Political scientists have conventionally assumed that achieving democracy is a one-way ratchet. Only very recently has the question of “democratic backsliding” attracted any research attention. Complexity science offers a range of powerful new tools to analyse the stability of social institutions in general, and democracy in particular. What makes a democracy stable? And which processes potentially lead to instability of a democratic system? We argue that democratic instability is best understood with tools from complexity science. We have analysed recently made available, detailed quantitative data on nations’ political state. We will present this novel complexity-science based analysis and the insights gained into the mechanisms of democratic (in)stability. The explanatory power of complexity science arises from several features of complex systems. We will discuss their relevance in the context of democracy and offer several policy recommendations to help (re)stabilize current systems of representative democracy.