To explain both the causes and the consequences of social spending, it helps to have a coherent framework, one that reveals the logical links between the host of points being made. Without a unifying theoretical approach, the numerous conclusions this book has reached may seem ad hoc and eclectic. Readers will be better served, and perhaps better persuaded, if the ideas are all part of a single logic, with a minimum of qualifications. Fortunately a unifying framework does fit the book's conclusions, and this chapter sketches it.

The approach offered here violates a particular scientific procedural code. According to one orthodox code, one is supposed to get a set of theoretical assumptions and work out their predictions, before testing the predictions empirically. This book travels the path in reverse. Induction comes first, and logical deduction comes last. I am definitely guilty of reverse engineering from the historical facts back to a set of model predictions and only then to the assumptions of the model. This chapter tries to optimize its theoretical design in a particular way, minimizing some combination of the falsehood and the complexity of the assumptions that deliver predictions that fit the facts. The quest, then, is for the simplest plausible set of assumptions that predict most of the book's main empirical conclusions.