'Testing theories with historical data has become increasingly popular among social scientists. Often, however, history is 'used' uncritically, after stripping social and political phenomena of their spatial and temporal context, or by forcing monocausal models on complex and distant realities. Kreuzer's impressive volume offers a very persuasive warning against such strategies and proposes a conceptual apparatus to use history consciously in the social sciences. Written in a very accessible language, this book should be required reading for all graduate students and social scientists who plan to engage with history.' Giovanni Capoccia, Professor of Comparative Politics, University of Oxford