Global Peace in Retreat? Trends, Drivers, and Divergence in the Global Peace Index, 2008–2023

Abstract

This paper analyses fifteen years of Global Peace Index (GPI) data — 163 countries from 2008 to 2023 — to document the magnitude and distribution of global peace deterioration, decompose it into domain contributions, and characterise the growing divergence between stable and fragile states. The global average GPI score deteriorated by 5.04 per cent over the period, from 1.971 to 2.071, with the Ongoing Conflict domain accounting for the largest proportional decline (+14.0%), followed by Societal Safety and Security (+5.4%), while the Militarisation domain registered the only improvement (−6.2%). Using variance decomposition, a two-way fixed-effects panel model, and k-means cluster analysis, we establish three findings. First, deterioration is overwhelmingly concentrated among a small set of crisis states — the top ten deteriorators account for 36.5 per cent of the total global decline — rather than reflecting a broad-based global trend. Second, the cross-national standard deviation of GPI scores widened from 0.402 to 0.487 over the period, indicating increasing global divergence: the gap between the 25 most and 25 least peaceful countries grew from 1.238 to 1.528. Third, regional trajectories differ sharply: MENA (+10.9%) and Sub-Saharan Africa (+9.0%) deteriorated most severely, while Europe and Asia-Pacific remained broadly stable. A two-way fixed-effects panel model identifies statistically significant year effects from 2016 onward, consistent with a structural break in the peace trend coinciding with the peak of Islamic State (IS) activity. The results imply that global peace improvement is a tractable policy challenge requiring concentrated intervention in a small number of conflict-affected states, and that the deteriorating Militarisation domain trend since 2022 poses a new threat to the sole domain that had previously improved.