1.2 The primary focus of the Magenta Book is on policy evaluation1 which examines how a policy or other intervention was designed and carried out and with what results.
1.3 Therefore, the focus is on the actual practice and experience of the policy and observations on what actually happened following implementation (rather than what was expected or intended, for instance, which is the topic of appraisal).
Evaluation can employ a variety of analytical methods to gather and assess information, and the choice of methods employed in any particular instance will depend on a wide range of factors which are the subject of the remainder of this book. In turn, this choice will affect what questions the evaluation might be able to answer and how strongly its conclusions can be relied upon. However, the focus on actual experience of a policy means that evaluation as described here is an impartial process which asks objective questions such as:
• What were the impacts of the policy?
• How was the policy delivered? and;
• Did the policy generate value for money?
1.4 Even when an evaluation asks a question on a subjective topic (such as stakeholder perceptions of effectiveness), it will seek to answer it in an objective way, such as:
• How successful did stakeholders think the policy was in achieving its objectives?
• Did the policy succeed in improving the public's perceptions of the problem?
1.5 In practice, of course, questions will be more complex and specific than this, and will often include consideration of how different features of the policy affected the way it performed and delivered, and how its outcomes varied across those it impacted upon: what worked for whom in what circumstances. The types of questions which different types of evaluation can answer are the subject of Chapter 2. Good evaluation, as described in this book, is an objective process, therefore the answers it provides will give an unbiased assessment of a policy's performance. For this reason, evaluation results might be challenging in real terms and from a presentational perspective.
1.6 However, good evaluations should always provide information which could enable less effective policies to be improved, support the reinvestment of resources in other activities, or simply save money. More generally, evaluations can generate valuable information and contribute to a wide range of initiatives and objectives. For instance good evaluation can:
• provide a sound scientific basis for policy making, by providing reliable understanding of which interventions work and are effective. An understanding of how and why policies work can also be used to inform the development of new policies, and to improve the effectiveness and reduce the burden of existing ones;
• underpin practical resourcing and policy making exercises such as Spending Reviews and the formulation of new strategies. They can contribute to the setting of policy and programme objectives, and can be used to demonstrate how those objectives are being met; and
• they can therefore provide accountability, by demonstrating how funding has been spent, what benefits were achieved, and assessing the return on resources. This can help to satisfy external scrutiny requirements and comply with sunset clauses and other formal requirements that make a link between evaluation and the continuation of the policy.
1.7 Good evaluation, and the reliable evidence it can generate, provides direct benefits in terms of policy performance and effectiveness, but is also fundamental to the principles of good government, supports democratic accountability and is key to achieving appropriate returns from taxpayers' resources. A good evaluation is therefore a normal and natural part of policy making and effective government and is a powerful tool available to the policy maker.
_______________________________________________________________________
1 The Magenta Book generally uses the term 'policy evaluation' to refer to evaluations covering projects, policies and programmes. How evaluations differ across these various types of intervention is discussed in Chapter 2.