HOW WELL DO THE MODELS FORECAST?

The previous chapters in this book have elaborated many different models of political decision-making in the European Union. To make their forecasts, some of these models focus on the incentives created by EU legal regulations or decision-making rules. Other models emphasise the power of bargaining in political decision-making. Still others start from logrolling, coalitions, or the spatial theory of voting.

The aim of this book is to set out all these models of EU decision-making, and then to evaluate how well the models predict actual decisions. Most of the discussion is quantitative. Yet in important respects, we mean to integrate prior case studies, formal theory, and statistical methods. For example, each of the modelling approaches represented in this book builds on one or more central aspects of political life known from dozens of skilful case studies of political decisions. All competent model-building depends on careful qualitative research in which explanatory factors are identified and tentative empirical generalisations are formulated. No model is worthwhile if, like some formal theorising, it applies to nothing in particular. Case studies have generated most of the interesting hypotheses in political science. They are the essential foundation for most model building.

Moreover, case studies play a crucial role in evaluating theory.