Replication Data and Code for "Shallow Circularity: Firm Resources, Institutional Context and the Depth of Circular Economy Practices in EU SMEs"

This dataset contains the data and Python code used in the article "Shallow Circularity: Firm Resources, Institutional Context and the Depth of Circular Economy Practices in EU SMEs". Contents: - country_panel_eu27.csv: EU27 country × year panel (2000–2025) with circularity indicators, GDP, industry share, Worldwide Governance Indicators and the coding of national circular economy strategies (data retrieved on 23 September 2026). - sme_country_wave_indicators.csv: weighted shares of SME practices, support and barriers by country and survey wave (2012–2024), computed from the survey microdata. - ce_strategy_adoption.csv: year of adoption of the first national circular economy strategy by country, with sources. - codebook_country_data.csv and codebook_sme_variables.csv: variable definitions and the crosswalk of harmonised survey variables across waves. - 01_country_data.py to 05_figures.py: Python scripts, run in numerical order. Script 01 optionally re-downloads the country data; 02 harmonises the survey microdata; 03 estimates the latent class models and measurement invariance tests; 04 estimates the class-membership models and the external validation; 05 produces the figures. Survey microdata: the firm-level data are the Flash Eurobarometer surveys "SMEs, Resource Efficiency and Green Markets" (waves 342, 381, 426, 456, 498 and 549), distributed by GESIS under studies ZA5777, ZA5874, ZA6648, ZA6917, ZA7806 and ZA8869 (DOIs listed under Related works). They are free for academic use after registration but may not be redistributed, and are therefore not included. To reproduce the analysis, download the Stata files from GESIS and place them in the same folder as the scripts. Files created by scripts 02–04 contain microdata and should not be shared. Requirements: Python 3.10 or later with pandas, numpy, scipy, matplotlib and requests. Licence: data files are released under CC BY 4.0 and code under the MIT licence (see LICENSE.txt). Please also cite Eurostat, the World Bank, the European Environment Agency and the GESIS studies used.