Lucy Gilson, Jane Doherty, Di McIntyre, Stephen Thomas, Vishal Brijlal, Chris Bowa and Sandi Mbatsha

Briefing Paper Centre for Health Policy, University of the Witwatersrand
Health Economics Unit, University of Cape Town
The SAZA Project
The report, The Dynamics of Policy Change, presents the findings of a study of health financing reforms in South Africa during the period 1994-1999 - the first term of the country’s first democratic government. The study was undertaken by the Centre for Health Policy, University of the Witwatersrand and the Health Economics Unit, University of Cape Town. It is part of a two-country project called Analysing the process of health sector reform in South Africa and Zambia (also known as ‘the SAZA project’).

The need for such a study
Health financing reforms that aim to improve resource availability and use are a central component of the current wave of health sector reforms - both in sub-Saharan Africa and in other parts of the world. However, there has been little systematic evaluation of reform experience in any country. This study was initiated to gain a better understanding of the process of developing and implementing such reforms and so to provide information which can support policy makers and planners – not only in the countries of focus but also around the world.

A focus on South Africa and Zambia
The study was undertaken in South Africa and Zambia, two sub-Saharan countries which have introduced wide-ranging health financing changes in recent years. The experience of these countries is expected to be of relevance to other countries within the region and around the world.

The unique features of SAZA
The study has three important features that distinguish it from other research on health reforms in middle and low-income countries:
1. The major contribution of the study is its emphasis on the process by which policies are developed and implemented, and the factors facilitating or constraining their impact (where ‘process’ covers the steps of policy change, their timing and the strategies used within these steps, for example to build legitimacy, consensus or capacity, as well as the specific mechanisms or bodies established to take forward any of the steps);
2. To ensure a comprehensive understanding of the reforms and how they relate to other relevant policy changes, the study has also considered the linkages between different financing reforms, and links between financing reforms and other health sector reforms (particularly decentralisation);
3. The study focussed on the issues of equity and health system sustainability, issues that have been subjected to less scrutiny internationally than, for example, efficiency.

The key reforms studied
The study considered the following range of reforms:
· geographic resource allocation formulae;
· user fees (in South Africa the removal of primary care fees and in Zambia the implementation of a full fee schedule);
· health insurance options (in South Africa, formal social health insurance and in Zambia, less formal, pre-payment schemes).

The research team
A multi-disciplinary research team of health economists and health policy analysts conducted the study, linking research institutions in two African and two European countries. The institutions are: The Centre for Health Policy, University of the Witwatersrand; Department of Economics, University of Zambia; The Health Economics Unit, University of Cape Town; Health Policy Unit, London School of Hygiene and Tropical Medicine (UK); Institute of Health Economics (Sweden).

This Briefing Paper presents a summary of the report of the
South African study of the SAZA project (Analysing the process of health sector reform in South Africa and Zambia)
We are sincerely grateful to the many people who have given us their time during the course of this project, agreed to be interviewed, lent us documentation and provided us with detailed feed-back on our analyses.

We would particularly like to thank the following reviewers of the draft report:

South Africa
Dr Jonathan Broomberg
Dr Jud Cornell
Dr Nicholas Crisp
Dr Steven Friedman
Mr T. Patrick Masobe
Prof. William Pick
Dr Yogan Pillay
Dr Malcolm Segall


Dr Sara Bennett
Dr Charlotte Leighton
Dr Gill Walt

Finally, we thank our colleagues from the University of Zambia, the Swedish Institute for Health Economics, and the London School of Hygiene and Tropical Medicine for their collaboration in undertaking this study.

The authors, of course, take full responsibility for the report and the final conclusions.

The SAZA Project was funded by:

· USAID’s Partnerships for Health Reform Project (through the Major Applied Research grants programme), Abt Associates Inc.
(contract number: HRN-5974-C-00-5024-00)

· The European Union (through INCO-DEV research funding, contract number: ERBIC18-CT97-0218)

· The South African Medical Research Council (through the Centre for Health
Policy’s MRC Research Group on Health Policy)
Lucy Gilson is a part-time member of the Health Economics and Financing Programme of the London School of Hygiene and Tropical Medicine, UK which receives funding from the Department for International Development, UK.
Towards improving policy development
In 1994 the first democratically elected South African government took office with a mandate to undertake wide-ranging and radical change to redress the apartheid legacy of inequity and inefficiency. Although much has been done, much inevitably remains to be done in order to ensure "delivery to the poorest of the poor" in all sectors, including health.

Five years on, what can be learned from the process of new policy development? What were the obstacles? Are the policies having the expected impacts? How can we draw on the experience of the past five years to effect positive change for the future? These are just some of the questions the SAZA project sought to address through its study of policy change in the area of health care financing.

Health care financing issues are central to health system development. They include the level of funding available for health care provision and how the funds are distributed through the health system. Health care financing policies strongly influence both the provision and use of health care services. They therefore have important consequences for the quality of care, the efficiency of resource use and the equity of the health system. Not surprisingly, Health Minister Tshabalala-Msimang has included financing-related issues, such as social health insurance and increasing the number of ‘fee paying’ patients in public hospitals, within the health sector’s top ten priorities for 1999-2004.

A key foundation for developing future policy change is evaluation of past experience. This study looked back at what has taken place in the area of health care financing reform. The report draws out some valuable lessons and insights that can inform future policy development. It specifically focuses on the period between 1994 and 1999 – a time during which the government initiated a radical overhaul of the health system. The report suggests that alongside some significant achievements there were also some setbacks and lost opportunities. It seeks to explain the strengths and weaknesses of financing policy making as a basis for determining the next steps of policy action.

The importance of process
Unlike most evaluations, which are limited to issues of content, i.e. ‘what’ the policy says, this study also considered the process of policy development. It recognised that the impacts of new policies are not only influenced by their content, but also by the way the policies are developed, who is involved in the process and how the policy will be implemented.

The report considers the influence of various actors (role players) on the design and impact of reforms, and the nature of their agendas. It highlights the different ways in which specific policies have been developed, drawing out the characteristics of the processes. Key contextual events and trends are also explored to help explain policy decisions.
Through such analysis, the report demonstrates that design problems and implementation failures are often rooted in weaknesses in the process of policy making. Action to tackle the equity and sustainability problems within the health system must include not only specific pieces of analysis and design work, but also clear steps to strengthen policy making around health care financing. This requires consideration of issues such as how policy design influences actors’ support or opposition for reforms, the mechanisms and strategies to use in engaging different actors, and when to involve and work with the various actors. The impact and evolution of policy can never be disconnected from the manner in which it is developed.

The report concludes that improved delivery of health services requires not only good policies or action to strengthen implementation capacity within the public service, but also development of policy analysis skills to manage the process of health system change. In effect, the development of such skills is the critical foundation for improved service delivery. "The gap between policy and implementation … really lies with shortcomings in policy-making."
How the study was carried out
The study examined the three areas of financing policy change over the 1994-99 period:
1. the removal of fees for both pregnant women and children under six, and general primary care;
2. the re-allocation of government budgets between provinces;
3. the development of proposals for social health insurance (SHI) – these also had links to proposals to strengthen the management of public hospitals as well as the introduction of new legislation to regulate the private insurance industry.

Although a simplification of reality, the conceptual framework used to guide the study’s data collection and analysis assumed a linear process of policy change, moving from agenda-setting around a reform of focus, to reform design, and then through implementation to the achievement of immediate and longer-term changes.

The framework’s primary focus was on the detailed investigation of factors that influence the overall process of policy formulation and implementation, and so, ultimately, shape the nature and extent of change achieved by the reform. Therefore, for each reform area, the study examined the ‘design’ of the policies, the way in which the policies were developed and implemented, and the key people who were involved in the process. Available evidence about the impact of fees and budget reallocations on aspects of equity and health system sustainability were also assessed.

Information was drawn from reviews of relevant documentation, broader literature concerning the reform areas and the context of South African policy development, newspaper analyses of health issues and relevant parliamentary speeches. Detailed interviews were conducted with key informants from various levels and sectors of government and from non-governmental organisations (NGOs). Finally, evidence from existing evaluations as well as some additional budget and expenditure data were used in the assessment of policy impact.
The analysis of information from these different sources involved a process of interpretation – and this was undertaken with great care. A first step was to compare and contrast the information from different sources in order to seek confirmation of views and opinions as well as to identify different perspectives. In addition, a first draft of the report was reviewed by a range of key informants, and their feedback was considered when finalising the report.

A rigorous process is always important in analysing qualitative information. It was critical in this study because both the institutions undertaking the research had been involved, at some time or another, in the policy development processes being studied. Another factor that may have influenced the analysis was the difficulty of accessing some key actors. Also, the focus on financing issues led to examination of particular sets of actors and processes, some of which were specific to the arena of health care financing reform. Therefore, analysis of these specific experiences may not always be transferable to broader health policy development.
Financing flows within the health system in 1994
The structure and problems of the health system inherited by the new government inevitably shaped the nature of health financing reform after 1994. Figure 1 provides a diagrammatic representation of that system. The figure highlights the main funding sources and channels as well as the division of care between the private sector - serving the minority, higher income, predominantly white population - and the public sector - serving the majority lower income, predominantly African population.

The financing flows of the public health sector include:
· the collection of taxes (to fund all public services including health care);
· the collection of fees at health care facilities;
· the allocation of budgetary resources between geographical areas and the different levels of care provided by the health system.

The two most important flows, in terms of the amount of resources being channelled through them, are tax collection and budget allocation to the health system.

In the private health sector (as inherited in 1994) high-income and some middle-income members of the population and their employers make contributions to voluntary insurance schemes known as medical aid schemes. These are the primary source of revenue for private health care providers, providing funding to them through different forms of reimbursement mechanisms. The insured also make co-payments directly to the providers of care, but these co-payments are smaller than the insurance premia. (Since the early 1990s a range of other forms of health insurance products have also developed.)

Figure 1 also highlights two other financial flows within the overall health system. First, the higher income insured population benefits from tax subsidies on its medical insurance contributions. Second, there is a relatively small financial flow from the private insurance sector to public facilities when insured patients use these facilities. Such use may result, for example, from the need for specialist services only available in the public sector or from the lack of private hospital facilities in some small towns.
[PMG Ed Note: Figure 1 not included]

Context and pattern of financing policy change
The ability of the new government to deal with the legacy of the apartheid health care system was shaped by contextual factors. These need to be considered to appreciate the extent of subsequent reform.

The reform of the government required a complete reconceptualisation and reorganisation of the state and its relationship with society. In particular, the unification of previous administrations and the restructuring of government into a quasi-federal arrangement was an enormous and complex task. In parallel, the new government catapulted activists from the liberation movements into positions of power as politicians and civil servants. There they were confronted with an outmoded, inefficient and rule-bound bureaucracy subject to contradictory decentralising and centralising processes.

The emergence of new patterns of intergovernmental relations, and the special place allocated to the private sector through the negotiation process, introduced a variety of new players to the process of policy formulation.

The limited resource base compounded the problems produced by such structural changes. Previous patterns of underdevelopment as well as new economic policies had the effect of limiting government taxation and spending. This constraint on budgetary resources was accompanied by strictures on human resource redeployment, as well as pressure to increase salaries through central bargaining arrangements.

Progress made towards implementing new policies
Despite the problems and constraints, the new government made strong moves towards re-orienting service provision towards the needs of the majority of the population and away from those of historically powerful interest groups, such as the urban wealthy.

To remove financial barriers to care, and so improve access, two free care policies were implemented (action 1 in Figure 1). In addition, budgetary resources were reallocated between geographic areas to promote equity in their distribution (action 2, Figure 1). Initially, a needs-based formula developed by the Department of Health was used to determine provincial health department allocations. This approach was, however, overtaken by the introduction of a fiscal federal system with the finalisation of the new Constitution in 1996. Since then the Department of Finance has determined the total block grant allocated to each province. Provincial administrations then allocate resources to each sector, including health.

Two other critical areas of financing policy debate between 1994 and 1999 were:
· The level of cross-subsidies from the private to public sector (action 3, Figure 1): The SHI proposals developed over the 1994-99 period sought to extend insurance coverage to a larger proportion of the population (particularly to lower income but formally employed people). Through this extension they sought to generate revenue through and for public hospitals by charging higher fees to the subsequent, larger pool of insured users of the public sector. Some of the proposals also sought to promote greater cross-subsidisation between the high and low-income earners and between the private and the public health sectors.

· Legislative change to re-regulate the private insurance industry (action 4, Figure 1): This was to tackle two of the industry’s critical problems - spiralling costs and the exclusion of certain groups from coverage. (This area of policy action was not an initial focus of the study but came to be considered as a spin-off of SHI debates.)

By 1997 the development of SHI proposals had become divorced from the development of private insurance regulation proposals, although these were initially seen as part of the same policy. The regulation proposals became law in the 1998 Medical Schemes Act. However, by 1999 the broader SHI proposals had still not been implemented despite repeated and intense debate. It remains unclear whether the most recent set of proposals, published in 1997, have sufficient backing from key actors to be implemented effectively. Moreover, only slow progress had been made by 1999 in strengthening public hospital management – an important first step towards SHI.

The process followed
All the health care financing reforms discussed above, had their roots in health policy debates amongst the ‘progressive health movement’ in the late 1980s. The debates then fed into the African National Congress (ANC) Health Plan, published in 1994, which included proposals on all of the financing reforms examined in this study. However, after 1994, the experience of taking forward the reforms differed in three important ways:

· the speed with which progress was made: the speedy ‘policy actions’ of creating a health sector resource re-allocation formula and the removal of some public care fees contrasted with the slow progress made on public hospital fees and with the uneven process of ‘agenda-setting’ for social health insurance;

· the implementation process: although initial policy action on both fees and resource reallocation occurred speedily, the way in which these actions were implemented differed – fee removal involved two special ‘one-off’ policy actions, but health resource allocation policy evolved in response to changes in the broader governance pattern and structure of the country (the creation of a quasi-federal system).

· routine vs. ‘special’ policy-making structures: while changes in resource allocation practice were almost immediately implemented through the routine budgetary process, the slow development of SHI policy through various special bodies led to policy proposals but no policy change.
The impact of the reforms and remaining challenges
The financing policy changes introduced between 1994 and 1999 had clear potential to tackle some of the equity and sustainability problems within the health system:
· the two free care policies improved financial access to public care for specific vulnerable groups;
· the use of the health sector resource allocation formula supported geographic reallocations of public sector health budgets in favour of the formerly under-resourced provinces;
· the 1998 Medical Schemes Act has the potential to tackle critical problems within the private insurance industry, such as cost inflation and risk selection, and so to have positive equity and sustainability impacts.

However, these policy actions also had drawbacks. Although the available data are limited, the two free care policies and the parallel budgetary geographic reallocations seem to have had a negative impact on the stability of the health system. Uncertainty in planning, poor morale of providers, declining quality of care and public disaffection with the public health system, are all by-products of reform which create a climate within which it becomes difficult to sustain system improvements. In addition, favourable early trends in health budget reallocations are being jeopardised by the post-1996 process of allocating unconditional block grants to provinces. Poorer provinces have seen initial budget gains reversed or halted. There are also continuing inequities in health resource allocations within provinces.

Finally, social health insurance, a reform ten years in the making, was not implemented. The failure to implement such a complex reform in only five years was, perhaps, inevitable, but the limited progress achieved in simply finalising a proposal that had adequate support to move towards implementation was disappointing. This represents a failure to find, within the context of constrained government expenditure, both an extra-budgetary source of funds and a mechanism to achieve greater cross-subsidisation between the private and the public health sectors. However, such action is essential if equity and sustainability are to be promoted throughout the entire South African health system. The limited action to date in tackling the weaknesses of public hospital funding and management has important implications not only for sustainability within the public health sector but also for the feasibility of introducing SHI.

Perhaps the most important factor of policy design that limited the positive impacts of the different health financing reforms was the weak linkage between financing policy change and other changes in the health system. For example, removing fees does not in itself ensure that the services are either geographically accessible or seen to be of sufficiently high quality to attract people to use them. Although important efforts were made to build new clinics in previously under-served areas, the impact of the speedy removal of fees on staff workloads (due to increasing patient numbers) had knock-on consequences for drug availability and staff morale. These, at least in the short-term, appeared to jeopardise patient perceptions of the quality of care.

Similarly, speedy and substantial budget reallocations between provinces were not turned into the same level of real resource reallocations, in part because the existing civil service regulations (including labour agreements) restricted provincial departments of health from moving personnel between areas and facilities. This also constrained reallocation between levels of care.

In general, during the 1994-99 period, financing policy appears to have taken second place to specific interventions such as drugs policy or abortion. Yet a combined package of institutional and financing reforms is necessary to enable health system change to achieve its intended objectives.
Some explanations …
What explains the mixed picture of health care financing change? The Department of Health has been held up as a department that delivered change – and the removal of fees has sometimes been given as a specific example of its commitment to addressing the legacy of the past. This policy action and the initial establishment of a sectoral resource reallocation formula both generated political capital and provided some equity and sustainability gains. But the picture of financing policy change over the 1994-99 period is flawed both by weaknesses in the design of these policies and by the lack of action on some other critically important issues. As a result, the pattern of change may lead to longer-term equity and sustainability losses.

A central conclusion from the study is that the weaknesses in the overall pattern of policy action and inaction on health care financing issues since 1994 reflects problems with the underlying policy-making processes. Although policy design influences impacts, the roots of implementation failure – both the manner in which some actions have been implemented and the failure to implement others – lie in these problems of process.

Fee removal
The rapid, top-down implementation of the fee removal policies and the health resource allocation formula, for example, captured a window of political opportunity, but prevented the policies from being implemented carefully in consultation with those implementing them and in ways that would promote sustainability. Indeed, the reactions of health providers to these policies only heightened the low morale of health providers. The strong political leadership shown in initiating policy change represented, at the same time, poor leadership for implementation.

Social health insurance
Speedy fee removal shaped the nature of subsequent SHI proposals by limiting the SHI benefit package to hospital care. Such a package may encourage inefficient use of the health system and does not address the concerns of some key actors. The trade unions, for example, appear to have sought private primary care access for their members. SHI policy development was also a site of recurring disagreement between the technicians (who even before 1994 had begun to develop their ideas on possible design options) and the new policy makers who had little grounding in the earlier debates. (The word ‘technicians’ refers to those inside and outside government with health economics knowledge.)

By allowing differential care between the insured and the uninsured, even if only in hotel-like amenities, and by incorporating a role for the private sector in health care administration/provision, several of the sets of SHI proposals sowed the seeds of opposition to them. The new policy-makers and their allies, the trade unions, simply did not accept that a system based on these lines was appropriate, whatever the technical arguments in its favour. As the new policy-makers sought ‘equal access for all’, in reaction to the discrimination of the apartheid era, the proposals simply made no sense to them. At the same time, the proposals never persuaded the newly empowered Department of Finance to set aside its opposition to any form of earmarked tax.

Yet the technicians repeatedly developed policy designs that had limited support, predominantly from the private insurance industry. The new policy makers, on the other hand, repeatedly brought the technicians into policy development through special committees without clearly engaging with their arguments. The process of policy formulation remained a largely technical affair that did not allow broader debate of either policy goals and relevant strategies, or the concerns of different interested, or potentially interested, actors. The end result was a stalemate in policy development.

Geographic reallocation of resources
The initial health sector resource allocation approach was overtaken by the introduction of new budgetary processes in line with the creation of a quasi-federal governance system. Under the new system provincial treasuries have final responsibility for allocating resources to the health sector. But the highly political nature of this process has generated severe concerns for health budgets in some provinces and has required a new approach to resource allocation policy development. Provincial departments of health, in collaboration with the National Department of Health, sought to adopt a collective planning approach in thinking through the budgetary needs of the health sector, and had some success in developing conditional grants to protect aspects of provincial health budgets. However, although they initially sought to protect primary care, they had to accept the Department of Finance’s preference to protect high-level hospital care through these grants. They also had not, by 1999, implemented the ‘norms and standards’ that can influence budget decisions within provinces. Interestingly, although the technical complexity of this area of policy development mirrors that of SHI policy development, limited health economics expertise was drawn on to support health policy makers in this task.

Common problems
Across the three areas of health care financing reform investigated in this study, the most critical problems of policy making were:

· the limited public debate about the appropriate and feasible goals for the health system in the post apartheid era, particularly in relation to equity, and strategies for achieving them;
· little recognition of the importance of health financing policies in shaping the nature of health care provision and its impact on equity and sustainability;
· a combination of strong political leadership with weak structures and processes for providing technical advice on complex issues;
· the limited availability of relevant technical expertise within the country, and weaknesses in the ways in which those with expertise were used in policy development;
· the failure of reformers to engage a broad range of actors in discussion on some of the most fundamental issues of financing policy development, including the link between policy design and actor support or opposition for policy change;
· inappropriate patterns of engagement and consultation with different actors within specific policy processes;
· inadequate consideration of implementation needs and strategies within both policy design and the process of policy formulation;
· the lack of preparation for policy implementation, such as limited action taken to develop the skills and systems required to allow effective implementation of financing (or other) policy;
· poor leadership of implementation, including a failure to set clear priorities for policy development, to build support before implementation, and to establish mechanisms for learning through experience.

Financing reforms are, of course, always highly contested and so difficult to implement. In addition, the radical transformation of the government and political bureaucracy since 1994 constrained all policy action. Nevertheless, through their actions key actors influenced the processes of policy formulation and the policies generated through them. These actors’ experience of managing policy change in a changing policy environment offers lessons for ‘how to do things differently’ in the future.
Lessons for the future
1. Strengthening policy formulation as a foundation for implementation

(a) Supporting leadership by providing technical analysis
Senior health policy makers exercised considerable personal influence over decision making in South Africa between 1994 and 1999, in part because it was a time of rapid transformation within the health and governance systems. At the same time, the structures for channelling information and advice to these policy makers on health economics issues were quite weak. Perhaps as a result, health care financing policy appeared to receive less attention than other aspects of health policy development despite its importance to health system change. Aspects of priority setting and design development for health care financing were also weak. To strengthen future decision making it may be important to review the existing mechanisms of providing technical support to health policy makers and to establish procedures that enable regular contact with technical advisers. Opening up some policy issues, including overall health system goals and particular policy goals, for broader debate might also allow a timely flow of relevant information to decision makers.

It is particularly important to strengthen the National Department of Health’s Directorate of Health Financing and Economics by improving its access to the highest level of government and promoting systematic dialogue between it and key reform managers and policy makers at both national and provincial levels. To make best use of the small pool of health economists available within the country, it would be useful to clarify research needs for particular issues as well as the objectives and nature of any involvement of external analysts. Capacity shortfalls in the area of health economics can also be addressed through long-term partnerships between government and key training institutions. However, the independence of non-government groups is an important characteristic of their particular contribution to policy debates.

Example: Policy-makers require good technical support in relation to the establishment and use of norms and standards to influence provincial health resource allocations. These norms must be compatible with national affordability criteria, so that provincial treasuries can realistically be expected to fund the services from their unconditional block grants. Consideration must also be given to the services to be covered by the norms. Specific attention needs to be given to maintaining the secondary referral services that are critical in the effective provision of primary care, but that may be overlooked in policy development. Technical analysis is also important in tackling intra-provincial health budget inequities.

(b) Strengthening the strategic skills and awareness of technicians
It is vital for government technicians, as well as non-government policy analysts, to recognise the importance of strategy as a complement to technical analysis. The 1994-99 experience shows, for example, that where the technical design of a financing policy matched the values of policy elites, and the broader political goals with which they were
associated, there were fewer barriers to implementation. Political buy-in was a necessary pre-condition for further policy development. Understanding the power, value bases and concerns of major actors in the health sector is a first step in managing the process of policy change. It provides the basis for developing strategies that create alliances in support of reforms and offset opposition. It also allows the actors’ interests to be taken into account in the design of the reforms. In addition, as inappropriate presentation of technical inputs can impede reform, consideration of the dominant values of key political actors will help analysts appropriately frame their policy inputs. Complex reforms anyway need to be communicated clearly and simply both to policy makers and the broader public. This need not sacrifice technical detail.

Technicians need to have the knowledge and skills to develop effective strategies of policy change. The particular role of special committees, for example, needs to be thought through carefully. They can help strengthen policy formulation, for example, by protecting the task from the administrative burden of government, supplementing government’s skill base, or allowing broader representation of key actors. However, if planned poorly, they may also frustrate policy development.

Example: Technical and political analyses need to be combined to clarify the possible alternatives for a future social health insurance system. Five key steps that will provide a foundation for further SHI policy development are:

1. undertaking a comprehensive analysis of the extent of cross-subsidisation within existing health care financing and spending patterns;
2. reconsidering the motivations and objectives of cross-subsidisation in health care financing and spending within the context of South Africa’s social objectives, in part through dialogue with key actors;
3. analysing the extent to which alternative SHI proposals achieve the desirable level of cross-subsidisation;
4. considering the acceptability of new proposals (and other options) to key actors, and the risks associated with implementing the reform as presently configured;
5. reconsidering the separation of SHI from the conventional medical schemes environment and the nature of the benefits to be covered through SHI (in part to ensure that they offer members advantages over current services).

(c) Building implementation concerns into design development
As happens in many countries, the development and design of the health care financing policies did not involve those who would be responsible for implementing the policies. The consequences of this included poor implementation practices and negative impacts of policy, as noted in relation to the free care policies. Involving representatives of mid-level health system managers and health care providers as advisers in policy formulation may be important in strengthening the eventual implementation of reforms. In addition, ‘policy champions’ within government should be identified to enable the work of special policy processes to be fed forward into implementation.

2. Strengthening implementation processes
(a) Working within a changing policy environment
The structural change and institutional flux during the government’s first term of office undermined the capacity for both policy development and policy implementation, so weakening the impact of the reforms. In a changing structural environment it is important to recognise the costs of ‘trying to do too much too quickly’, even whilst accepting that problems demand urgent action. When institutions are being reformed and new lines of management, roles and responsibilities are being developed, the priorities for policy action need to respond to concrete and identifiable problems as well as building political support for a broader reform agenda. In addition, these priorities must be rooted in clear analysis and understanding of the key health problems, how the reforms could address such problems and the sequencing of policy actions needed to support the implementation of complex reform. Enhancing the institutional capacity for future policy development and implementation is also likely to be an important building block for continuing reform.

(b) Enabling implementation through leadership
It is critical to improve central co-ordination and facilitation of policy implementation. It is particularly important that the policy-making process allows for more dialogue and consultation with those who will implement policy. This requires a clear specification of roles and responsibilities between the tiers of government, and structures for co-ordinating action between managers at the different levels of government. Strengthening skills and systems for implementation is also important. More gradual implementation processes can enable such capacity development.

Example: Although some technical analysis is needed in conceptualising norms and standards, their effective implementation will require a well-managed process. It will be particularly important to build support for the policy with provincial treasuries, as well as with medical professionals and health system managers.

(c) Planning for implementation
Although it may be important to take advantage of ‘windows of opportunity’ to introduce policy change, rapid change at such times can generate problems for further reform. Important steps in planning for implementation always include identifying the potential obstacles to change, preparing guidelines to support change, and identifying the capacity needs to implement change. The gradual implementation of reforms can also allow some details of policy design to be determined through experience. Complex reforms may need to be broken down into phases or stages to enable their implementation, particularly where they involve establishing new institutions. Such stages also need to be planned in advance, as the policy environment may change and leave the complex reform only partly implemented. Reformers must ask whether the individual steps toward the reform meet objectives in themselves. If not, then the cost or risk of failing to implement all the steps must be considered beforehand.

Example: The implementation of SHI must be preceded by improved hospital billing and fee systems, revenue retention as part of enhanced decentralised hospital management, and a mechanism for ensuring the equitable distribution of all resources available to public sector services. Similarly, the effective implementation of any mechanisms for influencing resource distribution between and within provinces requires that budgetary changes are accompanied by new approaches to personnel decision making, and that capital and recurrent budgeting is more closely linked.

(d) Securing better policies through monitoring and evaluation
Monitoring and evaluation systems are vital to ensure that reforms are adapted over time to meet their objectives. They should allow assessment of the progress made towards the objectives and of the factors that influence the degree of progress (i.e. assessment of both what has been achieved and what has constrained or encouraged the progress). The views of the broader public should always be considered in evaluation. Existing information systems may need to be strengthened to provide regular data on key indicators. Non-government analysts can play a role in designing effective evaluation systems and in implementing some evaluation activities.

Examples: There is an opportunity to prepare for the evaluation of the 1998 Medical Schemes Act in advance of its implementation, and to keep a watching eye over its equity and sustainability impacts. It is always important to monitor the impact on geographic equity and level-of-care spending patterns of resource allocation policies. Such assessment of the existing conditional grants, as well as any new policies developed in the future, will be critical in informing further policy development. Equity could also be introduced as a key criterion for monitoring the implementation of the government’s medium-term expenditure framework.

3. Towards delivery - strengthening the policy process
Finally, the study highlighted six key principles to guide the development of a policy process that will lead from new policy formulation to change in health care delivery practices. They are:

1. Financing reform should pay attention to the ‘art’ of politics rather than just the ‘science’ of technical analysis
- to enable change and to ensure that it does not become the preserve of the few with the relevant knowledge.

2. Financing reform should be placed at the heart of health system development
because it has a wide-ranging influence over the patterns of health care
provision and use, and because it must be supported by parallel institutional changes.

3. Financing policy should be developed through a relatively open and transparent process
- to allow broader, public debate about the goals and strategies of policy; however, ‘closed’ decision making may be useful to identify policy options on the basis of publicly debated goals or to develop detailed design proposals in relation to specified options.

4. Information is a critical element in financing policy development
- both formal data and, despite their informality, the views and opinions of the
public and key actors.

5. As technical analysis is important in supporting financing policy development, the roles of different groups of technicians and analysts must be clear
- to enable best use of the limited pool of available health economists.

6. Implementation should be an integral element of financing policy development
- rather than being seen as an activity that somehow automatically follows policy development and that does not require policy management skills.
Overall, the study emphasises that in implementation as in policy formulation:
"policy analysts cannot continue to ignore the how of policy reform" .