{"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.q071.gold (closed, evaluate_result.py; no --auto-rescale)","open_interval":false,"rel_tol":null,"required":true,"source_id":null,"table":null,"tool":null,"value":[-1.2158604039748053,8.862153380473373],"weight":1.0,"where":null},{"abs_tol":2.5709218837878005e-06,"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.q071.gold","table":null,"tool":null,"value":3.8231464882492836,"weight":1.0,"where":null},{"abs_tol":2.5709218837878005e-06,"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.q071.gold","table":null,"tool":null,"value":2.5709218837878005,"weight":1.0,"where":null}],"datasets":{"ds_0041":{"columns":null,"description":"| Variable Name | Description |\n|---------------|-------------|\n| twonames | Indicator from voter file for whether two registered people share the household name. |\n| strata | Sampling strata code used in the survey frame. |\n| treat | Randomly assigned mailing status: 1 if selected to be sent the January 22, 2008 nonpartisan letter explaining Connecticut's primary-registration rules, accompanied by a blank party-affiliation form; 0 if not sent this study mailing. This records mailing assignment, not confirmed reading or subsequent party identification. |\n| voted2006 | Voter-file indicator for whether the person voted in 2006. |\n| voted2004 | Voter-file indicator for whether the person voted in 2004. |\n| voted2002 | Voter-file indicator for whether the person voted in 2002. |\n| voted2000 | Voter-file indicator for whether the person voted in 2000. |\n| voted1998 | Voter-file indicator for whether the person voted in 1998. |\n| voted1996 | Voter-file indicator for whether the person voted in 1996. |\n| age | Respondent age in years at the time of survey. |\n| regyear | Year the person first registered to vote (from voter file). |\n| voted2008 | Administrative voter-file indicator of voting in the February 5, 2008 Connecticut presidential primary; unavailable for some respondents where town reporting was incomplete. |\n| pid7 | Baseline January 2008 party-identification scale: 1=strong Democrat, 2=weak Democrat, 3=independent leaning Democratic, 4=independent without a party lean, 5=independent leaning Republican, 6=weak Republican, 7=strong Republican. Only baseline values 3 and 5 appear in this extract. |\n| interest | Self\u2011reported interest in the election (survey item). |\n| votein2000 | Pre\u2011survey report of who the respondent voted for in the 2000 presidential race. |\n| vote2000bush | Indicator (pre\u2011survey) for reporting a vote for Bush in 2000. |\n| vote2000gore | Indicator (pre\u2011survey) for reporting a vote for Gore in 2000. |\n| econretro | Pre\u2011survey retrospective evaluation of the economy. |\n| unemployment | Pre\u2011survey view of unemployment / job market performance. |\n| bushapp | Pre\u2011survey approval rating for President Bush. |\n| congapp | Pre\u2011survey approval rating for Congress. |\n| pt_gotmail | Post\u2011survey indicator reporting whether the respondent received campaign mail. |\n| pt_gotphone | Post\u2011survey indicator reporting whether the respondent received campaign phone contact. |\n| pt_gotface | Post\u2011survey indicator reporting whether the respondent experienced face\u2011to\u2011face campaign contact. |\n| pt_gotmoney | Post\u2011survey indicator reporting whether the respondent received a fundraising request. |\n| pt_gotad | Post\u2011survey indicator reporting whether the respondent saw campaign advertising. |\n| pt_persuadeothers | Post\u2011survey report of attempts to persuade others about politics. |\n| pt_soughtinfo | Post\u2011survey report of seeking political information. |\n| pt_interest | Post\u2011survey reported interest in upcoming elections. |\n| pt_pid7 | June 2008 follow-up party-identification response, using the same 1\u20137 coding as pid7. |\n| pt_votein2000 | Post\u2011survey report of who the respondent said they voted for in 2000. |\n| pt_vote2000bush | Post\u2011survey indicator for reporting a vote for Bush in 2000. |\n| pt_vote2000gore | Post\u2011survey indicator for reporting a vote for Gore in 2000. |\n| pt_novvote | Post\u2011survey reported intended vote for the November election. |\n| pt_stayiniraq | Post\u2011survey attitude about whether to keep troops in Iraq. |\n| pt_taxrich | Post\u2011survey attitude about taxation of high\u2011income individuals. |\n| pt_evalreagan | Post\u2011survey evaluative rating of Ronald Reagan. |\n| pt_evalcarter | Post\u2011survey evaluative rating of Jimmy Carter. |\n| pt_evaldem | Post\u2011survey evaluative rating of the Democratic Party. |\n| pt_evalrep | Post\u2011survey evaluative rating of the Republican Party. |\n| pt_econretro | Post\u2011survey retrospective evaluation of the economy. |\n| pt_postunemployperf | Post\u2011survey perception of unemployment performance. |\n| pt_bushapp | Post\u2011survey approval rating for President Bush. |\n| pt_congapp | Post\u2011survey approval rating for Congress. |\n| pt_campaigndonation | Post\u2011survey indicator for whether the respondent donated to a campaign. |\n| pt_educ | Respondent education level as recorded in the postsurvey dataset. |\n| pt_black | Post\u2011survey indicator for self\u2011reported Black race. |\n| pt_hispanic | Post\u2011survey indicator for self\u2011reported Hispanic ethnicity. |\n| pt_church | Post\u2011survey measure related to church attendance or religiosity. |\n| pt_yob | Post\u2011survey year of birth field. |\n| pt_female | Post\u2011survey female indicator. |\n| townshowsnoturnoutinvoterfile | Town\u2011level flag indicating absence of turnout reporting in the voter file for certain towns. |\n| R_persondead | Administrative flag indicating the voter\u2011file record shows the person as deceased. |\n| R_movedpre | Administrative flag for a move detected prior to the study period. |\n| R_movedpost | Administrative flag for a move detected after the intervention period. |\n| R_badmailadddress | Administrative flag for a problematic mailing address. |\n| R_anybad | Any bad\u2011record flag summarizing address/person problems. |\n| R_STRATACOAFILTER | Administrative/sample filter flag used in constructing the analytic sample. |\n| SD_2 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_3 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_4 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_5 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_6 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_7 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_8 | Sample/design indicator (one of several strata/dummy indicators). |\n| SD_9 | Sample/design indicator (one of several strata/dummy indicators). |\n| SDA_1 | Auxiliary sample/design indicator used in weighting/stratification. |\n| SDA_2 | Auxiliary sample/design indicator used in weighting/stratification. |\n| SDA_3 | Auxiliary sample/design indicator used in weighting/stratification. |\n| SDA_5 | Auxiliary sample/design indicator used in weighting/stratification. |\n| SDA_6 | Auxiliary sample/design indicator used in weighting/stratification. |\n| SDA_7 | Auxiliary sample/design indicator used in weighting/stratification. |\n| stratactr | Numeric stratification or centering variable for sample design. |\n| altstrata | Alternative strata identifier used in some analyses. |\n| SI_2 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_3 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_4 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_5 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_6 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_7 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_8 | Additional sample/indicator variable (survey instrumentation or strata). |\n| SI_9 | Additional sample/indicator variable (survey instrumentation or strata). |\n| STI_1 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_2 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_3 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_4 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_5 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_6 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_7 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_8 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| STI_9 | Set of indicator variables used for survey/strata adjustments (STI series). |\n| post_unif | Post\u2011survey indicator related to unaffiliated registration status. |\n| post_affil | Post\u2011survey indicator for any party affiliation in administrative records. |\n| post_dem | Post\u2011survey indicator for Democratic registration in administrative records. |\n| post_rep | Post\u2011survey indicator for Republican registration in administrative records. |\n| post_DUR | Administrative/post variable capturing duration or timing of post\u2011registration changes. |\n| vffemale | Gender as recorded in the voter file (female indicator). |\n| vfgendermissing | Flag for missing gender in the voter file. |\n| age2 | Age squared (for use in models). |\n| twophones | Indicator for whether two phone numbers were associated with the record. |\n| regyearmissing | Indicator that the registration\u2011year field was missing. |\n| combined_female | Combined female indicator (harmonized across survey/voter file). |\n| pre_direct_bushap | Pre\u2011survey direct question on approval of President Bush. |\n| pt_direct_bushap | Post\u2011survey direct question on approval of President Bush. |\n| pre_direct_congapp | Pre\u2011survey direct question on approval of Congress. |\n| pt_direct_congapp | Post\u2011survey direct question on approval of Congress. |\n| pt_direct_bin_novvote | Post\u2011survey binary indicator of intended November vote (binary encoding). |\n| pre_aligned_vh | Pre\u2011survey measure of alignment between partisan leanings and vote/party views (baseline index). |\n| post_aligned_vh | Post\u2011survey measure of alignment between partisan leanings and vote/party views (follow\u2011up index). |\n| pre_direct_econ | Pre\u2011survey direct evaluation of the economy. |\n| pre_direct_unemp | Pre\u2011survey direct evaluation of unemployment. |\n| post_direct_econ | Post\u2011survey direct evaluation of the economy. |\n| post_direct_unemp | Post\u2011survey direct evaluation of unemployment. |\n| post_aligned_iraq | Post\u2011survey alignment index for Iraq policy position relative to partisan leaning. |\n| post_aligned_taxrich | Post\u2011survey alignment index for taxation policy relative to partisan leaning. |\n| post_aligned_reagan | Post\u2011survey alignment index for Reagan evaluation relative to partisan leaning. |\n| post_aligned_carter | Post\u2011survey alignment index for Carter evaluation relative to partisan leaning. |\n| post_aligned_rep | Post\u2011survey alignment index for Republican party evaluations. |\n| post_aligned_dem | Post\u2011survey alignment index for Democratic party evaluations. |\n| pre_lean_dem | Baseline leaning indicator: 1 for Democratic leaners (pid7=3), 0 for Republican leaners (pid7=5) in this extract. |\n| pt_voteevalalignindex | Additive follow-up alignment score combining reported 2000 presidential vote choice, intended November 2008 vote, and evaluations of the Democratic and Republican parties, oriented toward the respondent's baseline party lean. Larger values indicate greater alignment; this is an index score, not a probability. |\n| abspid7 | Absolute value or intensity measure derived from the seven\u2011point party identification scale. |\n| pt_id_with_lean | Follow-up self-identification with the party leaned toward at baseline: 1 for a baseline Democratic leaner reporting weak or strong Democratic identification in June, or a baseline Republican leaner reporting weak or strong Republican identification; 0 otherwise. This is a survey measure, distinct from administrative party registration. |\n| pt_direct_pid7 | Post\u2011survey direct 7\u2011point party identification item. |\n| post_reg_lean | Post\u2011registration indicator of party lean in administrative records. |\n| pt_voteevalalignindexwithcb | Variant of the vote/evaluation alignment index that includes additional control adjustments. |\n| pt_voteevalalignindexwitcbhist | Historical/alternative variant of the vote/evaluation alignment index. |\n| pt_policyevalalignindex | Composite index measuring alignment between policy positions and partisan leanings (post\u2011survey). |\n\nData Description: This person-level extract links telephone interviews with Connecticut administrative voter-file records around the 2008 presidential primary. The sampling frame consisted of registered voters who were not formally affiliated with a party. The study targeted ages 18\u201349 and recent voters, with an additional eligibility route for younger recent registrants, and screened contact records for telephone and mailing availability. The baseline survey was conducted January 11\u201316, 2008. This released analytic extract contains 411 respondents who initially described themselves as independents leaning toward either the Democratic or Republican Party: 265 Democratic leaners and 146 Republican leaners.\n\nA randomly selected half of the experimental participants were sent a nonpartisan letter on January 22, 2008, with a blank party-affiliation form. The letter explained that participation in a party's February 5 presidential primary required registration with that party. The remaining participants were not sent this study mailing. Mailing assignment is recorded by treat; becoming a party identifier or changing administrative registration was not itself assigned.\n\nA follow-up telephone survey in June 2008 measured party identification, intended November voting, recalled earlier voting, party evaluations, policy attitudes, and campaign contacts. A voter-file update supplied registration changes and primary turnout. The file therefore distinguishes mailing assignment, subsequent self-reported party identity, administrative party registration, and follow-up political attitudes. Respondents were not all reinterviewed in the original study; the released analytic subset should not be interpreted as the entire originally sampled population.\n","file_sha256":"567d4130760d86a5a30d59ec7df9f8ca003ac24411fdeb6ae3e2c1f1538868ac","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":"567d4130760d86a5a30d59ec7df9f8ca003ac24411fdeb6ae3e2c1f1538868ac","transforms":[],"url":"https://huggingface.co/datasets/syrgkanislab/CausalReasoningBenchmark/resolve/edd0fe1c1d1c5c0a4f0e3e67a50481ac679fb888/data/ds_0041/data/data.csv"},"ref_iv":{"columns":null,"description":"Respondents with R_STRATACOAFILTER == 0 and pid7 in (3, 5), complete rows","file_sha256":"567d4130760d86a5a30d59ec7df9f8ca003ac24411fdeb6ae3e2c1f1538868ac","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":"567d4130760d86a5a30d59ec7df9f8ca003ac24411fdeb6ae3e2c1f1538868ac","transforms":[{"args":{"query":"R_STRATACOAFILTER == 0 and pid7 in [3, 5]"},"op":"filter"},{"args":{"columns":["pt_voteevalalignindex","pt_id_with_lean","treat","pre_lean_dem","age","age2","regyear","regyearmissing","twonames","combined_female","voted2006","voted2004","voted2002","voted2000","voted1998","voted1996","interest","pre_aligned_vh","pre_direct_unemp","pre_direct_econ","pre_direct_bushap","pre_direct_congapp","R_STRATACOAFILTER","pid7"]},"op":"dropna"},{"args":{"columns":["pt_voteevalalignindex","pt_id_with_lean","treat","pre_lean_dem","age","age2","regyear","regyearmissing","twonames","combined_female","voted2006","voted2004","voted2002","voted2000","voted1998","voted1996","interest","pre_aligned_vh","pre_direct_unemp","pre_direct_econ","pre_direct_bushap","pre_direct_congapp"]},"op":"select"}],"url":"https://huggingface.co/datasets/syrgkanislab/CausalReasoningBenchmark/resolve/edd0fe1c1d1c5c0a4f0e3e67a50481ac679fb888/data/ds_0041/data/data.csv"}},"difficulty":"M","id":"crb_papers.q071","level":"T2","meta":{"agent_reach":{"blocked_by":["reference_frame:ref_iv","needs_op:iv_se_hc0"],"dtypes":{"ds_0041":{"pt_id_with_lean":"float64","pt_voteevalalignindex":"float64"}},"named":{"outcome":["pt_voteevalalignindex"],"treatment":["pt_id_with_lean"]},"reachable":false,"rule":"magi_bench.reach/1"},"answer_format":{"figure_transform":"identity","round_to":null,"unit_scale":1.0},"cluster":"ds_0041","crb":{"data_path":"data/ds_0041/data/data.csv","dataset_id":"ds_0041","gold":{"ci95":[-1.2158604039748053,8.862153380473373],"ci_rule":"effect \u00b1 1.96\u00b7|standard_error| (evaluate_result.ci_bounds)","effect":3.8231464882492836,"standard_error":2.5709218837878005},"identification":{"bad_controls":["voted2008","pt_gotmail","pt_gotphone","pt_gotface","pt_gotmoney","pt_gotad","pt_persuadeothers","pt_soughtinfo","pt_interest","pt_pid7","pt_votein2000","pt_vote2000bush","pt_vote2000gore","pt_novvote","pt_stayiniraq","pt_taxrich","pt_evalreagan","pt_evalcarter","pt_evaldem","pt_evalrep","pt_econretro","pt_postunemployperf","pt_bushapp","pt_congapp","pt_campaigndonation","pt_direct_bushap","pt_direct_congapp","pt_direct_bin_novvote","pt_direct_pid7","post_unif","post_affil","post_dem","post_rep","post_DUR","post_direct_econ","post_direct_unemp","post_aligned_iraq","post_aligned_taxrich","post_aligned_reagan","post_aligned_carter","post_aligned_rep","post_aligned_dem","post_reg_lean","pt_voteevalalignindexwithcb","pt_voteevalalignindexwitcbhist","pt_policyevalalignindex","R_movedpost"],"causal_quantity":"LATE","controls":["pre_lean_dem","age/age2","regyear","regyearmissing","twonames","combined_female","voted2006","voted2004","voted2002","voted2000","voted1998","voted1996","interest","pre_aligned_vh","pre_direct_unemp","pre_direct_econ","pre_direct_bushap","pre_direct_congapp"],"instrument":["treat"],"is_encouragement_design":true,"minimal_controlling_set":["pre_lean_dem","age/age2","regyear","regyearmissing","twonames","combined_female","voted2006","voted2004","voted2002","voted2000","voted1998","voted1996","interest","pre_aligned_vh","pre_direct_unemp","pre_direct_econ","pre_direct_bushap","pre_direct_congapp"],"outcomes":["pt_voteevalalignindex"],"reason_for_bad_controls":"All listed variables are measured after the randomized mailing (instrument) and can lie on, or be descendants of, the causal path from encouragement -> party identification -> political attitudes/behaviors. Per the good/bad controls guidance, conditioning on post-treatment variables (mediators or their descendants) blocks part of the causal effect or introduces collider/selection bias. Examples: post_affil/post_dem/post_rep directly capture post-registration outcomes that the mail explicitly shifts; pt_* campaign contacts and attitudes (e.g., pt_gotmail, pt_interest, pt_bushapp) are downstream of treatment and party ID; pt_votein2000 (post report) is shown in the paper to shift with treatment (partisan projection), so controlling for it would block part of the causal channel; voted2008 and R_movedpost occur after intervention. Therefore these are bad controls for IV estimation of the LATE of partisanship on the alignment index.","reason_for_minimal_controlling_set":"\u201cM are pretreatment covariates included to increase efficiency and correct for any imbalance on observed measures that exists after randomization. Note that in this specification we also measure other covariates (M) prior to the random assignment.\u201d (Research Design section, p. 725).","strategy":"Instrumental Variable","treatments":["pt_id_with_lean"],"variant":"encouragement design"},"identification_strategy":"Instrumental Variable"},"headline_keys":{"infer.iv_2sls":["coef.pt_id_with_lean"]},"naive":{"dataset":"ds_0041","treatment":"pt_id_with_lean","y":"pt_voteevalalignindex"},"native_id":"71","reference":{"cite":1,"note":"Reference path transcribed from the item's own estimation script.","steps":[{"args":{"dataset":"ref_iv","endog":["pt_id_with_lean"],"exog":["pre_lean_dem","age","age2","regyear","regyearmissing","twonames","combined_female","voted2006","voted2004","voted2002","voted2000","voted1998","voted1996","interest","pre_aligned_vh","pre_direct_unemp","pre_direct_econ","pre_direct_bushap","pre_direct_congapp"],"instruments":["treat"],"se":"HC1","y":"pt_voteevalalignindex"},"tool":"infer.iv_2sls","why":"2SLS by matrices: X = [1, D, controls], Z = [1, treat, controls]"}]},"reference_path":{"calls":[{"args":{"dataset":"ref_iv","endog":["pt_id_with_lean"],"exog":["pre_lean_dem","age","age2","regyear","regyearmissing","twonames","combined_female","voted2006","voted2004","voted2002","voted2000","voted1998","voted1996","interest","pre_aligned_vh","pre_direct_unemp","pre_direct_econ","pre_direct_bushap","pre_direct_congapp"],"instruments":["treat"],"se":"HC1","y":"pt_voteevalalignindex"},"tool":"infer.iv_2sls"}],"needs":["needs_op:iv_se_hc0"],"provenance":"benchmark_script","quotes":{},"reference_datasets":["ref_iv"],"source_id":"crb_papers.q071.script"}},"question":"Among registered Connecticut respondents who initially leaned toward a party but were unaffiliated, what is the causal effect of subsequently identifying with that party (i.e., post-survey self-reported party identification) on the alignment between their voting intentions and evaluations of parties and candidates (the Voting and Party Alignment index)? Report the local average treatment effect for the subgroup whose identification was changed by the study's intervention.","requires":[],"runner":null,"sources":["crb_papers.q071.gold","crb_papers.q071.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"}