National Road Mortality Trajectories, 1996-2021: The Independent Roles of Governance Quality and Cultural Values
Description
This dataset supports the manuscript "National Road Mortality Trajectories, 1996–2021: The Independent Roles of Governance Quality and Cultural Values" (LaJeunesse, submitted to Accident Analysis & Prevention). It is a country-year panel covering 74 countries from 1996 to 2021 (1,924 observations), combining road traffic mortality rates, governance quality indicators, GDP per capita, population-weighted density, and Schwartz cultural value orientation scores. The dataset was constructed by merging four public sources at the country and country-year level; no primary data collection was involved.
Files
Steps to reproduce
Run `primary_analysis_rebuild.py` in the same directory as `Panel_Master_5-4-2026_deposit.xlsx` (the script reads `Panel_Master_5-4-2026.xlsx` by filename; rename the deposit copy or update the path in the script if running it directly). It reproduces: - M0 (unconditional means model; ICC) - M1 (unconditional growth model; grand mean improvement rate) - G1 (primary GEE-Stationary model) -- Table 2 - S1, S2 (GEE-Stationary sensitivity specifications) -- Table 4 - D2 (lagged-governance REML MLGCM specification) -- Table 3 Environment used: Python 3.12.3, pandas 3.0.2, numpy 2.4.4, statsmodels 0.14.6. Full output is in `analysis_output_log.txt`.
Institutions
- University of North Carolina at Chapel HillNorth Carolina, Chapel Hill