Rationality

The concept of rationality is central to policy formulation and implementation. Rationality refers to the quality or expectation of being logical, reasonable or scientific, and therefore points to a set of universal definitions and appropriate standards against which the usefulness of policy decisions in achieving certain goals and outcomes can be judged to be internally consistent, valid, reliable, even replicable. This view of policy decisions or advice as rational presents an ideal representation of policy as evidence-driven, namely that the policy process is a scientific or apolitical practice structured around hypothesis testing that uses value-free knowledge to generate contextual or causal explanations for social change. Leaving aside the idea that policy is evidence-or fact-based, which is arguably false in some cases, there is the notion that the rationality of the policy process can be traced to its proclivity for and demonstration of science.

But science has a communicative context according to Andrews (2007) since scientific knowledge requires both consensus and authority in order to be considered legitimate. In other words, scientific knowledge claims never exist intrinsically and independent of human activity, but rather emerge through their relation to other sources of authority or processes of legitimation and consensus formation that determine their ‘politicisation’ and role in the policymaking process. Moreover, even assuming that the policy process is evidencebased and scientific in the Enlightenment sense of universal and value-free, it is important to separate out the rationality of the policy process (the means by which policy makers assimilate perfect information into policy decisions so that they are fair or accurate) and the rationality of the policy actor. This is because, as Dunleavy (1991) reminds us, some people may ‘operate through intransitive preference orderings’ (p 249) or exercise their decision-making through emotional attachments and ethical commitments not captured by a standard or comprehensive rationality that privileges forms of instrumental choice. Moreover, according to Bevir and Brentmann (2007), economistic models of rationality are inadequate for making sense of the ways in which diverse rationalities combine in unique ways to produce iterations of local reasoning and individual experience.