European Housing Policy Inventory (EHPI)

Dataset Description

The European Housing Policy Inventory (EHPI) is a harmonised dataset of housing policy indicators covering 30 European countries across six policy domains: rent control, tenure security, social housing, construction subsidies, housing allowances and tax treatment. It comprises 16 indicators, with social housing and rent control coded at subnational levels where meaningful spatial variation could be identified, and other domains coded nationally.

Methodology

Housing policy indicators were developed using literature-informed definitions and standardised coding procedures. Targeted LLM-assisted searches for ChatGPT or Claude were used to identify potentially relevant sources across the rent control, social housing, tenure security and tax treatment policy domains, prioritising official legislation and government guidance, followed by academic literature and non-governmental organisation reports, with property and legal guidance sources used where higher-priority sources were unavailable or insufficient. Identified sources were then reviewed manually, supplemented through targeted desktop research where necessary, and translated into the predefined coding rules. The dataset underwent source tracing, bilateral coding and reconciliation, independent recoding, and external validation against comparable datasets. Coding was performed on the EU-27 countries plus the United Kingdom, Switzerland and Norway, as these countries are also frequently classified in European housing system analyses.

The subsequent indicators were then harmonised across the EU NUTS0–LAU spatial hierarchy using downward and dwelling-weighted upward imputation where required. The coding process was undertaken iteratively by a team of researchers over nine months to identify, interpret and harmonise policy information across 30 countries and multiple policy domains.

Framework

Housing policy indicators were selected based on their prevalence in comparative literature, potential for subnational differentiation, and scalability across the EU. Measurement approaches sought to maximise conceptual precision while retaining scalability, relying primarily on categorical, legislation-based classifications and, where feasible, continuous census-based indicators.

The resulting EHPI indicators are:

• Rent control: initial rent control (categorical; 0–2) and rent increase control (categorical; 0–2);
• Tenure security: minimum tenancy length (categorical; 0–2);
• Social housing: social housing dwelling stock (%), social housing providers (categorical; public / not-for-profit / for-profit / other non-state), and allocation criteria (categorical; universalistic / generalist / residual);
• Construction subsidies: expenditure as a share of total government expenditure (%), expenditure as a share of total government housing subsidy expenditure (%), per-capita expenditure adjusted for inflation and purchasing power (continuous), and Z-stat of adjusted per-capita expenditure using EU-27 expenditure figures (continuous);
• Housing allowances: expenditure as a share of total government expenditure (%), expenditure as a share of total government housing subsidy expenditure (%), per-capita expenditure adjusted for inflation and purchasing power (continuous), and Z-stat of adjusted per-capita expenditure using EU-27 expenditure figures (continuous);
• Tax treatment: mortgage interest deduction (MID) (binary) and dual-income tax (DIT) (binary).

The targeted rationale, operational definitions, and measurement approach for all indicators, and a comprehensive description of the entire methodology is detailed in the attached Data Descriptors paper.

Purpose / Usage Note

The EHPI dataset is designed to support comparative housing policy research at the subnational scale and analysis of interactions between housing policy domains. It is not intended as a comprehensive municipality-level policy glossary, but as a validated macro-level dataset that provides a consistent starting point for quantitative comparative research. Several limitations should therefore be considered when interpreting individual spatial-unit observations.

Data Format

The repository provides separate CSV and GeoJSON files for each EU spatial nomenclature level (NUTS0–NUTS3 and LAU). The datasets contain 30 NUTS0 regions, 106 NUTS1 regions, 304 NUTS2 regions, 1,390 NUTS3 regions, and 97,943 LAUs. Each row corresponds to a single spatial unit and contains the EHPI policy indicators, intermediate classification variables, applicable legislation year, and spatial-unit source and provenance information. The policy datasets are provided as at present (2026), and as of 2024 to align with existing European subnational housing outcome datasets. The CSV and GEOJSON files are structured using standard EU spatial nomenclature codes, facilitating interoperability with other European subnational datasets. These files are labelled “EHPI_lau_26.csv”, “EHPI_nuts3_26.csv”, …, “EHPI_nuts0_26.csv”. The GeoJSON files additionally contain the corresponding spatial boundaries, enabling direct use in GIS and spatial analyses. These files are labelled “EHPI_lau_26_sf.geojson”, “EHPI_nuts3_26_sf.geojson”, …, “EHPI_nuts0_26_sf.geojson”.

Metadata describing the variables and coding conventions are provided in the “EHPI metadata” file. Supplementary qualitative coding documentation is provided for the rent control (“EHPI_RC_qual”), social housing (“EHPI_SH_qual”), tenure security (“EHPI_TS_qual”) and tax treatment (“EHPI_TT_qual”) domains. These files document the underlying classification process and provide additional legislative and policy summaries, source information, and classification tags that are not fully represented in the quantitative datasets. The “EHPI_validation” file documents the internal and external validation procedures and results.