Delfs Erbo Andersen, DavidSkaaning, Svend-ErikKnutsen, Carl Henrik2023-10-232023-10-232023-101653-8919https://hdl.handle.net/2077/78927State weakness is often emphasized as a key determinant of democratic breakdowns. However, previous studies have failed to appreciate how different aspects of state weakness pose different challenges. Against this backdrop, we examine the relationships between two fundamental dimensions of state capacity (coercive capacity and administrative capacity) and different modes of democratic breakdown, i.e., incumbent-driven and nonincumbent driven takeovers. We propose that coercive capacity mainly enables containment of rebels and coup-plotters, which reduces the risk of nonincumbent takeovers. Conversely, we expect that administrative capacity mainly serves to prevent executive aggrandizement, which reduces the risk of incumbent takeovers. Global analyses of democratic breakdowns between 1789 and 2020 support only the second expectation. Coercive capacity, reflected by territorial control and military personnel per capita, usually drops below accepted significance levels for both modes of democratic breakdown. In contrast, indicators of meritocracy, impartial public administration, and predictable enforcement that proxy administrative capacity show a significant, negative relationship with the risk of democratic breakdown, but only for incumbent-driven takeovers.engText