The Irawadi Dynamic Model: When Economies Worsen and Trust Falls

The Irawadi Dynamic Model Version 13.0 delivers three major upgrades: (1) geographic expansion from 37 to 53 countries via the Global Populism Dataset (GP Dataset, 1,081 country-year observations, 2000–2024), confirming that the ITI effect is larger and more significant globally than in the European sample (Country FE: βIT I = −85.7, p < 0.001 global vs −29.4, p < 0.001 EU); (2) formal threshold estimation via Hansen-style grid search, yielding τIT I = 0.787 as the ITI level minimising total within-regime RSS, and τGDP = +11.6% as the 3-year GDP change below which the economy is classified as deteriorating — providing the first quantitative content to Equation 4 in the IDM’s theoretical framework; and (3) updated Granger causality test on the full 2000–2023 panel confirming that past A3 electoral gains predict future institutional decline (F = 24.1, p < 0.001, country fixed effects), strengthening the ratchet hypothesis. The non-European sample (28 countries: Asia, Americas) shows an ITI coefficient of β = −103.2 (p < 0.001) in country FE specifications, indicating that institutional decline is an even stronger predictor of right-populism outside consolidated European democracies. Parliamentary polarisation (REPDEM RILE divergence) adds β = +0.133 per unit alongside ITI (p = 0.213, country FE) — the sign is correct but does not reach significance in this sample. All prior findings from Versions 9.0–12.0 are retained: OOS forecast validity (r = 0.622, p < 0.001), median income interaction (p = 0.005), ESS survey validation, A4 executive corruption trajectories, and cabinet A3-in-government coding.