Cultures of Evidence series 2:The Normative Turn in Evidence and Policy From ‘Evidence Use’ to the ‘Good Governance of Evidence’ 

The Normative Turn in Evidence and Policy From ‘Evidence Use’ to the ‘Good Governance of Evidence’ 


By Justin Parkhurst 
Associate Professor, London School of Economics and Political Science 
twitter: @justinparkhurst  


How do we improve the use of evidence for policymaking? This seemingly simple question does not have a correspondingly simple answer. Since the 1990s at least in the UK and several other Western countries, there has been an explicit embrace of ‘evidence based policy’ or increasing the ‘use’ of evidence within policy sectors. Yet many scholars have noted for years that it is over-simplistic to assume that evidence can simply be ‘used’ in this way when considering public policy choices.  

Those who reject the concept of ‘evidence based policy’ typically critique how the concept fails to capture the political realities of decision-making. This includes both the processes involved in using evidence (involving time-limited decision makers working through specific institutional structures),  but also critically how the political nature of policy decision-making involves competing interests, values, social goals. In short, policymaking requires value-based choices evidence cannot make for you – and those choices will dictate which evidence is relevant or appropriate to use. You cannot simply assume there is such a thing as an ‘evidence based policy’ when there is disagreement over what the evidence should even be of. Ignoring the importance of values and politics is not, as some might think, a way to remove political considerations, but rather it ‘depoliticises’ decision making – rendering hidden the political choices made around the pieces of evidence used.   

Academic research on evidence use has developed greatly in recent years to better understand when, why, or how pieces of evidence are utilised within public policymaking spaces. Drawing on concepts of cognitive bias, bounded rationality, institutionalism (in various forms), political-economy, or other fields – many authors have noted that the ultimate form or features of ‘evidence use’ will be a result of the nature of the policymaking process – including who is making decisions, where, and for what goals or ends. These insights allow a great deal of understanding of why idealised notions of ‘evidence based policy’ may not be achieved (or may not ever be achievable), as well as knowledge about where or when types of evidence use may come about.  

Yet what has been much less discussed in the academic literature is the question that many advocates of evidence use are ultimately concerned with – understanding what better use of evidence would actually look like, or how to establish a system that improves evidence use for policy.  

Considering this requires taking an explicitly ‘normative turn’ in thinking about the use of evidence – a turn that allows us to ask (and potentially work towards answering) three key questions: 

  1. What constitutes good (or better) evidence for policy needs? 
  1. What would represent improvements in the process of evidence utilisation? And 
  1. How can we establish better systems of evidence use for policy at national (or other decision making ) levels?  

Each of these questions has explicitly normative components – allowing discussion of the values by which evidence use itself can be judged.  

The first of these requires moving beyond the traditional hierarchies of evidence – which focus on evidence of intervention effect, but not necessarily policy importance. Instead, it allows us to consider what is appropriate evidence for particular policy needs. One way to consider appropriate evidence can be to start with goals or tasks of a decision-making body. Colleagues and I developed a ‘programmatic approach’ to evidence use by asking programme officials what their goals were, and which pieces, forms, and uses of evidence were most appropriate to those goals (which we did for a set of malaria programmes in Africa, but which could conceptually be applied to almost any public sector technical programmes). This approach takes as given that programme officers have established goals to achieve – yet another way to reflect on good evidence for policy, however, is to critically reflect on which goals are being pursued in the first place.  

In a recent book analysing the use of evidence to inform Public Policy in Brazil, I argue in the preface chapter that a good use of evidence from a public policy perspective is one which can be judged to serve the public interest. Establishing whether or not a use of evidence serves the public interest, however, requires a clear process of ‘goal clarification’ – in which the desired outcomes of a course of policy action may be stated, debated, or challenged by the public and/or their legitimate representatives.  This crucial step – to have an explicit goal clarification process within evidence advisory systems – is often absent from formal evidence synthesis processes and bodies. This may be due to a fear that discussing policy goals could be seen as political interference by scientists, or it may be due to an assumption that evidence provision is a purely technical process. Yet having an explicit goal clarification step (or having evidence providers demand clarity on goals) is not the same as imposition of values by scientists – indeed it is a process by which scientists can make transparent the public values being pursued, and ensure they provide the appropriate evidence for those needs.  

Taken as a whole, taking a normative turn in evidence use allows us to reflect on a broad set of normative concerns related to evidence, the use of evidence, and the appropriate evidence for particular needs. This requires drawing both on principles of scientific good practice, as well as principles of democratic representation and accountability. It also allows us to reflect on the third of the questions listed above in relation to how to build institutional arrangements, processes, and systems within countries that can serve to improve evidence use. In a recent book, I have described the institutionalised arrangements of evidence advice as ‘governing’ the use of evidence in policy making – with the normative principles discussed here allowing the development of a framework to consider what the good governance of evidence might look like (as visualised below).  


The goal of building such a framework is to help facilitate explicit discussions around what constitutes improved or better uses of evidence within public policy processes. As such, it is hoped it can be applied to critical or comparative discussions of evidence use from this perspective. So, for example, in a recent paper, Isaac Weldon and I attempted to apply these GGE principles to consider if the process of developing the 2020 Canada Food Guide might have used evidence in a better way than had been done previously, while Shaxson has applied it to reflect on evidence utilisation in both US and UK public sector contexts. While these represent initial efforts, such an approach can be applied to a much larger range of policy decisions. Indeed, the framework can further be used to consider whether or how entire systems of evidence are designed in relation to both scientific and democratic needs – so as to work to improve evidence use within those systems, and ultimately to serve the public interest.