Discover Society

Christina Boswell and Eugenia Rodrigues

The Policy Briefing section of Discover Society is provided in collaboration with the journal Policy & Politics. The section is curated by Sarah Brown.  

It has long been observed that policies can get lost in implementation. The best intended legislation or programme adopted by central government can get reinterpreted, distorted or even subverted when applied at local level, or across different areas of government. This was certainly the case with the British Labour government’s system of targets rolled out in the 2000s. Number 10 and the Treasury (the ‘core executive’) adopted a series of quantified performance targets designed to improve public services. And the government even monitored how far they were being achieved through rigorous reporting arrangements. But the targets were appropriated and applied in quite different ways across departments. What factors shaped how different parts of government implemented targets?

In our recent Policy & Politics article: Policies, politics and organisational problems: multiple streams and the implementation of targets in UK government, we develop John Kingdon’s idea of ‘multiple streams’ to try to understand differential implementation across sectors. According to the multiple streams approach, policies are adopted where three elements converge: a policy, a problem, and amenable political conditions. The policy ‘stream’ refers to the policy idea or proposal. This policy idea needs to find a good fit with prevailing policy problems: policies or social problems considered to be in need of change. And the political environment needs to be conducive to such change: the policy idea needs to chime with the ‘national mood’, and be supported by key political actors and organized interests. When these three streams converge, a political ‘window’ can open, allowing astute ‘policy entrepreneurs’ to press for policy change.

The multiple streams approach was originally developed to explain agenda-setting in US public policy. However, a number of scholars have suggested that the multiple streams approach is also useful for understanding policy implementation. By understanding policies, problems and politics as separate streams, it provides a neat way of conceptualising how a centrally adopted policy may be understood and implemented in very different ways, dependent on the problem and politics streams. In other words, different levels of government, or different policy sectors, may be tasked with implementing the same policy. But they might have quite distinct understandings of policy problems, or may be guided by different political dynamics. So these two streams – problems and politics – can help explain the often patchy or varied implementation of policies in different political contexts.

However, we suggest that in order to apply the multiple streams approach to understand implementation, we need to make two adjustments. First, we suggest that it is useful to focus on implementation in the departments or ministries responsible for implementing policy. So we adopt the perspective of organisations in the public administration, rather than focusing on (party) political dynamics. This makes sense because implementation is most clearly dependent on the decisions and behaviour of the actors responsible for overseeing and implementing policy. In order to understand implementation, we need to understand how these organizations in the public administration perceive the policy problems they are facing, and what political pressures they are subject to.

Second, we suggest that two factors are particularly important in shaping implementation: the strength of the core executive’s commitment to the policy, which is another way of specifying the ‘politics’ stream; and the policy’s ‘match’ with organisational problems, or what we can refer to as the ‘problem’ stream. In other words, does the proposed policy seem useful or appropriate to the organisation as a way of addressing what it perceives to be the challenges facing it?

We suggest that different combinations of political commitment and fit with organisational problems yield four possible modes of implementation. First, where there is strong central commitment and the proposed policy matches organisational problems, then we can expect effective implementation. However, where the policy doesn’t appear to fit with the organisation’s perception of policy problems, then we can expect either weak implementation (if central government is not that committed either). Or, if the core executive is strongly committed to the policy but the organisation isn’t, then the organisation may engage in ‘decoupling’: complying with the policy in its rhetoric and formal structures, but in practice failing to carry out the changes required to implement it. A fourth possibility is that the core executive is not strongly committed to the policy, but the organisation is – in this case, we can expect ‘bottom-up’ implementation, driven by the organisation.

                                                                    Politics Stream

We examine this model by looking at the implementation of targets on carbon emissions, defence procurement and asylum. Each of these policy areas was subject to a series of performance targets, imposed by the Labour government between 2000-2010. Our analysis explores how these targets were implemented in the key departments responsible for meeting targets. In the case of carbon emissions, this was the Department of Energy and Climate Change (DECC), and its predecessor the Department for Environment, Food and Rural Affairs (DEFRA); for defence procurement, it was the Ministry of Defence; and for asylum, the responsible department was the Home Office. The analysis draws on 54 interviews carried out with officials, special advisors and ministers. The research was carried out as part of an ESRC sponsored project on “The Politics of Monitoring”.

We find support for our model. In the case of asylum, there was a confluence of the target ‘policy’ and the politics stream in the early 2000s, but only loose alignment of these streams with the organization’s problem construction. This initially produced decoupling, and then coercive implementation as the core executive stepped up its intervention and monitoring to drive through organizational adjustment. Towards the end of the decade, however, the politics stream became detached from the policy and problem streams. By this time, the organization had internalised targets as an appropriate solution to a range of organisational problems, generated by adjustments to its problem stream. This produced what we refer to as bottom-up implementation.

In the case of defence we can observe a combination of decoupling and non-implementation. Political commitment to the targets was weaker than in the case of asylum, and the core executive lacked the will and the capacity to impose reform on the Ministry of Defence (MoD). The MoD was therefore able to avoid introducing substantial adjustments to organizational practice. However, the organization did perceive the need to adjust certain of its practices in order to be seen to be meeting targets, decoupling rhetoric from informal practice. We can depict this as weak confluence of policy and politics, detached from the problem stream. This produced a combination of our decoupling and non-implementation modes.

Finally, the climate change case offers another example of a switch in modes of implementation over time. In this case, we can observe an initial confluence of the three streams, with the politics stream being represented by both international and national political considerations. However, targets were thoroughly embraced by senior management in DEFRA and especially its successor DECC as a solution to organizational problems. Thus implementation came to be driven by bottom-up commitment to implementation. In this case the political stream did not diverge, but was replaced or superseded by the organizational problem stream.

One of our key findings is that organisations can switch between modes of implementation, as core executive commitment changes over time, and as new organisational problems emerge. Our analysis also reinforces how difficult it is for core executives to steer implementation. Even where they set clear, specific targets and monitor them rigorously, there is huge scope for organisations to decouple formal compliance from informal deviation. As the asylum case shows, the core executive may need to resort to quite resource-intensive, intrusive – and damaging – forms of intervention to achieve its goals. In the case of the Home Office, one has to ask if this was a price worth paying.

Christina Boswell is Professor of Politics and Director of Research in the School of Social & Political Science at the University of Edinburgh. Eugenia Rodrigues is Research Fellow in Science Technology and Innovation Studies, in the School of Social & Political Science at the University of Edinburgh.

Image from Freephoto: CC BY-NC-ND 3.0