THE MIDDLE EAST DEMOCRACY TRAP: Why Electoral Competition Undermines Property Rights in Islamic Countries

This dataset accompanies the research paper "The Middle East Democracy Trap: Why Electoral Competition Undermines Property Rights in Islamic Countries" which examines how electoral democracy affects property rights protection across 93 developing countries from 2007-2023. Key Finding: Electoral democracy's effect on property rights is conditionally negative, depending on sectarian fragmentation and institutional capacity. In Islamic-majority countries, electoral democracy significantly undermines property rights (β=-4.73, p<0.01), while in non-Islamic developing countries, democracy shows positive or neutral effects (β=+1.82, n.s.). The interaction between electoral democracy and Islamic context is highly significant (β=-6.55, p=0.002). Dataset Contents: Cross-sectional data (2023) for 93 developing countries with complete variables Panel data (2007-2023) with 1,526 country-year observations from 102 countries Regression output tables (cross-sectional, panel, lagged variables, interaction models) Visualization of regional heterogeneity in democracy-property rights relationships Original data sources: International Property Rights Index (IPRI 2007-2025) and Varieties of Democracy (V-Dem v15) Methodology: The analysis employs OLS regression with robust standard errors, country fixed effects panel models, lagged independent variables for causal identification, and difference-in-differences estimation using the Arab Spring as a natural experiment. Regional subsamples demonstrate the pattern is specific to MENA/Islamic countries, not universal. Policy Implications: Results suggest democracy promotion requires institutional sequencing—building rule of law, judicial capacity, and civil society before competitive elections, particularly in sectarian contexts. Rapid democratization without institutional foundations can worsen property rights protection. Replication: All analysis code (Python/Jupyter notebook), raw data files, and regression outputs are included for full reproducibility.