Promised Abundance, Deflating Stakes: European Public Opinion on Robots, AI and Jobs, 2017–2024

## Source data (not redistributed)The respondent-level microdata are subject to GESIS terms of use and must be obtained directly from GESIS:- Eurobarometer 87.1 (March 2017), ZA6861, version 2.0.0, doi:10.4232/1.13738- Eurobarometer 101.4 (April–May 2024), ZA8844, version 1.0.0, doi:10.4232/1.14471Save them as `data/ZA6861.sav` and `data/ZA8844.sav`.## RunningPython 3.11+ with pandas, numpy, statsmodels, pyreadstat, matplotlib. mkdir out python code/01_harmonise.py data out # EU27 harmonised two-wave file (UK 2017 removed; weights w87 / w92; DK codes kept) python code/02_trends.py out # Table 3, typology (Figure 2), country and regional tables python code/03_models.py out # robustness (no-CAVI), predictor models (Table 4), 2024 welfare and rules (Table 5) python code/04_validation.py out # full-base validation against Special Eurobarometer 554; steal-based typology python code/05_figures.py out # Figures 1–3## Contents- `results/T1_EU27_item_trends.csv` – weighted item shares, change and country fixed-effects odds ratios- `results/T2_typology_EU27.csv` – four CAF publics by wave- `results/T3_country_change.csv` – country estimates (national weight)- `results/T4_region.csv` – regional estimates- `results/T5_2024_welfare_rules_by_type.csv` – Redistribution Premise and rules by CAF public (2024)- `results/T_corr_2017.csv`, `T_corr_2024.csv` – Spearman correlations among the four items- `results/res_*.json` – all model estimates reported in the article- `figures/` – Figures 1–3 (300 dpi)Licence: CC BY 4.0 (code and derived aggregates).