In Chapter 3 we paid homage to those who have advanced the systematic study of PB. Nevertheless, the refrain remains that previous explanations of PB's achievements in deepening democracy have either relied on small numbers of cases or eschewed systematic comparison. They have produced long lists of conditions that are deemed essential for democratic deepening. If we add up all the conditions that are claimed to be necessary for any explanatory theory of deep democracy in PB, it makes a long list – contentious civil society, organized civil society, absence of co-opting strategies, political commitment, demonstration effects, rules, size, wealth, support of officials, and many intersections thereof have been offered as explanatory conditions to name just a few. What is more they have been offered to explain a range of possible outcomes from changes in state–society relations, to redistribution of wealth, increases in education, efficiency of government spending and increasing vote share for the parties implementing PB, to again, name but a few. The first chapter of this part of the book explains the considerations in specifying an inclusive, but robustly specified, examination of more than a handful of diverse PB cases.

I must begin by explaining the necessary tools for identifying necessary conditions. Bent Flyvbjerg argued that ‘good social science is problem driven and not methodology driven in the sense that it employs those methods that for a given problematic, best help answer the research questions at hand’ (2006: 242). In setting out on this research journey my aim was to harness the knowledge provided by existing ethnographies, drawing on single-case examples and small-N comparative work to cumulate knowledge of causes and outcomes of citizen control of spending decisions worldwide in a systematic but case-sensitive way. I was also guided by Sartori's caution regarding how comparative politics is practised. He wrote over 50 years ago that: