2026-05-17T12:10:20Z https://datacatalogue.cessda.eu/oai-pmh/v0/oai

021428bf9c88819712c7554fb35a8168a7874c96d9497c1863df8f6b3f57813a 2026-05-04T02:59:59Z language:enopenaire_data

DDI study level documentation for study 10.7802/3038 ClimateGen ClimateGen ClimateGen 10.7802/303810.7802/303810.7802/303810.7802/3038 Rittershaus, Simon Rittershaus, Simon [Funded by the Deutsche Forschungsgemeinschaft (DFG – German Research Foundation) under Germany‘s Excellence Strategy – EXC-2035/1 – 390681379] GESIS Data Archive for the Social SciencesGESIS Datenarchiv für Sozialwissenschaften policy designdistributive politicspolicy feedbackpolicy instrumentscross-national panel datagenerative ailarge language models (LLM)automated text analysisclimate inequalityclimate policycomparative political sciencepolitical economypolicy designdistributive politicspolicy feedbackpolicy instrumentscross-national panel datagenerative ailarge language models (LLM)automated text analysisclimate inequalityclimate policycomparative political sciencepolitical economy Climate change mitigation requires policies that are both effective and politically feasible, making policy design a crucial focus of comparative research. However, systematic cross-national data on design features have been scarce due to the resource-intensive nature of manual coding. This dataset, ClimateGen, addresses this gap by providing a comprehensive, harmonized, event-based classification of climate policy outputs. Building on the International Energy Agency (IEA) ‘Policies and Measures’ database, ClimateGen covers over 3,400 climate policy expansions across 23 affluent democracies from 1990 to 2022. To process the extensive unstructured textual data, the dataset leverages an innovative, reproducible pipeline powered by generative artificial intelligence. Specifically, it employs the Large Language Model GPT-4o, utilizing an iteratively refined prompt architecture rigorously validated against human expert coding. By employing multiple classification trials and leveraging the model’s parametric knowledge and emergent reasoning capabilities, the pipeline systematically infers latent distributive effects from sparse textual policy records. The core contribution of ClimateGen is its systematic categorization of climate policies along three theoretically grounded design dimensions: (1) Instrument choice, distinguishing between fiscal support and regulatory measures; (2) Target groups, separating consumer-oriented policies (highly salient to households) from broader economy-oriented measures; and (3) Distributive incidence, identifying whether consumer policies allocate costs and benefits in a progressive or regressive manner across income strata. By quantifying these micro-level design choices, ClimateGen provides a unique empirical yardstick. It enables researchers to benchmark national climate policy portfolios, examine the trade-offs between political feasibility and equity, and empirically test theories of distributive politics, policy feedback, policy expansion, and climate policy sequencing.Climate change mitigation requires policies that are both effective and politically feasible, making policy design a crucial focus of comparative research. However, systematic cross-national data on design features have been scarce due to the resource-intensive nature of manual coding. This dataset, ClimateGen, addresses this gap by providing a comprehensive, harmonized, event-based classification of climate policy outputs. Building on the International Energy Agency (IEA) ‘Policies and Measures’ database, ClimateGen covers over 3,400 climate policy expansions across 23 affluent democracies from 1990 to 2022. To process the extensive unstructured textual data, the dataset leverages an innovative, reproducible pipeline powered by generative artificial intelligence. Specifically, it employs the Large Language Model GPT-4o, utilizing an iteratively refined prompt architecture rigorously validated against human expert coding. By employing multiple classification trials and leveraging the model’s parametric knowledge and emergent reasoning capabilities, the pipeline systematically infers latent distributive effects from sparse textual policy records. The core contribution of ClimateGen is its systematic categorization of climate policies along three theoretically grounded design dimensions: (1) Instrument choice, distinguishing between fiscal support and regulatory measures; (2) Target groups, separating consumer-oriented policies (highly salient to households) from broader economy-oriented measures; and (3) Distributive incidence, identifying whether consumer policies allocate costs and benefits in a progressive or regressive manner across income strata. By quantifying these micro-level design choices, ClimateGen provides a unique empirical yardstick. It enables researchers to benchmark national climate policy portfolios, examine the trade-offs between political feasibility and equity, and empirically test theories of distributive politics, policy feedback, policy expansion, and climate policy sequencing. JapanJapanPortugalPortugalBelgiumBelgiumAustraliaAustraliaAustriaAustriaGermanyGermanyFranceFranceSwitzerlandSwitzerlandUnited States of AmericaUnited States of AmericaSwedenSwedenUnited Kingdom of Great Britain and Northern IrelandUnited Kingdom of Great Britain and Northern IrelandDenmarkDenmarkNew ZealandNew ZealandCanadaCanadaLuxembourgLuxembourgFinlandFinlandGreeceGreeceItalyItalyIrelandIrelandNetherlandsNetherlandsNorwayNorwaySpainSpainKorea (Republic of)Korea (Republic of) National-level climate policy adoptions (specifically fiscal support measures and regulations) in 23 affluent democracies between 1990 and 2022. 1990-2022; annual; Longitudinal (panel study)1990-2022; annual; Longitudinal (panel study) Total Universe / Complete enumerationTotal Universe / Complete enumeration ObservationObservationBeobachtungObservation Free access (with registration) - The research data can be downloaded by registered users.Freier Zugang (mit Registrierung) - Die Forschungsdaten können von allen registrierten Nutzerinnen und Nutzern heruntergeladen werden. http://dbkapps.gesis.org/dbkoai/oai.asp oai:dbk.gesis.org:SDN/10.7802_3038 2026-04-29 ddi:codebook:2_5