The recent global financial crisis has been a serious stress test for representative democracies. Voter support has supposedly become more volatile, fragmented, and polarized, leaving elites with an intricate mix of economic and political challenges. However, a closer look at a new data set of European party systems during three major crises (1929, 1973, and 2008) reveals that the reality is less dramatic than the popular impression suggests. We propose a novel theory of party-system change that explains both the impact of economic crises as well as the robustness of party systems to more serious destabilization. Since voters and elites are risk averse, economic crises tend to disturb party systems that are generally “restrained” but, at the same time, help consolidate more complex systems. This explains why party systems rarely fall apart, nor do they reach ultimate stability. We provide quantitative evidence and qualitative illustrations of “restrained change” in various party-system dimensions.