{"part":"task","text":"# Datasets\nPass a dataset's id as the `dataset` argument.\n{\n \"id\": \"ds_0013\",\n \"rows\": 980,\n \"description\": \"| Variable Name | Description |\\n| --- | --- |\\n| voted | State-year turnout rate (0\u20131), aggregated from the Census Bureau\u2019s Current Population Survey Voting and Registration Supplement. Computed as the share of adult residents surveyed in that state and year who report having voted in the November general election. Includes both presidential and midterm cycles and reflects CPS\u2019s biennial November fielding and weighting. |\\n| temp_var | Indicator (0/1) that residents could register and cast a ballot on the same day in that state and year. Status reflects the first general election in which the option was implemented, and remains 1 thereafter. Codings draw on legislative and administrative sources and treat Election Day registration as part of the same \u201csame-day\u201d category. |\\n| state | U.S. state name corresponding to each panel unit. |\\n| year | Calendar election year (even-numbered), matching the CPS Voter Supplement cycle and covering both presidential and midterm general elections. |\\n\\nDataset Context:\\nThis file is a state-by-year panel spanning 1980\u20132018 that aligns survey-based measures of participation with the timing of state registration rules. Turnout is derived from the Census Bureau\u2019s Current Population Survey (CPS) Voting and Registration Supplement, a large, nationally representative survey of the civilian, noninstitutionalized population conducted every two years in November. The CPS asks respondents whether they voted in the most recent general election and provides sampling weights to recover population-representative estimates. Aggregating these responses to the state-year level yields a consistent series of turnout rates across both presidential and midterm elections. As with any self-reported voting measure, the CPS can modestly overstate participation, but it is widely used for its coverage, consistent questionnaire, and long time span.\\n\\nPolicy status for same-day registration is compiled from legislative and administrative references, including nonpartisan compilations and state records, and is coded to reflect the first election in which residents could register and vote on the same day. Jurisdictions that permit registration on Election Day itself are included under the same umbrella as those allowing same-day registration during an early voting window. Once a state implements the option, its indicator remains active in subsequent even-year elections unless the policy is rescinded.\\n\\nBy harmonizing CPS-based turnout with the timing of registration options, the dataset supports comparisons of participation patterns across places and over time. The panel includes both high-salience presidential contests and midterm cycles, allowing researchers to examine differences that can arise across election types while working with a uniform measurement framework.\",\n \"columns\": {\n  \"voted\": \"float64\",\n  \"temp_var\": \"float64\",\n  \"state\": \"str\",\n  \"year\": \"int64\"\n }\n}\n{\n \"id\": \"ref_panel\",\n \"rows\": 92,\n \"description\": \"State rows of 1992 and 1994: temp_var 0 then 1 (treated) or 0 and 0 (control)\",\n \"columns\": {\n  \"voted\": \"float64\",\n  \"temp_var\": \"float64\",\n  \"state\": \"str\",\n  \"year\": \"int64\",\n  \"nyears\": \"int64\",\n  \"t_pre\": \"float64\",\n  \"t_post\": \"float64\",\n  \"treated\": \"int64\",\n  \"post\": \"int64\",\n  \"treated_post\": \"int64\"\n }\n}\n{\n \"id\": \"ref_post\",\n \"rows\": 49,\n \"description\": \"Each state's max of temp_var in 1994 (t_post)\",\n \"columns\": {\n  \"state\": \"str\",\n  \"t_post\": \"float64\"\n }\n}\n{\n \"id\": \"ref_pre\",\n \"rows\": 49,\n \"description\": \"Each state's max of temp_var in 1992 (t_pre)\",\n \"columns\": {\n  \"state\": \"str\",\n  \"t_pre\": \"float64\"\n }\n}\n{\n \"id\": \"ref_years\",\n \"rows\": 49,\n \"description\": \"How many of 1992 and 1994 each state is observed in\",\n \"columns\": {\n  \"state\": \"str\",\n  \"nyears\": \"int64\"\n }\n}"}