Introduction

In most welfare states, the quality of education greatly determines the future opportunities of children and of society in general. Therefore, high quality education is considered an important public service and schools that are underperforming are considered a policy problem. In the Netherlands, the government has developed policies and interventions that deal with underperforming schools – referred to as ‘very weak schools’. These schools are included in a black list that is published online annually and then they become subject to increased supervision by the Inspectorate for Education. However, the results are diverse. In some schools, the approach has led to major improvements in quality and positively surprising results, whereas in other schools the situation has spiralled out of control in a series of undesired policy effects. Similar interventions have had different effects; some intended, others unintended; some desired, others undesired. This empirical observation is the starting point of this chapter: what causes the differences in outcomes of similar policies in similar contexts? Can patterns and causation be found in what seem to be unpredictable, unstable, and chaotic systems?

In general, policy is constructed along the lines of intentions and expectations of policy makers. Policy makers often assume, or desire, a ‘magic button’ that they can push to achieve intended outcomes. However, for decades, social scientists, historians, and economists have all been fascinated by the ubiquity of the so-called ‘unintended’, ‘unanticipated’, or ‘unexpected’ consequences of policies; policy plays out differently, sometimes positive, but often tragic. How can such ‘surprises’ be explained and understood? What mechanisms underlie the intentions and anticipation – or lack thereof – of public policy? Why are ‘great expectations dashed’ (Pressman and Wildavsky, 1984)?

Throughout its 40-year history, the Policy & Politics journal has devoted much attention to explaining the – often unexpected – outcomes of policy interventions. From the first volume in 1972 (see Sigsworth and Wilkinson, 1972) to the current volume (see Lindsay and Dutton, 2013) many articles have tried to open the ‘black box’ of the mechanisms through which policy interventions reach their impacts, and to identify conditions that affect the transfer of policy interventions into policy outcomes.