p and weights_o both equal 1 throughout this extract. chng_lib records the change in provincial Liberal vote share from 2007 to 2011.\n","file_sha256":"5e162a6286b73906441536ef428f81ba09456bb80ab810f43eaefaa7a6572a70","format":"csv","frame_sha256":null,"generator":null,"kind":"file","licence":"redistributed by CausalReasoningBenchmark (MIT) under the original replication data's terms","member":null,"object":null,"params":{},"path":null,"read_kw":{},"records":null,"role":"agent","sha256":"5e162a6286b73906441536ef428f81ba09456bb80ab810f43eaefaa7a6572a70","transforms":[],"url":"https://huggingface.co/datasets/syrgkanislab/CausalReasoningBenchmark/resolve/edd0fe1c1d1c5c0a4f0e3e67a50481ac679fb888/data/ds_0051/data/data.csv"},"ref_iv":{"columns":null,"description":"Complete rows of the script's columns, with ed_id dummies","file_sha256":"5e162a6286b73906441536ef428f81ba09456bb80ab810f43eaefaa7a6572a70","format":"csv","frame_sha256":null,"generator":null,"kind":"file","licence":"redistributed by CausalReasoningBenchmark (MIT) under the original replication data's terms","member":null,"object":null,"params":{},"path":null,"read_kw":{},"records":null,"role":"reference","sha256":"5e162a6286b73906441536ef428f81ba09456bb80ab810f43eaefaa7a6572a70","transforms":[{"args":{"columns":["chng_lib","prop_3km","avg_pwr_log","mindistlake","mindistlake_sq","longitude","long_sq","latitude","lat_sq","long_lat","ed_id"]},"op":"select"},{"args":{"columns":["chng_lib","prop_3km","avg_pwr_log","mindistlake","mindistlake_sq","longitude","long_sq","latitude","lat_sq","long_lat","ed_id"]},"op":"dropna"},{"args":{"column":"ed_id","drop_first":true,"prefix":"ed_id"},"op":"dummies"}],"url":"https://huggingface.co/datasets/syrgkanislab/CausalReasoningBenchmark/resolve/edd0fe1c1d1c5c0a4f0e3e67a50481ac679fb888/data/ds_0051/data/data.csv"}},"difficulty":"M","id":"crb_papers.q081","level":"T2","meta":{"agent_reach":{"blocked_by":["reference_frame:ref_iv","needs_op:iv_se_hc0"],"dtypes":{"ds_0051":{"chng_lib":"float64","prop_3km":"float64"}},"named":{"outcome":["chng_lib"],"treatment":["prop_3km"]},"reachable":false,"rule":"magi_bench.reach/1"},"answer_format":{"figure_transform":"identity","round_to":null,"unit_scale":1.0},"cluster":"ds_0051","crb":{"data_path":"data/ds_0051/data/data.csv","dataset_id":"ds_0051","gold":{"ci95":[-0.13094526994891065,-0.023037402860295993],"ci_rule":"effect \u00b1 1.96\u00b7|standard_error| (evaluate_result.ci_bounds)","effect":-0.07699133640460332,"standard_error":0.027527517114442512},"identification":{"bad_controls":["votes_lib","perc_lib","voter_turnout","perc_turnout","votes_lib_fed","perc_lib_fed","turnout_fed","perc_turnout_fed","op","treat_o","op_1km","treat_o_1km","op_2km","treat_o_2km","op_3km","treat_o_3km","op_4km","treat_o_4km","op_5km","treat_o_5km","op_6km","treat_o_6km","avg_home_val_log/avg_home_val"],"causal_quantity":"LATE","controls":["mindistlake/mindistlake_log","mindistlake_sq","longitude","long_sq","latitude","lat_sq","long_lat","ed_id"],"instrument":["avg_pwr_log/avg_pwr"],"is_encouragement_design":false,"minimal_controlling_set":["ed_id","mindistlake/mindistlake_log","mindistlake_sq","longitude","long_sq","latitude","lat_sq","long_lat"],"outcomes":["chng_lib"],"reason_for_bad_controls":"These variables are post-treatment or downstream of treatment and would induce over-control or collider bias if conditioned on. Specifically: (a) Electoral outcomes in 2011 (votes_lib, perc_lib, voter_turnout, perc_turnout) and their federal counterparts (votes_lib_fed, perc_lib_fed, turnout_fed, perc_turnout_fed) are outcomes/mediators affected by nearby turbine proposals; conditioning on them blocks part of the treatment effect or opens collider paths. (b) Operational turbine indicators and buffers (op, treat_o and op_1km\u2013op_6km, treat_o_1km\u2013treat_o_6km) occur after proposals and lie on causal paths from proposals to voting; controlling for them would condition on mediators. (c) avg_home_val_log is a contemporaneous housing-price measure that can be affected by turbine proposals; including it risks over-controlling for a mediator. Per the bad-controls guidance, post-treatment variables and mediators should not be included in adjustment sets.","reason_for_minimal_controlling_set":"The IV specification conditions on flexible geographic controls and district fixed effects to absorb spatial trends unrelated to politics: \u201cWind speed is theoretically orthogonal to precinct boundaries but predicts the placement of wind turbine locations\u2026 Geographic controls, in both the first and second stage, included longitude, latitude, both variables squared, and their interaction\u2026 the model also controls for distance to [the Great] lakes and this variable squared.\u201d (pp. 965\u2013966). The paper argues siting was driven by wind resources, not politics: \u201cWind project treatment assignment to communities in Ontario\u2026 was plausibly exogenous to political boundaries: Wind farms were proposed and built where wind resources were strongest, and wind resources are conditionally orthogonal to the political preferences of voters.\u201d (p. 963). District fixed effects (ed_id) are included in both stages (Table 2, p. 969).","strategy":"Instrumental Variable","treatments":["prop_3km"],"variant":"2SLS (cross-sectional) with Mahalanobis pre-matching, district fixed effects and geographic controls"},"identification_strategy":"Instrumental Variable"},"headline_keys":{"infer.iv_2sls":["coef.prop_3km"]},"naive":{"dataset":"ds_0051","treatment":"prop_3km","y":"chng_lib"},"native_id":"81","reference":{"cite":1,"note":"Reference path transcribed from the item's own estimation script.","steps":[{"args":{"dataset":"ref_iv","endog":["prop_3km"],"exog":["mindistlake","mindistlake_sq","longitude","long_sq","latitude","lat_sq","long_lat","ed_id_10.0","ed_id_14.0","ed_id_18.0","ed_id_19.0","ed_id_21.0","ed_id_22.0","ed_id_28.0","ed_id_29.0","ed_id_34.0","ed_id_36.0","ed_id_40.0","ed_id_41.0","ed_id_42.0","ed_id_55.0","ed_id_58.0","ed_id_67.0","ed_id_69.0","ed_id_70.0","ed_id_73.0","ed_id_85.0","ed_id_87.0","ed_id_90.0","ed_id_91.0","ed_id_98.0","ed_id_105.0"],"instruments":["avg_pwr_log"],"se":"HC1","y":"chng_lib"},"tool":"infer.iv_2sls","why":"IV2SLS.from_formula('chng_lib ~ controls + C(ed_id) + [prop_3km ~ avg_pwr_log]').fit(cov_type='robust')"}]},"reference_path":{"calls":[{"args":{"dataset":"ref_iv","endog":["prop_3km"],"exog":["mindistlake","mindistlake_sq","longitude","long_sq","latitude","lat_sq","long_lat","ed_id_10.0","ed_id_14.0","ed_id_18.0","ed_id_19.0","ed_id_21.0","ed_id_22.0","ed_id_28.0","ed_id_29.0","ed_id_34.0","ed_id_36.0","ed_id_40.0","ed_id_41.0","ed_id_42.0","ed_id_55.0","ed_id_58.0","ed_id_67.0","ed_id_69.0","ed_id_70.0","ed_id_73.0","ed_id_85.0","ed_id_87.0","ed_id_90.0","ed_id_91.0","ed_id_98.0","ed_id_105.0"],"instruments":["avg_pwr_log"],"se":"HC1","y":"chng_lib"},"tool":"infer.iv_2sls"}],"needs":["needs_op:iv_se_hc0"],"provenance":"benchmark_script","quotes":{},"reference_datasets":["ref_iv"],"source_id":"crb_papers.q081.script"}},"question":"What is the causal effect of a proposed wind turbine being located within 3 km of a precinct (prop_3km) on the change in the provincial Liberal Party's vote share between 2007 and 2011 (chng_lib), estimated for comparable precincts in the 2011 selected comparison sample?","requires":[],"runner":null,"sources":["crb_papers.q081.gold","crb_papers.q081.script"],"split":"public","suite":"crb_papers","tags":["econ-bench","external","crb","iv","ref_benchmark_script"],"target":"agent","tool":null,"truth":"CausalReasoningBenchmark gold at the pinned revision: causal_queries.json effect and standard_error (the value the item's estimation.py prints), scored as evaluate_result.py scores estimate_within_solution_ci"}