Climate Policy Design Dataset

Domestic Consumer Climate Subsidies & Regulations (1990–2022)

Work in progress.
This repository documents and (gradually) packages the code and data workflow behind a climate policy design dataset introduced in Chapter 3 of my dissertation.

---

Overview

This repository accompanies a novel large‑N dataset of domestic consumer climate subsidies and regulations (DCCSR) based on entries from the IEA Policies and Measures Database.

At a glance

• Coverage: 23 advanced democracies
• Time: 1990–2022
• Unit of observation: national policy adoption events (IEA policy records)
• Content focus: fiscal support policies & regulations
• Dataset size (final dataset reported in dissertation): 2,190 fiscal support policies + 1,214 regulations
• Key design dimensions (examples): consumer orientation, target groups, income-group targeting (see “Data dictionary”)

Note: Some parts of this repository are still being cleaned up for public release (paths, secrets, and documentation).

---

Repository structure

The repository currently contains three main directories:

.
├── classification/            # AI-assisted classification + post-validation workflows
├── processing/                # data cleaning, reshaping, and dataset compilation
└── descriptive replication/   # scripts/notebooks for descriptive analyses & figures

Data dictionary (lite)

A full codebook is being prepared. For now, below are the main variable groups documented in the dissertation appendix:

• Policy orientation & targeting
• Consumer_Oriented
• Target_Group
• Income_Group, Income_Confidence
• Ownership_Requirement
• Investment_Costs_and_Financial_Burden
• Policy type & sector mapping
• Policy_Instrument, Sub_Category
• Sector_Classification, sector & instrument counts
• Metadata
• Policy_ID, Country, Year, Description

⚠️ Data availability release note:

The dataset is currently under embargo until the dissertation is formally published. Once released, this folder will contain:

• data/raw/ (snapshots of source exports where licensing permits)
• data/processed/ (intermediate files)
• data/final/ (DCCSR release files)

Until then, this repository focuses on code structure, documentation, and reproducible processing steps.

---

Reproducibility / Quickstart (minimal)

This repo is currently organized as R Markdown notebooks (plus Python via reticulate where applicable).

General expectations

• Use repo-relative paths (no local absolute paths).
• No secrets are stored in the repository.

Environment variables (if needed)

• OPENAI_API_KEY should be provided via your local environment (e.g., .Renviron, shell env vars, or keychain/keyring set interactively).
• The repository code should only read secrets (never write them).

---

Changelog

See CHANGELOG.md.

---

Citation

A citation entry will be added once the dissertation/paper is public.
For now, please cite the dissertation chapter introducing the dataset.

---

License

License to be confirmed (TBD).