How can government departments deliver substantial and sustained improvements in public services given that that they are often one or more levels removed from the workers who serve the public at the front line? The primary method used by the centre to trigger and deliver performance improvements in the field is issuing the circular - setting out what must be achieved and by when. Sometimes this is informed by pilots or bolstered with guidance. Whilst this approach works in some circumstances, civil servants need to widen their approach to change management if they are to succeed in bringing about significant change.
Four possible approaches to managing performance change in public services are described below. They differ mainly in the extent to which the change effort is influenced and directed by local circumstances and management.
Centre-out model
At its simplest, a circular is issued from the centre to all providers. It describes the desired performance improvement and some time later there is a check to see whether what was asked for materialised. The required change is tightly defined and involves a new way of working, or a target relating to a specific aspect of work. The centre does not trouble itself unduly with what the position may look like at ground level, for example the level of local resources, or socio-demographic factors that make the challenge more difficult in some areas than others. Such complications are left to local management to sort out. There are variants to this theme which include:
- Issuing of guidance
- Running of regional seminars
- Use of pilot sites to prove feasibility and inform central guidance
- Frequent checks on progress, and pressure applied to those performing poorly.
The centre-out model can only work where the centre is very clear at the outset about what it wants to achieve, that it is achievable, and that there are sufficient skills and motivation available at a local level to deliver. Resource requirements – certainly at the centre – are low. By focusing on desired outcomes rather than prescribing processes, it allows service managers to develop locally appropriate means to centrally defined ends.
The centre-out model is unlikely to succeed, however, where the change is complex, resistance is high, or in services which do not have a strong managerial culture.
Local audit model
The local audit model has two main elements. The first is to carry out a national study of the particular service area to identify variation in performance, scope for improvement, and what constitutes ‘best’ practice. The second element is the design and rollout of a programme of audits of local performance. The national study and local audits act as catalysts to drive and direct improvements in local services towards best practice standards.
This model has been used for many years primarily by the Audit Commission to improve the value for money (VFM) of local government and health services. Government inspectorates and bodies such as the Office for Standards in Education (OFSTED) and Her Majesty’s Inspectorate of Constabulary use similar approaches to raising standards. In contrast to the Centre-Out model, it is not necessary to know exactly what improvement you require at the outset, since the national study helps provide the detail to broader policy objectives and define benchmark performance standards.
The methods for conducting the local audits are designed during the course of the national study that will include detailed examinations – or trial audits – of a number of service providers as well as analysis of national data. Local auditors are brought up to speed through training, audit manuals and other materials. Ongoing support is available from the national study team, which will have developed a high level of subject expertise.
The main call on central resources in this model is the diagnosis stage which results in the production of a set of solutions customised into local performance improvement programmes. Initial testing and discussion of the full range of potential solutions during the national study with professional experts ensures a fair level of confidence that they are both workable and will enjoy the support of the relevant professions.
Armed with the local audit recommendations, service providers are left to make the changes happen on the ground. In a follow up visit, the auditor will ask the provider to confirm that has happened.
This model is most useful where the required change is limited in scope, such as being focused on a single department or professional group, or where there is a limited set of changes required to working practice. Examples would be the introduction of the national literacy strategy in primary schools, or the national crime recording standard in the police forces.
Managing change, however, is not just about knowing what to do but how to do it. What seems like a sensible and achievable set of audit recommendations may prove devilish to implement in practice. Initial momentum generated by the audit can be slowed to a halt by a series of implementation obstacles; for example failure to reach agreement on what to do between agencies or professional groups, the emergence of competing local priorities, or the loss of key members of staff.
Local support model
The local support model is similar to the local audit model in that it also uses a national study to identify best practice and performance variation. The key difference is that local providers receive support from an expert team during implementation There is no one-off, in-depth local audit, but continuous performance monitoring, feedback and bursts of analysis over a longer period. The team supports by identifying specific problems and designing effective corrective action. Through its continuing involvement it works to persuade reluctant providers of the need for change.
This model is suited to change initiatives where implementation is likely to be particularly complex, for example where the change requires action by more than one agency. Here the support team can act as an independent third party to broker agreements and reduce inter-agency conflict.
The support team monitors performance throughout the implementation phase so it can reallocate resources such that they remain focused on those providers that are having the biggest impact on the national picture. The support team is also able to devise solutions to problems that have emerged since the policy launch – changes to policy, developments in adjacent areas and difficulties that were not predicted prior to full-scale implementation.
In this model central resources are thinly spread – geographically and over time – so complex subject areas that require multi-faceted and detailed diagnosis in each locality will be unsuitable for this approach.
Decentralised model
In the decentralised model, the centre sets out the broad policy outcomes it wishes to see, and then leaves local service providers to decide both how to sub-divide those outcomes and how to deliver. Unlike the centre out model, the decentralized model does not break down broad policy outcomes into specific ways of working or process targets. So, for example, rather than requesting criminal justice agencies to increase the number of crimes detected by 10%, it asks agencies to reduce overall crime by 10%. Local providers can then decide whether to achieve this by increasing detections, or whether they would be more successful focusing on crime prevention or improved criminal rehabilitation programmes. This hands-off approach relies on the motivation and skills of professionals working in accordance to a set of values shared with the centre.
Local enterprise is facilitated through the provision of knowledge management systems (often web-based), which enables providers to share approaches and compare results. These may be provided by the centre or through professional bodies. No notion of best practice is uncovered and universally applied. Instead, local initiative is encouraged to generate innovative and effective service responses. Professional staff are freed from excessive and intrusive central intervention, and are able to fully use their skills to develop bespoke solutions.
Whilst there are some clear advantages to using the decentralised model, an inevitable consequence of encouraging local responses will be a variation in methods and outcomes – “postcode” service provision – which may not be acceptable to the public or politicians.
Increasing the chances of success
What should a central government department consider when embarking on a major change initiative? Failure to plan a clear and convincing overall approach is frequently at the heart of the policy-implementation gap. To introduce change successfully, departments should select an approach or combination of approaches from a broader repertoire than just variants on the ‘centre-out’ model. In deciding which model(s) to use civil servants should consider scaling the challenge. The task of identifying the right change approach needs insight into the practical realities of service delivery. The assessment of complexity should include: extent of the change required; timescale; number of geographical areas; number of agencies and individuals; number and strength of external uncontrollable factors; potential resistance to change; local skill levels; and available resources.
Different models may be more appropriate at different stages of a change initiative, or in different localities due to factors such as managerial motivation and competence, and the scale of problems faced. The centre needs to recognise this, be flexible in its approach and be prepared to vary their approach as circumstances change.
The demands each approach makes on resources – both central and local – should not be ignored. The centre must ensure that local areas are adequately resourced and that there are sufficient central resources to manage and support the change. This is both a qualitative and a quantitative issue.
Finally, consistency of goals will be a key factor in determining success. Change will be made much more difficult if objectives are continually added to and refined. Local managers need to be given a relatively small number of change priorities and be allowed to focus on them. Fewer targets and objectives, means more achievement.
