Abstract

The public social policy and programme decisions that are made in low-income countries have critical effects on human social and development outcomes. Unfortunately, it would appear that inadequate attention is paid to analysing, understanding and factoring into attempts to reshape or change policy, the complex historical, social, cultural, economic, political, organizational and institutional context; actor interests, experiences, positions and agendas; and policy development processes that influence policy and programme choices. Yet these can be just as critical as the availability of research or other evidence in influencing decision making on policies and their accompanying programmes and the resulting degree of success or failure in achieving the original objectives. Ghana, a low-income developing country in sub-Saharan Africa, embarked on a national policy process of replacing out-of-pocket fees at point of service use with national health insurance in 2001. This paper uses a case study approach to describe and reflect on the complex interactions of context with actors and processes including political power play; and the effects on agenda setting, decision making and policy and programme content. This case study supports observations from the literature that although availability of evidence is critical, major public social policy and programme content can be heavily influenced by factors other than the availability or non-availability of evidence to inform content decision making. In the low-income developing country context there can be imbalances of policy decision-making power related to strong and dominant political actors combined with weak civil society engagement, accountability systems and technical analyst power and position. Efforts at major reform need to consider and address these issues alongside efforts to provide evidence for content decision-making. Without an analysis and understanding of the politics of reform and how to work within it, researchers and other technical actors may find their information to support reform is not applied effectively. Similarly, without an appreciation of the need for critical technical analysis to support decision making rather than an indiscriminate use of political approaches, political actors may find that even with the best of intentions, desired policy objectives may not be attained.

KEY MESSAGES

  • Though social policy reform has technical as well as political challenges, inadequate attention is paid to recognizing, analysing and dealing effectively with the political challenges.

  • Other low-income countries can learn from Ghana's national health insurance policy and programme development and implementation, where despite high-level political commitment, popularity of the proposed reform and availability of information on the technical challenges, limited understanding and management of the political challenges resulted in potentially avoidable difficulties.

  • There is a need to promote better dissemination, understanding and use of analytical frameworks on the political economy of reform in developing countries to assist reformers to manoeuvre within the challenges of the environmental context, agenda-setting circumstances and policy characteristics of reform.

Introduction

“Would you tell me, please, which way I ought to walk from here?”

“That depends a good deal on where you want to get to,” said the Cat.

“I don't much care where,” said Alice.

“Then it doesn't matter which way you walk,” said the Cat.

“– so long as I get somewhere,” Alice added as an explanation.

“Oh, you’re sure to do that,” said the Cat, “if you only walk long enough!”

    (From Alice in Wonderland by Lewis Carroll)

Sometimes those interested in the effective development and implementation of health reform in developing countries appear to think that focusing on technical content generation, advice and support will lead to the desired endpoint. Inadequate attention is paid to how this content is going to be adopted and implemented within the national and international context. Yet observation suggests that reform is political as well as technical. In the words of Walt and Gilson (1994): ‘The traditional focus on the content of policy neglects the other dimensions of process, actors and context which can make the difference between effective and ineffective policy choice and implementation’. Glassman et al. (1999), in a political analysis of health reform in the Dominican Republic, comment: ‘the process of health sector reform involves a continual tension between the technical and political dimensions’. Thomas and Gilson (2004), in their paper on actor management in the development of health financing reform in South Africa, make similar observations; and Cassells (1995) reflects: ‘health reform is inherently political’. Among the findings of the SAZA study (Gilson et al.2003), which explored health reform policy-making processes in South Africa and Zambia, was the strong influence of political factors and actors over which health care financing policies were implemented and which were not, as well as the details of policy design.

The work of Grindle and Thomas (1991) on the political economy of reform in developing countries is one from which several other authors quote. Based on an analysis of policy and organizational reform in recent decades from several developing countries, they propose a multivariate framework for understanding the emergence, discussion, implementation and sustainability of policy reform in developing countries. They focus especially on the role of policy elites in shaping policy agendas, weighing policy options and managing political and bureaucratic challenges of policy reform. By policy elites, they refer to ‘those who have official positions in government and whose responsibilities include making or participating in making and implementing authoritative decisions for society’. Key factors in their framework are the environmental context of reform, the agenda-setting circumstances and the policy characteristics.

The environmental context includes the individual characteristics of policy elites such as their ideological predispositions, professional expertise and training, memories of similar policy situations, position and power resources, political and institutional commitments, loyalties and personal attributes and goals. It also includes the context of policy choice such as societal pressures, historical, economic and international context, administrative capacity and other policies. The agenda-setting circumstances include whether there is a perception of a crisis situation or not. In a perceived crisis situation there is strong pressure for reform and immediate action, the stakes are considered high, change is seen as innovative and high-level decision-makers are involved in the reform process. On the other hand, where there is no perception of a crisis situation, the agenda-setting circumstances are ‘politics as usual’. The reform issue is one that is chosen by policy makers rather than pushed onto the agenda. The stakes are seen as low and decisions are often left to middle level decision-makers, incremental change is acceptable and there is flexibility in timing. Decision making in perceived crisis situations tends to be dominated by concerns about macro political relationships whereas in politics as usual, policy decision making tends to be dominated by concerns about micro political and bureaucratic relationships. Policy characteristics include whether arenas of conflict in reform are public or bureaucratic; and the resources for implementation and sustainability. In perceived crisis situations, the arenas of conflict tend to be public. Conflict can be highly and immediately visible and may have a direct impact on large sectors. The stakes are such that the legitimacy of the political regime, the stability of the regime, the sustainability of the reform and the capacity to pursue other reforms may be called into question. In politics as usual, the arenas of conflict tend to be less public and more bureaucratic.

Walt and Gilson (1994) refer to the work of Grindle and Thomas (1991) in suggesting a more simplified policy analysis model that considers the content of reform, the actors involved in policy reform, the processes contingent on developing and implementing change and the context within which policy is developed. Actors and context in their model overlap elements in the environmental context in the model by Grindle and Thomas (1991). Similarly, there is some overlap between agenda-setting circumstances and policy characteristics in the Grindle and Thomas (1991) model and processes in their model.

Aryee (2000), in his analysis of why public policies and programmes succeed or fail in Ghana and other African countries, also refers to the work of Grindle and Thomas (1991) among others. He focuses on characteristics of the environmental context, especially the role of policy elites, but also on the context of policy choice, and argues: ‘Successful public policies and programs are rare because it is unusual to have progressive and committed politicians and bureaucrats (saints) supported by appropriate policy analysts with available and reliable information (wizards), that manage hostile and apathetic groups (demons) and consequently insulate the policy environment from the vagaries of implementation (systems).’ From his point of view, political and bureaucratic leadership that is prepared to champion a policy, take the risks as they come, and stick to the efforts required to make policies succeed are probably the most important of the actors in determining public policy success or failure. However, they need to be supported by appropriate policy analysts with reliable and available technical content information. They also need the skills and the willingness to recognize and deal with the group of actors he refers to as ‘demons’. In his words:

‘Perhaps the most difficult lesson to learn, much less accept in public policies and programs, as in religion, is that demons do exist and can be very destructive. Identifying demons is one of the most difficult but important tasks in public policy analysis … For public policies and programs to succeed, the demons have to be managed and neutralized. Demons are a very small set of public officials and individuals who engage in corruption or rent seeking activities.’

Aryee's (2000) concerns about the influence of the group of actors he calls ‘demons’ on public policies and programmes are related to issues of accountability within health systems. Brinkerhoff (2004) suggests two aspects of accountability—answerability and sanctions. Actors need to be answerable in terms of provision of information on their actions and justification for their actions to someone or some groups who have available sanctions they are empowered to apply for illegal or inappropriate actions and behaviour uncovered through answerability.

There is an increasing body of literature on the technical challenges of implementing health insurance in low- and middle-income countries (e.g. Carrin 2003; Carrin and James 2004; Preker 2002), but much less on the political challenges. This paper examines the political challenges of reform in the development of national health insurance in Ghana using the framework of Grindle and Thomas (1991) in analysis but also referring to the related ideas of Walt and Gilson (1994) and Aryee (2000).

Objectives

In 2001, Ghana, a low-income developing country in sub-Saharan Africa, embarked on a process of developing and implementing policy and accompanying programmes for a National Health Insurance Scheme (NHIS) to replace out-of-pocket fees at point of service use as a more equitable and pro-poor health financing policy. As of December 2006, 38% of the approximately 21 million population had registered in the NHIS. Twenty-one per cent (21%) had been issued with ID cards and were effectively protected from out-of-pocket fees at point of service use by the NHIS (NHIC 2007). Government policy objectives in setting up a NHIS are stated in the national health insurance policy framework for Ghana (MOH 2002, 2004) as:

‘Ultimately, the vision of government in instituting a health insurance scheme … is to assure equitable and universal access for all residents of Ghana to an acceptable quality package of essential healthcare. The policy objective is “within the next five years, every resident of Ghana shall belong to a health insurance scheme that adequately covers him or her against the need to pay out of pocket at the point of service use in order to obtain access to a defined package of acceptable quality of health service”.’

Achieving universal health financial protection in a low-income country like Ghana is a laudable ideal, but technically difficult and challenging. There is currently no low-income country that has achieved universal health insurance coverage. With the exception of Thailand, a lower middle-income country, that has achieved virtually universal coverage since 2001 using a big bang approach (Tangcharoensathien and Jongudomsuk 2004), countries which have achieved universal coverage are wealthier nations such as those of Western Europe, Canada, Japan and South Korea. They have done so over the long rather than short term. Related to income, the structure of Ghana's economy, with many citizens employed in the non-formal sector and living in rural communities and small towns with poor road access, telecommunications and health service access, is a major challenge. Despite these challenges, Ghana was determined to try. What was perhaps not well recognized at the beginning of the process was that apart from the technical challenges of the proposed reform, there were challenges related to the politics of social policy reform in a developing country. This paper describes and analyses the experience with these challenges and what can be learned from them.

Methodology

The paper is a qualitative case study of a policy process using a mixture of participant observation and review of public sector, media and historical records and policy documents. Participant observation is based on the authors’ participation as well as observations as actors in the public sector in Ghana for over a decade. Participant observation is a method commonly used in anthropology. In the words of Patton (1990):

‘Experiencing the program as an insider is what necessitates the participant part of participant observation. At the same time, however, there is clearly an observer side to this process. The challenge is to combine participation and observation so as to become capable of understanding the program as an insider while describing the program for outsiders.’

The in-depth personal engagement of a participant observer yields rich and ‘thick’ descriptive material and insights. However, the nature of the investigation that gives this approach its strength also provides its weakness. Given that the observers are participants and actors in the processes observed, described and reflected upon, with positions and opinions on the issues at stake, a possibility of bias in objectivity of description and reflection exists. To give the reader the chance to make their own decisions about the possibility of subjective bias in this paper, the role and position of the authors as participants in the processes and decisions is summarized as an endnote.1 The article also steers away from passing value judgments as to ‘good’ or ‘bad’, ‘success’ or ‘failure’, and focuses on analysis, reflection and learning.

The context of policy choice

Socio-economic and socio-demographic context

Ghana has a gross national income (GNI) per capita (Atlas method) estimated at US$380 (World Development Indicators Database April 2006). It is an agricultural country, with main exports of cocoa, timber and gold. Most of its estimated 21 million population are employed in the non-formal sector, and about half the population is under 15 years. Approximately 44% of the population live in urban localities with a population of 5000 or more. Urbanization varies from a high of 88% in Greater Accra region, where the capital is located, to a low of 16% in Upper East region (Ghana Statistical Services 2002). The 2003 Core Welfare Indicators Questionnaire (CWIQ) survey gave a national adult literacy rate of 53%, with higher male (66%) than female (42%) literacy (Ghana Statistical Services 2003). Youth literacy rates were higher (69%) than adult rates.

Health status and health service access

The 2003 Ghana Demographic and Health Survey (GDHS) gave an infant mortality rate (IMR) of 64/1000 live births, compared with 57 in 1998 and 66 in 1993 (GSS/NMIMR/ORC Macro 2004). Malaria remains the most frequently reported cause of morbidity and a major cause of childhood mortality. Other frequently reported diseases are diarrhoea, acute respiratory infections, skin diseases, pregnancy related complications, anaemia and malnutrition. The most common chronic diseases are hypertension and diabetes (PPME-GHS 2005).

One of the objectives of Ghana's health sector reforms that started implementation in 1997 was to increase geographical access to basic clinical and public health services (MOH 1996). In the CWIQ survey (Ghana Statistical Services 2003), access to clinical services is defined as living within 30 minutes of any kind of modern health facility. Between the 1997 and 2003 CWIQ survey, the percentage of people stating that they had access to a health facility increased from 37% to 58%, while those reporting themselves as sick or injured in the 4 weeks before the survey and needing clinical services remained constant at 18%. Despite this, the reported use of care fell from 22% in the 1997 survey to 18% in the 2003 survey, raising questions as to whether quality of care and ability to pay may sometimes be more important barriers than geographic access.

Political context

Ghana attained independence from British colonial rule in March 1957 under the leadership of Dr Kwame Nkrumah and his Convention Peoples Party (CPP). The CPP adopted far-left socialist politics and eventually declared a one party state. It was ousted by a military coup in 1966 and Ghana subsequently went through three cycles of attempts to establish multiparty democratic governance, with each cycle cut short by a military coup. In 1992, the Provisional National Defence Council (PNDC), which had ruled Ghana since coming to power through a military coup in 1981, allowed multi-party democratic elections. The PNDC re-organized itself into a political party, the National Democratic Congress (NDC), contested and won the December 1992 election as well as the December 1996 multiparty election 4 years later. The PNDC started with far-left socialist politics, but had evolved into a centre-left leaning political party by the time it became the NDC. The NDC lost the December 2000 election to the New Patriotic Party (NPP), the new face of the Danquah-Busia tradition (conservative centre-right leaning political tradition) that had been the party in opposition at independence in 1957 and had briefly ruled the country from 1970 to 1972 before it was ousted by a military coup. The NDC tended to have a stronger following in the more economically depressed northern areas of the country, Volta region and poorer urban communities. The NPP tended to be stronger in the more economically prosperous middle and southern forest belts and urban communities.

One of the major election campaign promises of the NPP in the 2000 election was to replace out-of-pocket fees at point of service use, popularly called ‘cash and carry’, with national health insurance and to assure access to basic clinical services for all Ghanaians regardless of ability to pay. Replacing ‘cash and carry’ with insurance was a popular election promise. The pinch of out-of-pocket fees was felt not only by the poorest, who usually suffer most from regressive taxation, but also by middle-income groups. Even those in the higher income brackets felt the pinch, especially given that traditional extended family structures and social responsibilities are still strong in Ghana. Better off members of the extended family are socially obliged and pressured to provide the safety net for poorer members in financial crisis, such as exposure to sudden unplanned catastrophic health expenditure. In the words of Arhinful (2003), people saw a primary motivation to participate in an NHIS based on ‘the self interest to cater for themselves and their immediate relatives’.

Health financing in Ghana

Prior to independence, financial access to modern health care was predominantly by out-of-pocket payments at point of service use (Arhinful 2003). Following independence, the government switched to tax-based financing of public sector health services and all such services were made free. Private sector health services continued to be paid for by out-of-pocket fees at point of service use. By the early 1970s, general tax revenue in Ghana, with its stagnating economy, could not support a tax-based health financing system. In 1972, very low out-of-pocket fees at point of service use were introduced in the public sector to discourage frivolous use. The stagnation of Ghana's economy was followed by a decline and in the health sector there were widespread shortages of essential medicines, supplies and equipment, and poor quality of care. In 1983, the PNDC government adopted a traditional IMF and World Bank economic recovery programme. In 1985, public sector user fees for health care were raised significantly as part of structural adjustment policies and became known as ‘cash and carry’. The aim of the 1985 user fees was to recover at least 15% of recurrent expenditure for quality improvements. The financial aims were achieved (MOH 2001). Shortages of essential medicines and some supplies improved. However, these achievements were accompanied by inequities in financial access to basic and essential clinical services (Waddington and Enyimayew 1989, 1990).

Historical context of health insurance in Ghana

The first community health insurance (CHI) scheme in Ghana was the Nkoranza health insurance scheme started by the St Theresa's Catholic Mission Hospital in 1992. It proved popular and endured the test of time (Atim and Madjiguene 2000). In the mid 1990s, a unit was created in the Ministry of Health (MOH) to establish national health insurance as an alternative to ‘cash and carry’. The unit focused its efforts and resources on consultancies and feasibility studies for a pilot social health insurance (SHI) scheme for the formal sector and organized groups such as cocoa farmers in the Eastern region. By 1999, the proposed SHI pilot had died a stillbirth without insuring anybody. No public acknowledgement or explanation was given for its demise. However, it appeared to be partly related to lack of leadership, consensus and direction in the MOH as to the way forward; as well as a failure to sufficiently appreciate the difficulties of implementing centralized social health insurance in a low-income developing country (Atim et al.2001; Arhinful 2003). Following the demise of the Eastern region pilot, the social security and national health insurance trust (SSNIT) started planning for another centralized health insurance scheme to be run by a company called the Ghana Health Care Company. Like the Eastern region pilot, it never took off despite some public expenditure on personnel, feasibility and software.

In 1993, UNICEF funded exploratory research on the feasibility of district-wide community health insurance (CHI) for the non-formal sector in Dangme West (Arhin 1995), a purely rural district with a subsistence economy and widespread poverty. The study had strong support from the MOH because the Director of Medical Services was interested in the issues. The study showed enthusiasm among community members for the concept of CHI. A pilot district-wide CHI was planned in the same district with MOH finance for scheme design and implementation, and EU finance for monitoring and evaluation (Community based prepayment and health insurance in rural Burkina Faso and Ghana, EC Number IC 18CT 96–0131). However, the Director of Medical Services retired and with him went MOH interest in the work. No MOH financing was made available and the EU grant for evaluation was not renewed after the initial instalment. Despite the central setbacks, at the local level the district health directorate and research centre, the district assembly (local government) and communities continued their collaboration and completed the design of the pilot district CHI scheme. The district assembly contributed part of its UNDP poverty reduction fund to support community mobilization and household register development, and WHO AFRO and DANIDA provided start-up funding. Registration of beneficiaries and delivery of benefits started in October 2000. Financing to continue implementation and evaluation was provided by the Ghana Health Service (GHS) and the MOH (DWHIS 2002, 2003; Agyepong et al.2006).

Several other CHI schemes, popularly called Mutual Health Organizations (MHO), were also springing up in Ghana. Their rate of development accelerated exponentially after 2001 (Baltussen et al.2006). Many were sponsored by faith-based organizations. Development partners that played a major role in their support were DANIDA (Danish International Development Assistance) and PHR-plus (Partnership for Health Reforms plus), an organization funded by the United States Agency for International Development (USAID). These two organizations also jointly supported the development of a training manual for administrators and governing bodies of MHO (Atim 2000). Many of the MHO were in the Brong Ahafo and Eastern regions, related in part to the fact that in these regions MHO had active ‘champions’ in the form of technically well-informed regional coordinators in the GHS. Also, the Christian Health Association of Ghana (CHAG), represented mainly by the Catholic Church, had many mission facilities in these regions and actively supported the growth of MHO around its facilities. The regional and district directors of health and district assemblies also took an active interest in the development of MHO.

Actors, agenda-setting circumstances and policy characteristics

The pre-legislation phase: 2001–2003

The NPP took over power in January 2001 and in March 2001 the new Minister for Health inaugurated a seven member ministerial health financing task force under the chairmanship of the Director for Policy, Planning, Monitoring and Evaluation (PPME) in the MOH. Members were premised to have some technical knowledge on the subject or to be important stakeholders and were from the MOH, Ghana Health Service (GHS), Dangme West District Health Directorate & Research Centre, Trades Union Congress and the Ghana Health Care company. The terms of reference of the task force were to support and advise the MOH on the development of a NHIS, the building up of systems and capacity for regulation of health insurance in Ghana, the development of appropriate health insurance legislation, and the mobilization of extra resources to support national health insurance.

The Minister for Health was keen on a centralized single payer SHI scheme. Though he was aware of their existence, he did not see MHO as a viable policy option. Many members of the task force, including the Chair did not agree and felt any policy that did not make room for the MHO would not achieve the policy objectives, given the large non-formal sector in Ghana. There was also concern that one of the lessons of the failed Eastern region pilot was the difficulty of implementing a centralized single payer health insurance scheme in Ghana. Conflict developed between the Minister and the Chair of the task force over this and other issues, and the work of the task force stalled as the Chair disengaged from the process because of the conflict. The members of the task force decided, given its importance, to continue the work despite the absence of the Chair. The dilemma as to the way forward was resolved by proposing a hybrid that comprised a classical single payer scheme for the organized formal sector, and multiple payer semi-autonomous MHO for the non-formal sector. Private commercial health insurance would also be allowed for those who felt they could afford it and preferred it. Legislation, a central coordination mechanism in the form of a national health insurance council (NHIC), and a national health insurance fund would be used to ensure equity, re-distribution and cross-subsidization between the multiple funds. Satisfied, in June 2001, the minister endorsed the four-page policy outline developed by the task force outlining these arrangements.

Just before the stakeholder consultation in June 2001 to present this draft policy for debate before finalization, there was a cabinet reshuffle and the Minister for Health who started the process moved to another ministry. The new Minister for Health continued with the process and the first stakeholder forums at national and sub-national level were held. Just before the cabinet reshuffle, the previous Minister for Health had changed the director of PPME and brought in a trusted associate to fill the position. The approach of the new director of PPME involved the gradual introduction of his own trusted associates onto the task force. The political links of the trusted associates were sometimes clearer than their awareness of the technical challenges of NHIS implementation in Ghana. There were increasing tensions on the task force between some new members and the old ones related to differences in opinion about technical proposals. The trusted close political associates increasingly dominated the process of policy and programme development, and inconvenient comments, suggestions and critiques from others appeared, to those who made them, to be ignored rather than analysed. In view of their increasing powerlessness, rather than act as a rubber stamp and endorse proposals they felt could cause implementation problems, original members of the task force resigned or quietly dropped off. There was no formal announcement of dissolution, but by the end of 2002, only one original member of the ministerial task force remained involved in the ‘policy elite’ group taking the final decisions on the NHIS in the MOH.

Ghana took the decision to access the Highly Indebted Poor Country (HIPC) initiative in March 2001 and reached decision point in February 2002 and completion point in July 2004. Areas of expenditure included funding of projects for poverty reduction and economic growth (The Chronicle2006). In February 2003, the MOH allocated HIPC funds to support the creation of government sponsored MHO in all districts where they did not already exist. The trusted political associates were appointed implementation consultants by the MOH to help the district assemblies utilize the money to set up MHO.

In July 2003, one week before parliament was due to go on recess, the final version of the national health insurance bill was placed before parliament under a certificate of urgency to be passed into law. Adverts were placed in the national dailies requesting comments from the general public on the bill before parliament. The original drafts of the bill had been done by the ministerial task force on health financing in 2001 and the earlier drafts discussed at stakeholder meetings in late 2001 and early 2002. After the trusted political associates took control, they did the final drafting and very few people outside this policy elite group had seen the full text of the bill that finally appeared before parliament. A one-week timetable was given for deliberation and passage of the bill into an act.

Organized labour comprising the Civil Servants Association (CSA) and allied groups such as the Ghana National Teachers Association, Ghana Registered Nurses Association, Judicial Services Workers Union, and the Trades Union Congress had shown a lot of interest in national health insurance. The CSA in the Ashanti region had converted its civil servants medical refund scheme into a MHO and the association in Greater Accra and other regions had started planning and organizing for similar purposes. The leadership of these groups convened a meeting, studied the bill and submitted a formal resolution to parliament protesting aspects of the content of the bill and the rushed passage, and requesting a deferment of passage, deeper consultations and amendments. The minority in parliament (the NDC) also raised concerns. In response to these concerns, the debate and passage of the bill was deferred and parliament went on recess.

A month later, in August 2003, parliament was recalled and the bill essentially as it was in July was once more laid before parliament. The protests of organized labour groups and the minority NDC on the failure to address their concerns were ignored. The acrimonious debates that erupted around the passage of the bill were widely covered by the media. The media appeared to focus more on the acrimony than on the content and the rationale behind the dissension over the bill. The debates degenerated into accusations and counter accusations of political motivation to damage the ruling government and prevent the average Ghanaian and the rural poor from having national health insurance. The minority NDC refused to be part of the passage of the bill and walked out of parliament in protest. The majority NPP went ahead and passed the bill since they had the numbers to do so under the requirements of the 1992 constitution. Organized labour groups went on a street demonstration in protest at the passage of the bill into law in spite of their concerns.

What was the content of the bill and why was there so much acrimony over its passage? The bill, which was to become the National Health Insurance Act of 2003 (Republic of Ghana 2003), required the formal and the non-formal sector to enrol together in government-sponsored district MHO. Government sponsorship for the district MHO was automatic and not clearly tied to efficiency and policy effectiveness or responsiveness criteria. All MHO that were not district-wide government-sponsored (public) were classified as private. Private MHO though recognized as not-for-profit solidarity organizations, and legally entitled to operate, would not receive any financial support from the national health insurance fund or any of the subsidies to cover groups exempt from premium payments such as the elderly and the poor.

Financing of the NHIS was to be by individual premium payments and a 2.5% National Health Insurance Levy to be collected using the same mechanisms as the already existing 12.5% Value Added Tax (VAT). Two and a half per cent (2.5%) of formal sector worker contributions to the Social Security and National Insurance Trust (SSNIT) towards retirement benefits were to be automatically transferred to the national health insurance fund on a monthly basis.

A National Health Insurance Council was to govern the NHIS. The object of the Council was ‘to secure the implementation of a national health insurance policy that ensures access to basic health care services to all residents’ (Republic of Ghana 2003). Its responsibilities included registration, licensing and regulation of health insurance schemes, and supervision of their operations. It was also responsible for granting accreditation to health care providers, monitoring their performance, and ensuring that health care services rendered to beneficiaries were of good quality. A chief executive officer and supporting secretariat were to support the National Health Insurance Council in the execution of its functions.

The major disquiet expressed by organized labour groups over the bill was the channelling of 2.5% of their monthly SSNIT contributions into the national health insurance fund with no clear linked benefit to them. It appeared to be assumed that they could afford to give this money away. Parliament's response to these objections was to modify the bill so that as a benefit for giving up some of their social security funds, formal sector workers did not need to pay a premium to be covered by the district-wide MHO. Organized labour groups were still not happy with this arrangement. They felt the deductions would affect the long-term viability of the social security fund and pension payments. These concerns were part of longstanding concerns in the formal sector over low and inadequate pension payments and poor management of pension funds. They wanted to be allowed to organize social health insurance with separate payroll deductions that would leave their SSNIT contributions alone, and not lump them into district-wide MHO with the non-formal sector. Government's response was that the transfer of 2.5% of SSNIT pension payments into the national health insurance fund would not affect the viability of the SSNIT pension fund and the SSNIT funds were needed for cross-subsidization. There did not appear to be enough data to conclusively prove either government or organized labour groups right or wrong in their position on the viability of the SSNIT pension fund.

Organized labour groups also expressed concerns about the functions of the National Health Insurance Council as being too wide and sweeping, and likely to end up creating an expensive and unwieldy bureaucracy that would not necessarily advance the cause of national health insurance in Ghana. They were also concerned about the implications of government subsidizing district MHO regardless of performance (TUC and JCF 2003). In the widely publicized debates and commentaries, these other concerns were overshadowed by the more dramatic row over the 2.5% SSNIT deductions.

The major expressed concern of the minority NDC was that the introduction of a 2.5% National Health Insurance Levy represented a rise in VAT from 12.5% to 15% and was an excessively high tax burden. The majority NPP insisted that the National Health Insurance Levy was not VAT despite the remarkable similarity. The Levy was a particularly contentious issue between the majority NPP and the minority NDC given the history of VAT in Ghana. The NDC introduced VAT during their second term of office as the majority (1996–2000). At that time the NPP was the minority in parliament, and organized mass street protests against the introduction of VAT and the proposed rate as an excessive tax burden. The NDC had to back down on the original VAT bill and lower the proposed VAT rate to 10% before reintroducing it and getting it passed through parliament. The extra 2.5% was added as an educational tax. The opposition also supported organized labour's concerns about the 2.5% SSNIT deductions and the long-term viability of the social security fund.

The MHO already in existence expressed concerns about being classified by Act 650 as ‘private’ and therefore ineligible for any government support or subsidy. They were unlikely to survive unless they converted from independent organizations into government-sponsored district MHO. However, they were poorly organized, and many of their constituents were the rural poor, hardly any of whom participated in the vociferous debates that erupted around the passage of Act 650. Most of the civil society engagement in the extensive media debates was in the larger urban areas, with higher literacy levels and a bigger formal sector that had almost no MHO prior to the passage of Act 650.

Implementation of Act 650: 2004 to 2007

After the passage of Act 650, the MOH organized stakeholder meetings and set up task forces, with membership from technical actors in the health sector as well as the trusted political associates, to provide recommendations on implementation and to finalize the legislative instrument to accompany Act 650. Any optimism among technical and bureaucratic actors that the development of implementation arrangements would be more technical and involve less skipping of essential analytical and consultative stages in a rush to accomplish proved a mirage. The decision-making processes continued to be dominated by political approaches. The trusted political associates remained so powerful in the decision-making process that sometimes technical working groups would find that decisions on the issues they had been charged to work on had, in effect, already been taken. A few key policy and programme implementation decision examples—specifically the minimum benefit package and premium, premium exempt groups and provider administrative claims processing formats—are described to illustrate.

The minimum benefit package covered almost all outpatient care, investigations and drugs, many dental and eye services as well as most inpatient care including the cost of a general ward and meals. A decision was taken to exempt all Ghanaians under 18 years from payment of a premium based on the hypothesis generated by the close political associates that 50% of the population of Ghana was under 18 years. This would therefore be a rapid way of achieving the pressing objective of fulfilling the election promise of replacing ‘cash and carry’ with NHIS. There were enough adults over 18 who would pay a premium to add to the National Health Insurance Levy and 2.5% SSNIT contributions to carry the financial burden of those under 18.

The suggestion that there were too many unproven assumptions and it was necessary to analyse more carefully the population and financial data for the ability of the NHIS finances to carry a large exempt population for a generous benefit package was dismissed. Common approaches of the close political associates for dismissing suggestions included labelling the proponent as a member of the political opposition whose motivation was to sabotage government policy, or labelling the suggestions as part of attempts to slow down a process that needed to be completed as fast as possible. It was better to cut a few procedural corners and get things moving rapidly. Mistakes could be corrected later. The same arguments held on how to calculate an appropriate minimum premium. Government wanted an NHIS here and now in keeping with its promise and a minimum premium could be rapidly estimated by picking a rough cue from what the already existing MHO were doing. The fact that their benefit packages (many covering only inpatient care) and exempt groups (some did not have any) were more restricted than the proposed national minimum benefit package and exempt groups was glossed over as a possible problem with this approach. The minimum annual premium of approximately US$8 per adult aged 18 years and above was therefore rapidly decided following this approach. A recent analysis by the ILO suggests that without more careful analysis and review of benefit package, premium and exempt group arrangements in relation to the income of the NHIS, the NHIS fund may not remain viable over the medium to long term (ILO 2006).

In the example of the minimum benefit package, premium and exempt groups, it was clear why technical suggestions were considered inconvenient from a political perspective. In others, the opposition to the technical suggestion was not immediately clear until later events clarified things. Despite several consultative meetings and feedback complaints by providers at their complexity and the time they would take to process, there was a total refusal to simplify provider claims reporting formats and an insistence on maintaining complicated and elaborate time-consuming formats designed by the close political associates. The forms were pushed to the extent that they were incorporated into the Legislative Instrument accompanying Act 650 and made mandatory by law for providers to use as LI 1809 Forms 4, 5, 6, 7 & 8 (Republic of Ghana 2004). Later events revealed that large amounts of the forms had already been advance printed at considerable cost on the assumption that there would be no challenge to their introduction.

The director of policy at the MOH was made Chief Executive of the National Health Insurance Council Secretariat. The Members of the council were subsequently appointed as required by Act 650. MHO were rapidly set up in all districts that did not already have them, and managers were appointed for them by the National Health Insurance Council working with the district assemblies. The trusted political associates continued their role as implementation consultants in many of these districts.

The NPP government won the December 2004 elections, and came back to power in January 2005. In the face of widespread protests across the country concerning the operations of the close political associates as implementation consultants of the NHIS, the new Minister for Health wrote a letter suspending their services and ordered a financial audit into the operations of the NHIS. Like most other NHIS issues, the audit report was picked up by several media houses (e.g. Daily Graphic, Tuesday, 3 October 2006) even before it was made publicly available as part of the Auditor General's report to Parliament for 2004 & 2005 (Republic of Ghana 2006). It suggested some mismanagement.

In November 2005, the National Health Insurance Council used the powers conferred upon it by Act 650 and dismissed its Chief Executive. Despite much public and media speculation, the Council refused to be drawn and kept its public comments on the dismissal minimal. They stated that the dismissal was because the appointment was administratively irregular and the Chief Executive lacked the health insurance expertise to effectively manage the operations of the Council (JoyNews 2005). A new Chief Executive was appointed for the National Health Insurance Council through the public services commission and endorsed by the President as required by law. NHIS policy and programme development and implementation entered a stage of more caution in dispensing with public sector procedure and technical advice in the name of urgency and political expediency.

Discussion and conclusions

Based on the analytical framework of Grindle and Thomas (1991), the agenda-setting circumstances of Ghana's NHIS reform were such that there was a strong perception of a crisis and a need for change among political as well as technical and bureaucratic decision makers and civil society. The arenas of conflict were in the public domain, with policy and programme implementation decisions often highly and immediately visible and constantly under public scrutiny. There was strong pressure for the reforms and the stakes were high. High-level decision makers including the president and his cabinet were concerned about getting national health insurance to work in Ghana, and made this clear in public statements. For the ruling government, the high stakes were related in part to the need to demonstrate effective performance on a major and popular election promise within the 4 years before another election was due. For the political minority in opposition, despite being perceived as a potential threat to the programme by the majority, it appeared to be important to be seen publicly as fighting for appropriate policy and programme decision-making and implementation arrangements for a successful NHIS rather than as against the concept of an NHIS.

There appeared to be a sense among political decision makers of a need for immediate action on most of the policy and programme content decisions, and a fear that the opposition might try to take political advantage of any delays or seeming weakness and inability to deliver. Thus the policy decision-making concerns appeared to be dominated by concerns about political relationships and stability and a ready susceptibility to suggestions of attempts at political sabotage. This strong sense of insecurity and the urgent need to prove regime legitimacy and competence on a major national policy platform like the NHIS may have been heightened by the relative newness and therefore fragility of Ghana's multiparty democracy. In addition, as already mentioned, since independence, the Danquah Busia tradition from which the NPP evolved had only had a brief 2 year chance to prove itself before being ousted by a military coup.

For researchers, technical and bureaucratic actors, the need for successful reform was great and the stakes were high, because this was an area in which there were clear problems in health sector performance and for which change was desired sooner rather than later. Though they did not have the pressures of having to face an electorate for a new mandate to rule in 4 years time, the political sense of crisis and urgency was seen as an opportunity to bring about needed reform. They were aware of the technical issues, difficulties and challenges of developing and implementing a viable NHIS. It appears, however, that they were slower to fully discern the political concerns, climate, influences and policy characteristics, their importance, and how to create appropriate space to manoeuvre to steer policy in the desired technical direction within them.

Ghana, as a former British colony, had inherited and maintained a tradition of an independent non-politicized civil and public service. Part of the code of conduct in the public sector requires that civil and public servants stay politically non-partisan in the discharge of their duties. An advantage of this tradition is that it can ensure a certain level of stability and objectivity in governance and thus some continuity in development despite political regime change. However, for politicians concerned about regime stability and continuity, this kind of non-partisan tradition and code of conduct can be seen as potentially threatening. Under such circumstances, blurring the traditions of independence between technical and political actors by casting close political associates whose political loyalty was felt to be more assured in the role of technical and bureaucratic actors would seem desirable.

The perception that the close political associates were strong political loyalists, with better understanding of and responsive to the concerns around the crisis nature of the reform and the need for quick action than the more traditional-style technical actors, appears therefore to have been of critical importance. The close political associates appear to have been skilful at understanding the agenda-setting circumstances and policy characteristics, and using this to attain and maintain their strong power position in policy decision making and to downplay the importance of analytical and technical competence and considerations as compared with political credentials and loyalty. In some cases, they widened the power imbalance further by skilfully labelling the non-partisan stance of technical and bureaucratic actors as political opposition. The fact that in the process, there was sometimes a conflict of interest between actions that would promote their personal financial gain from the position so attained as compared with the national interest in terms of progress of the NHIS, and the potential damaging effects of this on the desired rapid progress of reform, appears to have been less quickly recognized by their sponsors. The observations of Aryee (2000) about the difficulties, despite the importance, of acknowledging corruption and its effect on the policy cycle are relevant here.

The glossing over of essential analytical stages seemed to be reasonable and essential short cuts in a perceived crisis decision-making situation with high stakes and pressure to take action for innovative change rapidly. Unproven short cuts, however, can sometimes prove longer than the tried and tested method. The essential analytical stages skipped before decision making came back later to haunt the programme with some level of potentially avoidable implementation difficulties. To effectively transform the health sector of developing countries, it is important to generate and make available scientific evidence on technical challenges and how to deal with them. However, for this evidence to be used effectively, it is also necessary to equip decision makers and stakeholders with skills in analysing, negotiating and working effectively within the politics of reform in a given context.

Much has been achieved in Ghana, as witnessed by the continuing survival of the NHIS programme and rising insurance coverage. However, complications have been added to the technical difficulties of implementing an NHIS by the failure to recognize early on and deal effectively with the political challenges. These complications have occurred despite a genuine commitment at the highest level of government to the NHIS. Others do not need to make the same mistakes. Even for Ghana, to review, remember and learn from the lessons of the past is to strengthen the present and the future. It is not only technical and bureaucratic actors who need to understand and plan for the political challenges of health reform alongside planning for the technical challenges. Politicians and policy elites in developing countries who genuinely wish to achieve a stated reform also need to understand and balance the technical and the political challenges. In doing this, it is useful to remember that a proper balance of power can help to attain a desired policy goal.

In the Ghana case study, the stronger power position of political actors to control and direct the policy process, combined with some political sense of insecurity and therefore relatively non-discriminatory reliance on trusted political associates for technical guidance, alongside the weaker power position of technical and other interest group actors, somewhat weakened the checks and balances inherent in a democratic system to protect the processes of a reform genuinely in the public interest. Technical and bureaucratic actors in traditions like that of Ghana are trained to maintain scientific integrity, political neutrality and independence in decision making. This is an ideal that is not always perfectly attained and it is possible for such actors to have partisan political biases. Nevertheless, political actors need to understand that it is a useful ideal to strive for if development is a goal. It is important to have disinterested actors in the policy process. Political neutrality rather than partisanship does not necessarily mean political opposition. Moreover, political partisanship does not necessarily achieve desired developmental or even political goals. Conversely, technical and bureaucratic actors who want to effectively inform public policy need to put efforts into understanding political actors and creating a relationship of mutual trust and respect between technical, bureaucratic and political actors without compromising scientific integrity. As Davis and Howden-Chapman (1996) observe:

‘… the relationship between research and policy making often involves a more fundamental relationship between researchers and policy makers. Increasingly, research evidence points to the importance of trust and ongoing commitment between parties when research is successfully translated into action.’

1
Author IAA was involved in the design, implementation and monitoring of the Dangme West district health insurance scheme experiment. She was a member of the Ministerial Task Force on Health Care Financing from 2001 to mid 2002 when she voluntarily resigned from formal membership. After mid 2002, she continued on an informal basis to provide technical advice on NHIS design as and when requested by the MOH. IAA was also involved in advising and assisting the Civil Servants Association in the Greater Accra region on a voluntary, unpaid basis, with proposals for an MHO along a classical social health insurance model before the passage of the NHIS act in 2003. Author SA was Director of the Health Research Unit of the MOH/Ghana Health Service from 1991–98 and Deputy Director General of the Ghana Health Service from 1999–2007. During the period of NHIS development in Ghana examined here, he provided advice to the MOH. He was appointed a member of the Ghana National Health Insurance Council for a 4 year term from 2004–07.

References

Agyepong
IA
Bruce
ES
Narh-Bana
S
, et al.  . 
Making health insurance and equitable and pro-poor financing mechanism in Ghana: some reflections
Medical Education Resources Africa (MERA), Ghana Edition
 , 
2006
, vol. 
21
 (pg. 
5
-
14
)
Arhin
DC
Rural health insurance: a viable alternative to user fees? A review and evidence from three countries
 , 
1995
London School of Hygiene and Tropical Medicine
 
PHP Departmental Publication No. 19
Arhinful
DK
The solidarity of self-interest: social and cultural feasibility of rural health insurance in Ghana
2003
University of Amsterdam
 
doctoral thesis
Aryee
JRA
Saints, wizards, demons and systems: explaining the success or failure of public policies and programs
 , 
2000
Inaugural lecture at the Amegashie Auditorium School of Administration
4 May 2000
University of Ghana
 
Accra, Ghana: Ghana Universities Press
Atim
C
Madjiguene
S
An external evaluation of the Nkoranza Community Financing Health Insurance Scheme, Ghana
 , 
2000
Bethesda, MD
Partnerships for Health Reform Project, Abt Associates Inc
 
Technical Report No. 50
Atim
C
Training of trainers manual for mutual health organizations in Ghana
 , 
2000
 
Sponsored by DANIDA Health Sector Support Office Ghana, United States Agency for International Development (USAID) and Partnerships for Health Reform (PHR)
Atim
C
Grey
S
Apoya
P
, et al.  . 
A survey of health financing schemes in Ghana
 , 
2001
Betheseda, MD
Abt Associates Inc
 
Partners for Health Reform plus (PHRplus)
Baltussen
R
Bruce
E
Rhodes
G
, et al.  . 
Management of Mutual Health Organizations in Ghana
Tropical Medicine and International Health
 , 
2006
, vol. 
11
 (pg. 
654
-
9
)
Brinkerhoff
DW
Accountability and health systems: towards conceptual clarity and policy relevance
Health Policy and Planning
 , 
2004
, vol. 
19
 (pg. 
371
-
9
)
Carrin
G
Community based health insurance schemes in developing countries: facts, problems and perspectives
 , 
2003
Discussion Paper No. 1
Geneva: World Health Organization
Carrin
G
James
C
Reaching universal coverage via social health insurance: key design features in the transition period
 , 
2004
Discussion Paper No. 2, EIP/FER/DP.04.2
Geneva: World Health Organization
Cassells
A
Health sector reform: key issues in less developed countries
 , 
1995
Forum on Health Sector Reform Discussion Paper No. 1. (WHO/SHS/NHP/95.4)
Geneva: World Health Organization
Daily Graphic
 , 
2006
 
NHIC flouts law. Tuesday, 3 October 2006. Story: Lucy Adoma Yeboah
DWHIS
Annual Report of the 1st Insurance year (1st October 2000 – 30th September 2001) Dangme West Health Insurance Scheme
 , 
2002
Dodowa, Ghana
Dangme West Health Insurance Scheme
DWHIS
Annual Report of the 2nd Insurance year (1st October 2001 – 30th September 2002) Dangme West Health Insurance Scheme
 , 
2003
Dodowa, Ghana
Dangme West Health Insurance Scheme
Davis
P
Howden-Chapman
P
Translating research findings into health policy
Social Science and Medicine
 , 
1996
, vol. 
43
 (pg. 
865
-
72
)
Ghana Statistical Services
2000 Population and Housing Census. Summary Report of Final Results
 , 
2002
 
Accra: Ghana Statistical Services, March 2002
Ghana Statistical Services
2003 Core Welfare Indicators Questionnaire Survey
 , 
2003
Ghana
Ghana Statistical Services
Gilson
L
Doherty
J
Lake
S
, et al.  . 
The SAZA study: implementing health financing reform in South Africa and Zambia
Health Policy and Planning
 , 
2003
, vol. 
18
 (pg. 
31
-
46
)
Glassman
A
Reich
MR
Laserson
K
Rojas
F
Political analysis of health reform in the Dominican Republic
Health Policy and Planning
 , 
1999
, vol. 
14
 (pg. 
115
-
26
)
Grindle
MS
Thomas
JW
Public choices and policy change: the political economy of reform in developing countries
 , 
1991
Baltimore, MD
John Hopkins University Press
GSS, NMIMR, ORC Marco
Ghana Demographic and Health Survey 2003
 , 
2004
Calverton, MD
Ghana Statistical Services (GSS), Noguchi Memorial Institute for Medical Research (NMIMR), and ORC Marco
ILO
Financial assessment of the National Health Insurance Fund
 , 
2006
Geneva
International Labor Organization
JoyNews
2005
 
Joy FM News, Ghana, 7, 8, 15, 16 & 17 November 2005, posted online at http://www.myjoyonline.com
MOH
Health Sector Five Year Program of Work
1996
Accra
Ministry of Health, Republic of Ghana
MOH
The Health of the Nation
2001
Reflections on the First Five Year Health Sector Programme of Work 1997–2001
Accra: Ministry of Health, Republic of Ghana
MOH
National Health Insurance Policy Framework for Ghana
2002
Accra
Ministry of Health
MOH
National Health Insurance Policy Framework for Ghana, Revised Version
2004
Accra
Ministry of Health
NHIC
2006 NHIS registration data
2007
Accra, Ghana
National Health Insurance Council
 
Unpublished
Patton
MQ
Qualitative evaluation and research methods
 , 
1990
Newbury Park, CA
Sage Publications
 
2nd
PPME-GHS
Facts and figures
2005
Accra, Ghana
Ghana Health Service, Policy Planning, Monitoring and Evaluation Division
Preker
AS
Health care financing for rural and low income populations: the role of communities in resource mobilization and risk sharing
 , 
2002
Washington, DC
IBRD/World Bank
Republic of Ghana
National Health Insurance Act, 2003
 , 
2003
 
Act 650. Accra, Ghana
Republic of Ghana
National Health Insurance Regulations, 2004
 , 
2004
Accra, Ghana
 
L.I. 1809
Republic of Ghana
Auditor General's Report to Parliament for 2004 & 2005
2006
Accra, Ghana
Tangcharoensathien
V
Jongudomsuk
P
From policy to implementation: historical events during 2001–2004 of Universal Coverage in Thailand
 , 
2004
Nonthaburi, Thailand
Distributed by National Health Security Office
The Chronicle
 , 
2006
(pg. 
10
-
11
Ministry of Finance and Economic Planning press statement on use of HIPC funds. Monday, 4 September 2006
Thomas
S
Gilson
L
Actor management in the development of health financing reform: health insurance in South Africa, 1994–1999
Health Policy and Planning
 , 
2004
, vol. 
19
 (pg. 
279
-
91
)
TUC and JCF
Understanding the proposed National Health Insurance Bill and its implications
Trades Unions Congress and Joint Consultative Forum
 , 
2003
 
Unpublished leaflet. August 2003
Waddington
CJ
Enyimayew
KA
A price to pay: the impact of user charges in Ashanti-Akim district, Ghana
International Journal of Health Planning and Management
 , 
1989
, vol. 
4
 (pg. 
17
-
47
)
Waddington
CJ
Enyimayew
KA
A price to pay: the impact of user charges in the Volta Region of Ghana
International Journal of Health Planning and Management
 , 
1990
, vol. 
5
 (pg. 
287
-
312
)
Walt
G
Gilson
L
Reforming the health sector in developing countries: the central role of policy analysis
Health Policy and Planning
 , 
1994
, vol. 
9
 (pg. 
353
-
70
)