Beyond top-down and bottom-up:hw do we currently understand policy implementation

Australia has experienced a number of high profile policy implementation failures in recent years, such as the 'Pink bats' scheme. Implementation, as a separate part of the policy process and as a scholarly endeavour, is creeping back onto the radar (thankfully). Today's post by Charlotte Sausman, Eivor Oborn and Michael Barrett discus orignally appared on the Politics and Policy Journal Blog as an overview of the paper - Policy translation through localisation: implementing national policy in the UK

It remains the priority of policy makers to show that they have put in place well designed policies that have demonstrable effect, in order to give a good account of their time in office. Whilst many depictions of the policy process focus on something that is driven from the ‘top down’, implementation scholars have over several decades provided particular understanding of the ‘bottom-up’, looking more qualitatively at organisational responses to policy initiatives. Through developments in New Public Management to current research on policy design, studies have moved away from the dichotomous ‘top-down’ versus ‘bottom-up’ and yet the problem of how to understand policy implementation endures.

At the same time, the current drive for ‘evidence-based policy’ is premised on the belief that if policies can be designed on the best evidence, it is more likely that they will be implemented with measurable effect in terms of desired outcomes. Policy makers believe both in the positive effects of evidence behind the policy and the translation of that evidence-based policy into practice. In the UK health sector, where our research was based, current policy design tends to favour such a rational approach, where putting policy into practice is a discrete linear process following clearly defined policy goals.

Despite developments in the study of policy design, including articles in Policy & Politics, less attention has been given as to how policy design itself influences implementation.  In the health sector, much of the policy implementation literature addresses specific concerns with the adoption and promotion of evidence-based guidelines, accumulating knowledge around the variables that affect implementation and the transfer of experimental evidence to ‘real world’ clinical situations. This understanding perpetuates the ‘implementation gap’ in health care, whereby local actors deviate from what the centre directs.

In studying the implementation of a large-scale mental health policy in the UK our research revealed the dynamic, iterative nature of the implementation process and its effect on reshaping policy. Implementation processes create important feedback mechanisms to policy makers and lessons for policy design. We also sought to understand the role of policy design in enabling coordination between multiple actors during the implementation process.

Adopting a ‘local universality’ perspective the research showed that an important tension in implementing policy that is neglected in the literature is the relationship of the new policy with the prevailing infrastructure, procedures and practice. The ‘local universality’ is the product of these translation processes.  What is enacted in each location – be that district, organisation or sector – is a unique product of the negotiations which are collectively produced.   They  also include the creation of new relations, new beliefs, new knowledge in the practices wherein policy is implemented. Hence, the local implementation site – whilst adhering to the overall policy design and specifications, will – through implementation processes – always be ‘unique’. Rather than ‘cookie cutter’ policy implementation which seeks exact replication in each location, the process is more akin to building a new development, where specifications are given for the number of houses, the size, and the quality, but how they are built, using local materials and trades, and how their overall design fits with the local terrain, may be different.

In practice, local realities and adaptions will always shape policy implementation in ways that could not have been predicted. Policy design therefore needs to include learning from implementation processes as well as from upfront evidence-gathering. Even pilot-based evidence is no substitute for practice-based experiential policy learning through large-scale implementation – and this needs to be fed back to policy makers.

Finally, balance is required between maintaining adherence to policies, whilst at the same time allowing local adaptation. In the healthcare sector, practice is now a highly complex world due to the range of overlapping policies, to continually updated, practice-based guidelines already in place and with a changing infrastructure and local population. Local actors need the discretion to make sense of, and adapt, where appropriate, national policies to the local setting.