This paper examines the social learning models of policy of Hall and May attempting to create a synthesis of the best elements of each. We then apply the revised model to three specific instances of macroeconomic policy in Britain; the introduction of ‘Keynesian-plus’ policy in the 1960s, the movement from Keynesianism to monetarism, and the experiment with monetarism in the 1980s. In each case study, the degree of policy change is assessed, and possible reasons for that level of change explored. We conclude that a more social constructionist approach is required to understand the link between policy instruments, indicators, and paradigms, and, alongside this, a greater need to understand the implications of the assumptions underlying policy.