Introduction

In the past decade, the discipline of policy analysis has regained momentum in the form of attention for ‘evidence-based policy’ (Adams, 2004; Walter et al, 2005; Pawson, 2006; Nutley et al, 2007; Boaz et al, 2008a, 2008b; Head, 2008, 2010; Argyrous, 2009; Banks, 2009; Leigh, 2009; Cartwright and Hardie, 2012). The concept of evidencebased policy was originally developed in the Anglo-Saxon context of British and US public administration (Sherman et al, 1997; Strategic Policy Making Team Cabinet Office, 1999; Solesbury, 2001; De Groot, 2010; Head, 2010). It rapidly expanded to other countries, in part due to the influence of international organisations such as the Organisation for Economic Co-operation and Development (OECD) and the European Union (EU) (Burns and Schuller, 2007; EU, 2007; Head, 2010). In the Netherlands, the rise of evidence-based policy was fuelled by various conclusions of successive parliamentary commissions examining policy failures, which indicated that minimal thorough analysis, and a lack of a sound scientific foundation, were significant, consistent deficiencies that contributed towards policy failure (Bax et al, 2010; De Wree et al, 2010; see also Chapter Twelve on the role of party-political think tanks). The message that resonated in Dutch policy organisations as a result was that successful policy requires sound and rigorous analysis, as well as underlying policy theory (CPB, 2002; Cornet and Webbink, 2004; De Wree et al, 2010).

One of the most common interpretations of evidence-based policy at the Dutch national level is that it relies on a quantitative analysis of effectiveness. In other words, it calls for randomised experiments or similar research designs. This often draws comparisons with the double-blind clinical trials that establish the efficacy of new medicines. In both cases, the objective is to establish a causal relationship between an intervention and its intended effects (Cornet and Webbink, 2004). Like proving medicinal effects, the allure of proven policy results is appealing in a political debate. The notion that there exists evidence, proof and, above all, ‘guaranteed value for public money’ resonates strongly in a debate that emphasises the effectiveness of policy. Measurable effect has become an important driver for political debate (Head, 2010). Public funds must be spent efficiently, and in order to do so, policy interventions must be tried and tested. However, effective policy is not the same as policy with a proven effect in a clinical trial experimental setting.