Samuels and Snyder in their recent article in this Journal presented an index of malapportionment (i.e., the discrepancy between seat shares and population shares by electoral districts) in national legislatures for seventy-eight countries. This short Research Note merges their index with the Gini index, a commonly used measure of income inequality, and explores cross-nationally the relationship between inequality in political representation and inequality in economic conditions.

In a recent article in this Journal Marsh, Smith and Richards (MSR) note the massive recent changes in the organization of British government and the attention the bureau-shaping model has received both at a theoretical level and as an explanation of changes. They suggest that the model has breathed new life into debates about the behaviour of officials and is important in the context of the ‘Next Steps’ agency reform. They state two aims of their article: ‘First, it is a critical contribution to the literature on the bureau-shaping model’, and secondly it examines ‘the model's utility as an explanation of the changes that have occurred in British central government in the past decade’. They also use their arguments as part of an assault upon rational choice and empirical political science more generally in favour of interpretative sociology. However, in this Note, we respond to their work on the bureau-shaping model and rational choice.

Dowding and James's response to our critique of the bureau-shaping model makes some important criticisms that we welcome. However, we suggest that they ignore the empirical evidence we presented in our original article (because of their narrow definition of acceptable data) and we believe that they misunderstand our arguments because of the positivist paradigm within which they operate. Our response covers three main issues – epistemology, methods and evidence.