COMMITTEE OF INQUIRY

Department of Health

Inquiry

Into the various Social Security Aspects of the South African Health System
Based on the Health Subcommittee Findings of the Committee of Inquiry into a Comprehensive System of Social Security
Title: Policy Options for the Future Covering

Public/Private integration

Public hospitals

Provincial budget process

Medical schemes

Social Health Insurance

National Health Insurance

14 May 2002
CONTENTS

1 Introduction *

1.1 Purpose *

1.2 Terms of Reference *

1.2.1 Overall terms of reference *

1.2.2 Interpretation of terms of reference *

1.3 Structure of Report *

1.4 Consultation Process Forward *

2 South African Health System: a review *

2.1 Overview *

2.2 Public sector *

2.2.1 Historical overview *

2.2.2 Reforms from 1994 *

2.2.3 Private revenue sources *

2.2.4 Compatibility with a contributory system *

2.2.5 Health Care Personnel *

2.3 Private sector *

2.3.1 Overview *

2.3.2 Prior to Supervision: 1889 to 1955 *

2.3.3 Supervision as Friendly Societies *

2.3.4 Supervision under the first Medical Schemes Act: 1967 *

2.3.5 The Regulation of Tariffs and Payments: 1968 to 1986 *

2.3.6 The "Free Market" Reforms: 1984 to 1988 *

2.3.7 The "Freedom" to Risk Rate: 1989 to 1994 *

2.3.8 Returning to Social Solidarity: 1994 to 1999 *

2.4 Concluding Remarks *

3 Stakeholder Views *

3.1 Overview *

3.2 Willingness to pay *

3.3 Ability-to-pay *

3.4 Earmarked tax *

3.5 Improvement of public sector services *

3.6 Injection of funds *

3.7 Phasing *

3.8 Revenue retention at facility level *

3.9 Benefits *

3.10 Findings and Concluding Comments *

4 The Achievement of Equity within the Health System *

4.1 Overview *

4.2 Defining Equity *

4.2.1 Review *

4.2.2 Equity Subject to a Budget Constraint *

4.2.3 Understanding a Definition Within the South African Context *

4.3 Strategic Elements of the Health System Affecting Equity *

4.3.1 Overall level of funding for the health system *

4.3.2 Income-based cross subsidies *

4.3.3 Health-related cross-subsidies *

4.3.4 Basic essential service and benefits *

4.3.5 Public sector *

4.3.6 Private sector *

4.3.7 Requirements for the future *

4.4 Concluding Remarks *

5 Financial Framework of the Public Health System *

5.1 Introduction *

5.2 Allocation of Funds arising from General Taxes *

5.3 Allocation of Funds arising from User Fees *

5.4 Allocation of Funds arising from Earmarked Taxes *

5.5 Alternative Options for Reform and their Implications *

5.5.1 Option 1 - Budget programmed at the national level *

5.5.2 Option 2 - Budget ring-fenced but not programmed at the national level *

5.5.3 Option 3 - Provincial discretion limited through use of national norms and standards *

5.5.4 Option 4 - Provincial allocations with full discretion *

5.5.5 Recommendation 1 *

5.5.6 Recommendation 2 (alternative to recommendation 1) *

6 Reform of the Tax Regime and Subsidies for Medical Scheme Cover *

6.1 Overview *

6.2 Value of the Tax Deduction *

6.3 National Health Insurance Committee proposals *

6.4 Assessment of the Tax Subsidy Framework *

6.5 Reform Options *

6.6 Prioritisation *

7 Risk-Equalisation *

7.1 Overview *

7.1.1 Policy Relevance *

7.1.2 NHI Committee Recommendation *

7.1.3 Need for Review *

7.2 Purpose of Risk-equalisation *

7.3 Definition of Risk-Adjustment *

7.4 International Review of Risk-Equalisation Mechanisms *

7.4.1 Criteria for the Selection of an Appropriate Risk-Equalisation Mechanism *

7.4.2 Criteria Used to Establish Risk-Equalisation *

7.4.3 Countries with Risk-Equalisation *

7.4.4 Review of Recommendations and Comments *

7.5 Evaluation of Residual Risk Selection in the South African Medical Schemes Environment *

7.6 Assessment of a Risk-Equalisation System for South Africa *

7.6.1 Introduction *

7.6.2 Risk Criteria Evaluated for South Africa *

7.6.3 Options for income cross-subsidisation *

7.6.4 Legislative Requirements *

7.6.5 Institutional Requirements *

7.7 Implementation *

7.8 Priority for Implementation *

8 Late Joiner Penalties and Mandatory Medical Scheme Cover *

8.1 Overview *

8.2 Evaluation of the Regulations *

8.3 International Experience *

8.4 Issues for South Africa *

9 Cost Containment in the Private Sector *

9.1 Introduction *

9.2 Cost Drivers *

9.2.1 Demand for health insurance cover *

9.2.2 Consumer demand for services *

9.2.3 Supplier induced demand *

9.2.4 Constraints on Developing Low-cost Healthcare Service Providers *

9.2.5 Competition Commission *

9.3 Supply Controls *

9.4 Regulation of Medical Schemes and Intermediaries *

9.4.1 Background *

9.4.1 Risk-selection *

9.4.2 Intermediaries *

9.4.3 Brokers *

9.5 Improving the Operation of the Market *

9.5.1 Enhanced amenities (buy-up options) *

9.5.2 Removal of Constraints on the Development of Staff Model Hospitals and Provider Services *

9.5.3 Intensive care and high care units *

9.5.4 Highly specialised services *

9.5.5 Radiology *

9.5.6 Dialysis *

9.5.7 Home-based care services *

9.5.8 Palliative care *

9.5.9 Step-down facilities *

9.5.10 Essential drugs *

9.6 Implementation *

9.7 Concluding Remarks *

10 Public Hospital Reform *

10.1 Purpose *

10.2 Overview of Policy Process and Trends Since 1994 *

10.3 Review *

10.3.1 Incentives to identify private patients and bill *

10.3.2 Flexibility to negotiate alternative forms of reimbursement *

10.3.3 Opportunities for making specialist units and services available to the paying market *

10.3.4 Treasuries and their approach to separate operational accounts for public hospitals *

10.3.5 Willingness to pay for public hospital services *

10.3.6 Consequences of a lack of revenue retention at public hospitals *

10.3.7 Consequences of a lack of public hospital autonomy *

10.3.8 Redistribution of retained revenue *

10.3.9 Opportunities for offering public sector services to private medical schemes *

10.3.10 Risks that public hospitals could lose their public character *

10.3.11 Cost implications for the private sector of public hospitals selling services to private medical schemes *

10.3.12 Public/private partnerships *

10.3.13 Equity and resource allocation *

10.3.14 Public health system incompatibility with the medical scheme reforms *

10.3.15 Arbitrary nature of relationships between provincial health departments and Treasuries *

10.3.16 Revenue generation undermined by treasuries *

10.3.17 Need for hospital autonomy *

10.3.18 Lack of political will with respect to decentralisation process *

10.3.19 Problems with dividing up management responsibility in public hospitals *

10.3.20 Problems with the centralisation of capital budgets for public hospitals *

10.3.21 Global budgets for public hospitals *

10.3.22 Relationship between hospital autonomy and social health insurance *

10.3.23 Link between the funding of public hospitals and national policy *

10.3.24 Centralisation of health budget coupled with decentralisation of operational control *

10.3.25 State of public hospitals *

10.3.26 Differential amenities versus differential services in public hospitals *

10.4 Hospital Decentralisation *

10.4.1 Comprehensive Approach *

10.4.2 Differential Amenities *

10.4.3 Financial Framework *

10.4.4 Governance Structure *

10.4.5 Human Resources *

10.4.6 Relationship to Private Sector *

10.5 Findings and Recommendations *

11 State-Sponsored Medical Scheme *

11.1 Overview *

11.2 Purpose of a State-Sponsored Low-Cost Scheme *

11.3 Target Group for Cover *

11.4 Benefits *

11.5 Contributions *

11.6 Relationship to Public Hospitals *

12 Civil Service Medical Scheme Cover *

12.1 Background *

12.2 Concerns with the Status Quo *

12.3 Discussion of Options *

12.3.1 Co-ordination of Civil Service Access to Health Cover and Services *

12.3.2 Mandating Cover for Civil Servants *

12.3.3 Restricted Medical Scheme: Proposal *

12.3.4 Restricted Medical Scheme: Benefit Options *

12.3.5 Restricted Medical Scheme: Administration and Intermediary Costs *

12.3.6 Accredited Medical Schemes and Limitation of Choice *

12.3.7 Equitable Subsidy System *

12.3.8 Funding the Post-retirement Subsidy *

12.3.9 Regionalisation *

12.3.10 Relationship to a Risk-Equalisation Fund *

12.3.11 Options in Relation to an Open State Sponsored Scheme Option *

12.4 Strategic Direction *

12.4.1 Overall Framework *

12.4.2 Potential Timelines *

12.5 Concluding Remarks *

13 Medical Savings Accounts within Medical Schemes *

13.1 Overview *

13.2 Discussion *

13.3 Industry Commission (Australia) *

13.4 Recommendations *

14 Key Strategic Challenges *

14.1 Introduction *

14.2 Context for Reform *

14.3 Evaluation of Current Policy Context *

14.3.1 Public sector *

14.3.2 Private Sector *

14.3.3 Mandatory Contributory System *

14.4 Concluding Remarks *

14.5 Role and scope of government involvement *

14.6 Role of the Public Sector *

14.7 Role of the Private Sector *

15 Integrated Strategy for Health Systems Reform *

15.1 Overview *

15.2 Principles *

15.3 Goals *

15.4 Reform Strategy *

15.5 Phase 1: Enabling Environment *

15.5.1 Objectives *

15.5.2 Preparation of the Public Sector Budget System *

15.5.3 Preparation of the Public Sector Hospital System *

15.5.4 Consolidation of Medical Scheme Reforms *

15.5.5 Development of a Policy Process on Basic Essential Services *

15.5.6 Development of Integrated Subsidy System *

15.5.7 Implementation of measures to Contain Private Sector Cost Increases *

15.6 Phase 2: Implement Preparatory Reforms *

15.6.1 Objectives *

15.6.2 Implement Risk-equalisation Fund for Medical Schemes *

15.6.3 Implement Risk-adjusted Subsidy to Medical Schemes *

15.6.4 Implement State-Sponsored Medical Schemes *

15.6.5 Implement Mandatory Environment for Civil Servants *

15.7 Phase 3: Implement Statutory Mandates *

15.7.1 Objectives *

15.7.2 Mandate Medical Scheme Membership *

15.7.3 Implement Voluntary Contributory Environment for Low-Income Groups *

15.8 Phase 4: Final Implementation of National Health Insurance *

15.8.1 Objectives *

15.8.2 Overall Framework *

15.8.3 Central Equity Fund (CEF) *

15.8.4 Public Sector Contributory Fund (PSCF) *

15.8.5 Subsidy to Medical Scheme Members *

15.9 Financial Implications *

15.10 Coverage *

15.11 Concluding Remarks *

16 Concluding Remarks *

Bibliography *

Tables

Table 2.1: Sources of Comprehensive Public Health Sector Financing, 1996/97-1998/99 (R million, real 1999/00 prices) *

Table 2.2: Distribution of healthcare professionals between the public and private sectors *

Staff *

Table 3.1: Willingness and ability to pay study: Reason for willingness to pay for public hospital services *

Table 3.2: Willingness and ability to pay study: Able-to-pay when last receiving care at a public hospital *

Table 3.3: Willingness and ability to pay study: Willingness to use and pay for public hospital care *

Table 3.4: Willingness and ability to pay study: Support for compulsory membership if Public Hospital Insurance *

Table 3.5: Willingness and ability to pay study: Support for compulsory Payroll deduction for covering public hospital costs *

Table 3.6: Willingness and ability to pay study: Support for payroll deduction with exemptions for Medical Aid members *

Table 3.7: Willingness and ability to pay study: Attitude towards a differentiated public health service *

Table 5.1: Evaluation of Alternative Options for Allocating the Health Budget *

Table 6.1: Impact on the Government Revenue from a removal of the employer tax deduction (R’billion) (based on 1999 registered medical scheme expenditure and 2000 prices) *

Table 7.1: Risk-adjustment systems in 10 countries *

Table 7.2: The practice of risk-adjustment in the United States *

Table 8.1: Premium penalties for late joiners *

Table 15.1 Summary of coverage by broad income category *

Figures

Figure 2.1: Real Total Medical Scheme Expenditure on Public and Private Hospitals (1995 prices) 1988 to 1999 *

Figure 2.2: Registered medical schemes: Per capita real expenditure and changes in beneficiaries *

Figure 2.3: Medical Scheme Beneficiary Changes for Open and Closed Medical Schemes, 1990 to 1999 *

Figure 2.4: Medical Scheme Beneficiary Changes for Open and Closed Medical Schemes, 1990 to 1999 *

Figure 2.5: Non-medical Cost Trends from 1993 to 2000 (Rands) *

Figure 2.6: Medical Scheme Real Cost per Beneficiary and Benefit Trends, 1993 to 2000 *

Figure 3.1: Willingness and Ability to Pay Study – Distribution of Respondents by Income *

Figure 7.1: Illustration of the Netherlands Risk-Equalisation Fund *

Figure 7.2: Price advantage/disadvantage for schemes representing 90 percent of the open scheme membership *

Figure 7.3: Possible Institutional Framework for a Central Risk-Equalisation Fund for South Africa *

Figure 9.1: Cost drivers in the private health sector *

Figure 12.1: Regional Structures for Civil Service Medical Scheme Strategy *

Figure 12.2: Framework for Universal Contributory Cover for Civil Servants *

Figure 12.3: Timelines for Implementation of a Strategy for Universal Medical Scheme Benefits for Civil Servants *

Figure 15.1 Reform Strategy and Approximate Timeline *

Figure 15.2: Framework for a Universal Contributory System *

Figure 15.3: Implications for Coverage over all Phases *

Introduction

Purpose

This report provides an evaluation of alternative strategic policy options for the South African health system based on a review performed by the Department of Health and the Health task group of the Committee of Inquiry into a Comprehensive System of Social Security.

Although certain aspects of the review cover existing policy accepted by Government, this report is a consultation document and does not represent a final policy position by the Department of Health.

The review proposes a broad course of action for achieving a move toward a more effective and unified health system. The purpose is to highlight key policy issues as a point of departure for consultation and the preparation of a final position by the Department of Health.

Terms of Reference

Overall terms of reference

The terms of reference given to the Committee requires the review of a broad number of elements relating to social security. The general objectives of this analysis include:

Options on ultimate objectives and targets for the social security system: Alternative options indicating an envisaged final structure should be provided. These should be extensively motivated and viable. (Terms of Reference, 2000, par. 2.1.1).

Options for immediate practical implementation: alternatives consistent with envisaged ultimate objectives should be outlined. These would need to be practical and focused on immediate needs, the current level of South Africa’s development and affordability. (Terms of Reference, 2000, par. 2.1.2).

Viability and implications of options considered: all relevant information concerning the viability and significant negative or positive implications linked to any options considered must be provided. (Terms of Reference, 2000, par. 2.1.3).

The specific social security areas that must be covered are:

National pensions system: This must involve an assessment of the entire environment providing for post-retirement cover, as well as general financial support for the aged. (Terms of Reference, 2000, par. 2.2.1).

Social assistance grants: This must involve an evaluation of the entire social assistance mechanism including all grants, their funding mechanisms, and the efficiency with which they achieve their goals. (Terms of Reference, 2000, par. 2.2.2).

Social insurance schemes: All social insurance schemes, including funding and protections for injury on duty and cover for road accident victims must be examined. (Terms of Reference, 2000, par. 2.2.3).

Unemployment insurance: The current system of unemployment protection must be examined. This must include the adequacy of all forms of support for the unemployed, including special employment programmes. (Terms of Reference, 2000, par. 2.2.4).

Health funding and insurance: The public and private sector environments must be examined with a view toward ensuring universal access to basic health care. (Terms of Reference, 2000, par. 2.2.5).

Each of the specific areas identified above must include the following analyses: (Terms of Reference, 2000, section 2.3).

Existing processes: In many instances there are existing policy processes examining specific funds and safety nets. The Committee will be expected to liase extensively with these initiatives in order to inform the final recommendations.

Core issues: Each policy area must be examined taking account of the following:

Adequacy of adherence to principles of social solidarity;
The legislative and general regulatory environment;
The social budget;
Institutional structure;
The tax environment;
Sources of finance;
Perverse incentives;
Significant gaps and the underlying reasons;
Macroeconomic environment;
Impacts on government as an employer;
Income distribution.

Key recommendations on future directions:

Long-term or ultimate objectives and targets;
Short-term or required intermediate reforms consistent with the long-term objectives

Implementation process: The Committee must make concrete recommendations on implementation steps and prerequisites.

In addition to the specific analyses indicated above, the Committee is also required to develop a social budget for all the key social security areas. (Terms of Reference, 2000, section 2.4).

The Committee must generate a detailed social budget for the country, outlining public and private expenditure on key areas of social policy.

The Committee must also set up the basis for the annual presentation of a social budget for the country. This will involve the creation of the relevant capacity within key government departments to ensure this can be done.

The Committee is also expected to enter into a fairly broad consultation process with all stakeholders. (Terms of Reference, 2000, section 2.5).

The Committee will be required to consult with all relevant stakeholders linked to the core issues under examination. The nature and structure of this consultation will be at the discretion of the Committee.

The Committee will be expected to take inputs from all relevant South African experts in the various policy areas under examination.

The Committee will be expected to consult directly with all government departments affected by the proposals.

The Committee will be expected to review all relevant material on international practice in both industrialised and developing country settings.

Interpretation of terms of reference

The specific section relating to health issues is very broad and effectively involves a review of the entire health system with a view to ensuring universal access. As such the terms of reference preclude reviewing issues which do not involve the provision of universal access. This is a significant limitation and implies a prior policy decision in this regard.

There is substantial international evidence that such a policy stance is appropriate and for this reason the limitation is accepted. The issue of the desirability of universal access and how this might be interpreted will however be addressed by the Committee as wide policy discretion is possible here.

Universal cover internationally is provided through a mixture of methods. These include non-contributory and contributory financing systems as well as various service delivery models. The contributory environments typically involve both earmarked taxes or various degrees of compulsion applied to private insurance markets.

Structure of Report

The initial sections of the Report, from sections 2 to 4 provide the context for later discussions and recommendations. Section 2 provides an historical review of South Africa’s public and private health sector. Section 3 summarises the results obtained from a number of stakeholder reviews performed by the Department of Health. Section 4 outlines key equity principles that universally underpin health systems policy.

Sections 5 to 12 provide evaluations and policy recommendations. Section 5 looks at the financing of the public health system and linkages to broader reform objectives. Section 6 looks at the tax subsidy provided to medical scheme members, and reviews how this could be altered and incorporated into a more integrated subsidy system linking the public and private sectors.

Section 7 evaluates the need for risk-equalisation within the medical schemes environment and indicates how this can be linked to the reform of the tax subsidy discussed in section 6. Section 8 analyses the current mechanism for dealing with systemic adverse selection within the medical schemes environment, late-joiner penalties, and the option of moving toward mandating membership of a medical schemes as an alternative and ultimate solution.

Section 9 looks at the issue of cost-containment in the private sector and options required to deal with the problem. Section 10 evaluates the need for public hospital reform in order to improve the management of hospitals and to permit them to obtain funding from medical schemes.

Section 10 discusses the option of a state-sponsored medical scheme and how this could influence the development of a low-cost provider market. Section 12 assesses the important question of mandating universal cover for civil servants. The options in this section relate to both to the development of a low-cost provider market (section 10), and the development of a state-sponsored medical scheme (section 11). Section 13 looks at medical savings accounts and their role in the health system.

Section 14 summarises some of the key strategic challenges facing the South African health system, based on international evidence and the reviews provided in this Report. Section 15 integrates all the various issues and recommendations raised in the Report into a strategic reform process.

Consultation Process Forward

This document provides information intended to generate comment from the public on a number of key questions affecting the future of the health system. The process needs to ensure that the response from the public extends further than those with a commercial stake in particular directions. The areas that are important for the purposes of feedback to the Department of Health are:

Central recommendations and proposed directions framed in this Report;
The development of a contributory environment for low-income groups (i.e. medical schemes);
Reform of the management and governance of public hospitals;
The full retention of revenue at public hospitals;
Budget options for the public health system, taking note of the need to ensure compatibility with either a mandatory of voluntary contributory environment (medical schemes);
The system of cross subsidies to be guaranteed by government;
The contents of a basic essential set of services which government must be provided by the public sector, the private sector, and any future mandatory contributory environment;
The role of the private health sector and its importance for achieving greater integration with the public sector;
Reform requirements for the medical schemes movement, ensuring that all key groups remain covered with access to a comprehensive set of benefits;
The implementation of a system of risk-equalisation between medical schemes;
Conversion of the existing tax subsidy for medical schemes into an on-balance-sheet per capita allocation for those covered in the medical schemes environment, and how such a system may form the basis for a future system of national health insurance;
Options to contain cost increases within the private health system including reforms required to enhance competition for lower-cost service provision (both hospitals and primary care) in the private sector;
The introduction of a system of direct controls to limit medical cost increases in the private health system; and
Medium- to long-term options for the development of social and national health insurance options.

South African Health System: a review

Overview

This section reviews the evolution of the South African health system, both public and private. This review provides a backdrop against which certain strategic challenges are identified in section 14, and provides the basis for recommended strategic reforms identified in later sections.

Public sector

Historical overview

One of the first pieces of legislation enacted of a purely medical nature was the "Contagious Disease Act" (No. 1 of 1856). This was to deal with regular outbreaks of measles and smallpox. In 1867 an epidemic struck Cape Town with a high mortality rate. This resulted in the enactment of the Contagious Diseases Act No. 25 of 1868. In Kimberly the government passed the Medical Tax Act in 1874 as a means of financing the provincial hospital. In terms of this legislation a fee of one shilling was levied upon each "native" worker for medical services on the diamond diggings. The diggers, however, opposed the Act and enforcement was deferred until 1882 when the various companies paid the levy for their "native" employees directly to the Cape government. (Ginwala, 1981).

The first Public Health Act was promulgated in South Africa in 1833 following a smallpox epidemic in Kimberly. For the first time vaccination and notification of infectious diseases was made compulsory in the Cape Colony. Extensive emergency powers were delegated to local authorities by the governor to permit officials to enter premises, and draw up and enforce quarantine regulations. Local authorities were given the power to establish hospitals and departments while the government advanced 50 percent of the costs of expenses and maintenance. (Ginwala, 1981).

The South Africa Act resulted from the National Convention of 1909 which created the Union of the four colonies, Cape of Good Hope, Natal, Transvaal, and the Orange River Colony. The Act made limited references to health care. The four Provincial Councils were endowed with various health and local government laws inherited from the Colonies. Local Authorities, by virtue of previous Colonial legislation and subsequent ordinances and under their local by-laws were responsible for environmental hygiene and measures to deal with outbreaks of infectious disease. There were however overlapping responsibilities and confusion with respect to public health. The influenza epidemic of 1919 exposed serious inadequacies in the existing responsibilities, safeguards and procedures. This resulted in the Public Health Act No. 36 of 1919. (Ginwala, 1981).

In terms of the Public Health Act provincial administrations retained their responsibility of administration of local government and the establishment, maintenance and management of general hospitals and matters relating to charitable institutions and for pauper medical relief. The Act established the Department of Health with executive responsibility. The intention of the Act was to decentralise. The Department of Health was given powers to advise, assist and if necessary coerce the local authorities and provincial administration in fulfilling their public health responsibilities. (Ginwala, 1981).

Local authorities had as their primary role the control of infectious diseases and environmental sanitation. These functions were facilitated by the statutory provision for refunds in respect of certain staff and certain services for infectious diseases. The Act made a distinction between communicable and non-communicable disease. The State took responsibility for persons with a communicable disease through isolation and prevention of spread of infection. Responsibility of a person with non-communicable disease was accepted as part of pauper medical relief at provincial hospitals and district surgeons. For the majority of people provision of health care was an individual responsibility. (Ginwala, 1981).

Private hospitals were subdivided into those that existed for gain and those that did not. The non-profit hospitals were divided into those established for philanthropic reasons and those established to fulfil statutory requirements. Into the latter category fell the Mine and Indian Immigration Bureau hospitals that were developed in response to peculiarly South African arrangements of labour supply. The pre-Union Natal government levied a special tax upon employers of Indians, the proceeds of which were paid into a fund administered by the Indian Immigration Bureau and utilised to establish special hospitals under its control, and relieving the tax-payer of this particular burden. (Ginwala, 1981).

The Union Legislature in 1911 passed the Native Labour Regulation Act which imposed on the gold and other mining industries the duty of providing hygienic housing, adequate diet and hospitals for Native labourers employed by them. Tax was not imposed on the employers, but in accordance with the regulations and its specifications, the employers provided hospitals. Equivalent legislation did not emerge in the instances of secondary industry. (Ginwala, 1981).

Mission hospitals were established in rural areas where no local authority capable of making a financial contribution. (Ginwala, 1981).

The Public Health Amendment Act of 1946 demarcated the functions of the Central Government and the Provinces. The provinces were responsible for general hospital services and outpatient services connected with their institutions while the government was to proceed with extra institutional services by the development of a system of health centres. The Act made provision for refunds to both provincial administrations and local authorities in respect of any outpatient services independent of general hospitals which either would institute. Difficulties in implementing services and funding however arose. (Ginwala, 1981).

In the period after 1948 health policy and planning became more determined by political rather than health criteria. The focus was essentially that of satisfying the needs of the white population. The Tomlinson Report of 1954 for instance recommended a separate "Bantu Health Service" which ended the moves of the Department of Health of the time to create a unitary system. The subsequent development of a homeland system in South Africa further fragmented service delivery and policy through the extension of numerous first tier government structures. In an attempt to co-ordinate the functions of the numerous health departments the Regional Health organisation of Southern Africa (RHOSA) was established in 1979. (van Rensburg et al, 1995, p.57).

The Health Act of 1977 for the first time included Provincial Administrations in the same way that local authorities were involved since the first Public health Act of 1883. Under this Act the Department of Health had the functions of co-ordinating health services rendered by Provincial Administrations and Local Authorities as well as to provide such additional services as may be necessary to establish a comprehensive health service for the population of the Republic of South Africa. (Ginwala, 1981).

According to Van Rensburg et al (1995) despite the fact that the Health Act 63 of 1977 intended to rationalise health care organisations by means of clearer definitions of the duties, powers and responsibilities of the respective authorities; to effect greater co-ordination between the various tiers of authority and to move to a nationally co-ordinated health policy, the Act had no real effect on the fragmentation embodied in the three tiers of authority and services. The provinces remained responsible for hospital services, local authorities were responsible for preventive and promotive care, and the central department was responsible for overall co-ordination. Two new bodies were created to achieve greater overall co-ordination: the Health Matters Advisory Committee and the National Health Policy Council.

The period from 1980 onward continued to be characterised by a high degree of fragmentation in the health services and policy co-ordination. The implementation of the homeland policy and the Tricameral system (in 1983) led to the breaking up of administrations into seventeen different political entities many of which had little political legitimacy. Although the four provincial administrations effectively covered the vast majority of the population, great disparities existed in resource allocations within and between provinces and between the provinces and homeland administrations. Public facilities were also segregated with separate services for the non-white population. In many instances this separation extended to entire facilities with separate white and non-white hospitals.

The Browne Commission, appointed in 1980 concluded in 1986 that there was excessive fragmentation of control over health services and a lack of policy direction, resulting in a misallocation of resources, duplication of services and poor communication between the various tiers.

The National Policy for Health Act 116 of 1990 repealed the sections of the Health Act related to policy-making structures. It made provision for the Minister of National Health and Population Development to determine policy. Three new bodies were established to assist in this, the Health Policy Council, and the Health Matters Committee. The Act attempted to co-ordinate health services and to diminish the role of provincial authorities. (Van Rensburg et al, 1995, p.59).

During 1989 to 1990 the public health system was officially desegregated technically generating a dual system based on income differentials. However, the former structure of the public sector system and its target population resulted in informal barriers to access that have survived the legal barriers. These barriers relate to location, i.e. many public facilities are situated in areas that make services inaccessible to the population now legally entitled to use them, and to the appropriateness of the service, i.e. many of the services and personnel are inappropriate for low-income and socio-economically deprived communities.

In the late 1980s and early 1990s the desegregation of public hospitals resulted in a dramatic growth in private hospitals. This growth in public hospital utilisation was mirrored by dramatic cost increases in medical aid costs experienced in these years.

Reforms from 1994

Subsequent to 1994 the public health system was reformed administratively along the lines of the new Constitution. Nine provincial health administrations were created responsible for the delivery of both hospital and primary health care. The provincial administrations transfer a portion of their budgets to local authorities who also render primary health care services. Overall responsibility for health policy resides with the national Minister of Health supported by the national Department of Health.

Since 1994 a number of significant changes have occurred in the financial arrangements of government in general with major implications for the rendering of health services. These have included the introduction of a fiscal federal system affecting the financing and budgeting of virtually all significant social services, including health, social development and education.

These financial changes have impacted in a number of ways that have relevance to the overall principles and objectives of the health system based on a large number of reports. These include:

Budget levels;
Inter-provincial equity;
Revenue raised from medical schemes and other user charges; and
Staff retention within the public sector.

The central issue is whether the changes occurring within the public health system reflect explicit policy decisions, or are merely a consequence of structural difficulties in co-ordination and implementation to achieve centrally determined policy objectives. A key issue in this discussion is the role and extent of provincial discretion relative to national policy requirements and whether a proper balance is currently maintained.

For the 1995/96 and 1997/98 financial years the health budget for the country was determined centrally based on recommendations by the Health Function Committee. In the 1997/98 financial year the public sector officially switched over to a fiscal federal system whereby budgets for health were determined by provincial legislatures and not advised by national policy. The funds made available to provinces were allocated through an unconditional (equitable share) grant allocated from the national budget.

According to the recent National Health Accounts review (NHA Review) there has been a systematic overall and per capita decline in public health expenditure since 1996/97. According to the NHA Review this expenditure decline is in part attributable to the peculiarities of the fiscal federal system rather than to any explicit national policy decisions. Various studies as well as the NHA review have pointed out the inability of the health system to achieve national policy objectives with respect to equity as a consequence of the fiscal federal environment. Initial progress toward equity, prior to the introduction of the fiscal federal environment, appear to have reversed in the years since 1996/97. However, a significant contributor to the decline in health budgets appears due to the decline in the overall allocations to provinces. (Thomas S. et al, 2000).

Table 2.1: Sources of Comprehensive Public Health Sector Financing, 1996/97-1998/99 (R million, real 1999/00 prices)

Sources of Finance	1996/97	1997/98	1998/99
General taxation	29,244	30,972	30,908
Local authority revenue	845	963	996
User fees from households	499	418	340
Provincial Government own revenue	334	578	384
Donors	18	33	68
Total	30,941	32,963	32,695

Source: Thomas S. et al, 2000, p.133

Various problems relating to the budget system for health care were brought to light by the Committee reviews. The problems appear pervasive and impact significantly on the performance of the health delivery system. These are listed below:

Policy decisions concerning health care at a national level cannot be backed by resource flows as the provincial governments are responsible for budget setting for health services. As a consequence a fundamental public health principle, that of equity, is potentially undermined. There is also significant potential for policy fragmentation.

Budget allocations for health departments are declining in real terms in all provinces despite substantial emerging needs. No specific measures have been undertaken to deal with the service-related impact of HIV/AIDS.

The conditional grants allocated for teaching and research and supra-regional services (highly specialised services that are only available in a few provinces) are not linked to any specific services, and are apparently being reduced in real terms without any clearly defined policy framework.

Conditional grant allocations to provinces are undermined by provincial treasuries, who dominate resource allocation decisions. As such the primary motivation for a "conditional" grant is substantially undermined. This appears to occur for all grants including the capital funds made available as conditional grants.

As yet no specific norms and standards that can be used for budget motivation and resource allocation have been satisfactorily developed. As such, the basis for resource allocation decisions cannot be defined.

The allocation of capital budgets for the health system is inefficient and in a number of provinces is still highly centralised, particularly where Public Works departments are used. The use of a dedicated Public Works department results in inefficiencies as the hospital or facility is several steps removed from the process of determining capital allocations and the procurement process. A large portion of the capital backlog in state hospitals can probably be attributed to the over-centralised budget process for capital allocations. In recent years this appears exacerbated by the centralisation at national level of hospital rehabilitation funds.

In recent years there has been a tendency for an increased level of funding for normal capital expenditure to be made available at a national level. Access to these funds is uncertain, difficult and bureaucratic. As a result, time periods between application, approval, draw-down and utilisation are so long that they exceed the financial year in which the funds were made available. The funding of capital backlogs in such a centralised manner is inefficient and probably leads to poor prioritisation. It also does not address the primary reason for the emergence of capital backlogs in the first place, i.e. the centralised and bureaucratic budgeting process for capital.

All funds raised in additional revenue from medical scheme and other private patients is not retained by the relevant cost-centre (i.e. the hospital). As such hospitals lose out financially when they treat private patients. The uncertainty associated with the return of funds is a fundamental factor undermining the ability of public hospitals to access non-tax revenue.

The Treasury Department is not applying a consistent approach with respect to mixed financing options within the public sector generally. This has particular relevance for the Health system, as a degree of flexibility between alternative revenue sources for public health institutions is a characteristic of all successful health systems. Such a framework generally regards additional revenue collected from user charges or pre-paid (capitation) user charges as additional to general revenues and hence not part of the tax system. This principle is not being applied in respect of public hospitals where such charges are effectively regarded as part of general tax revenue.

The current system of public finance for the health system is inflexible and inefficient. Institutions are generally uncertain about future budget allocations, despite the Medium-Term Expenditure Framework (MTEF). Additional allocations such as conditional grants or fees generated from private patients are undermined by budget cuts or not returned to the institution.

Hospital managers are disempowered in their ability to manage large and complex institutions. This is primarily a consequence of poor governance structures for hospitals rather than the quality of hospital managers. Despite the very large amounts of money they spend, the head of a hospital is not the accounting officer. They have little control over the appointment of staff, particularly their own support staff. Hospital boards do not have fiduciary responsibility for the hospitals they oversee, as a consequence the oversight value of the boards are minimal.

Hospital managers have little control over minor and major works budgets, expenditure and procurement. This, as much as tight financial conditions, plays a significant role in generating capital backlogs in institutions.

The budgeting system appears to involve an inconsistent mix of centralisation and decentralisation with respect to the public health. Firstly, budget allocations are decentralised to provinces resulting in a weakening of equity and other national policy objectives. Secondly, key aspects of the budget which are more appropriately managed with a high degree of decentralisation are at a national or provincial level. Thirdly, hospital managers have insufficient control over operational matters.

The system of user fees, charged at point of service, used by public hospitals relies on the application of a means test. This system is both dysfunctional and inequitable (as it is not evenly and consistently applied). This is discussed further below in section 2.2.3.

Private revenue sources

The public hospital sector currently charges fees for hospital services to higher income earners. Indigent patients receive free services. The fees charged are highly subsidized and invoiced at point-of-service. Invoices are also submitted to medical schemes. This system is highly inefficient as a source of revenue for public hospitals and needs to be substantially revised. The problems run fairly deep and cannot be resolved merely through improvements to the billing system or the setting of higher fees:

The system of billing patients at point of service is complex, especially where an assessment of the income status of the patient must be performed on the spot. Given the inability of the hospital to turn patients away, this assessment is not practical.

Claims from non-medical scheme members: Where patients are not members of a medical scheme, following up many small or even large unpaid accounts is costly and impossible to administer. Collections are only cost-effective where unpaid accounts are for greater amounts and smaller in volume. The majority of unpaid accounts from non-medical scheme members are however small and numerous (low-cost - high-volume). Given the enormous volume of activities occurring within health systems, the most efficient forms of funding involve bulk payments and billing.

Retention of revenue: As hospitals do not retain revenue, they have a reduced incentive to claim user fees or funds owed from medical schemes or related funds. However, the problem runs deeper than this. Retention of revenue would not materially alter the funding position of public hospitals if budgets are reduced to such an extent that they offset the additional revenue. Problems are:

New patients: Where additional revenue comes from patients formerly seen in private hospitals, the full cost of the services provided must be recovered. Where budgets are cut to offset new revenue from fees, public hospitals will be in a worse funding situation than before. New patients must be funded at full cost, and a net increase in the funding of the hospital by this amount should occur.

Previously non-paying patients: Where patients are made to pay where formerly they used the service free of charge, a reduction in budget reduces the hospital to the position it was in when patients were not compelled to pay. Budget reductions in these instances will result in a reduced incentive to recover fees, and a consequent drop in services and service quality.

Figure 2.1: Real Total Medical Scheme Expenditure on Public and Private Hospitals (1995 prices) 1988 to 1999

Source: Council for Medical Schemes Statutory Returns 1988 to 1999

Compatibility with a contributory system

The existing financial management system of government is not compatible with any form of contributory system, whether fee-for-service or pre-paid in one form or another. It is more compatible with a tax funded free service. If it is a policy decision to include a contributory component to the public system, as is apparently the case at present, then the appropriate technical reforms are needed.

Health Care Personnel

There has been a substantial shift of key healthcare professionals out of the public sector. Only 45.5 percent of all professional nurses work in the public sector (based o the 1999 estimates). Over the past 10 years general practitioners have also moved out of the public system with a shift from 38.3 percent in 1989 to 22.5 percent in 1999.

At present only 37 percent of all surgery related specialists function within the public sector. Seventy five percent of all anesthetists work exclusively in the private sector.

Table 2.2: Distribution of healthcare professionals between the public and private sectors

	Total SA		Public**		Private*
Staff		Total no.		%		Total no.	%
Period: 1989/90
Medical officer (GP)	12,889	4,936		38.3%		7,953	61.7%
Specialist	5,595	1,891		33.8%		3,704	66.2%
Pharmacist	8,262	909		11.0%		7,353	89.0%
Dentist	3,111	218		7.0%		2,893	93.0%
Period: 1998/99
Professional nurse***	90,923	41,401		45.5%		49,522	54.5%
Staff nurse	33,039	21,008		63.6%		12,031	36.4%
Nursing assistant	51,583	22,550		43.7%		29,033	56.3%
Total nursing	175,545	84,959		48.4%		90,586	51.6%

Medical officer (GP)***	15,376	3,458		22.5%		11,918	77.5%
Specialist***	6,136	1,741		28.4%		4,395	71.6%
Pharmacist	9,599	1,210		12.6%		8,389	87.4%
Dentist	3,482	471		13.5%		3,011	86.5%

Sources: *Sodelund et al, 1998, and **PERSAL 1999.

***These data were adjusted to full-time equivalents according to average salary costs and total expenditure for these personnel categories in 1998/99. Many doctors and some nurses only work part-time and therefore the Persal information used without adjustment distorts the actual number employed.

Private sector

Overview

No regulatory supervision of the private health system existed prior to 1956 after which schemes became regulated as Friendly Societies. From 1967 to 1975 medical schemes were regulated in terms of the Medical Schemes Act by both the Department of Health and the Registrar of Financial Institutions. From 1975 onward, medical schemes were regulated exclusively by the Department of Health. The policy underlying medical scheme regulation shifted considerably over time. A review is provided below of the historical evolution of the private health system in South Africa, drawing on a paper produced for the Melamet Commission (1994).

Prior to Supervision: 1889 to 1955

The first "medical scheme" in South Africa was the De Beers Consolidated Mines Ltd. Benefit Society, established in 1889. By 1910 seven such schemes were in existence.

At the creation of the Union of South Africa in 1909, no co-ordinated health system existed in the country. A centralised authority for co-ordinating health activities was only created ten years later in response to the influenza epidemic of 1919. This was incorporated in the Public Health Act, No. 36 of 1919, and resulted in a three tier public sector structure of central government, provincial administrators and local authorities. A system of private practitioners developed outside of the public system but there was no regulation of private sector schemes.

By the beginning of the Second World War in 1940 there were 48 medical schemes. After the War the significant increase in the number of schemes generated a need for some regulation. Consequently in 1950 the Advisory Council for Medical Fund Societies was formed. Its role was to act as a representative for the affiliated schemes in negotiations with the Medical Association of South Africa. Before 1956, however, there was no regulation of the behaviour of medical schemes.

Supervision as Friendly Societies

On 31 December 1956 the Friendly Societies Act, No. 25 of 1956, came into effect. All schemes (with a few exceptions) were required to register as Friendly Societies before they could operate. The controls applied by this Act were primarily financial in nature.

The feeling of the Reinach Committee (1962) was that although the registration of schemes would lead to greater stability, more comprehensive legislation would be required to control all other aspects of medical insurance. A consequence of the lack of legislation was that until the late 1960s no uniformity would exist within private sector healthcare. Schemes varied significantly in the coverage they offered.

Requests were made to the House of Assembly to set up a countrywide medical scheme in the 1950s to ensure coverage for the entire white population. In 1959 the private sector took the initiative and started the "Plan for Medical Services". The scheme was initiated, administered and controlled by doctors, and used its own schedule of tariffs which were higher than the preferential tariffs applied to other schemes. The plan offered 100 percent benefits for services offered by medical practitioners and hospitals, but excluded dentistry and medicines. Members had no say in the administration or determination of membership fees or benefits.

Some employee organisations advocated a National Health Scheme while others supported a comprehensive state-supported "existence-protection" scheme which included schemes. However, most were not in favour of amalgamation.

By 1960 there were 169 schemes providing cover for 368,890 members and 588,997 dependants. These schemes served the needs of the white middle class, especially those in urban areas. The importance of this type of scheme can be seen in the rapid growth in coverage that this form of scheme provided for the predominantly middle class white population. For whites, over a period of 15 years from 1945 to 1960, coverage grew from 48 percent to 80 percent of the eligible population.

By 1960 virtually all whites in South Africa had shifted away from the free services provided by government. On the other hand, 95 percent of non-whites were reliant on public sector services, which were largely free. By this time membership of medical schemes had effectively become mandatory for whites due to it being a condition of employment and given that virtually all whites were formally employed. Pensioner members of many schemes received the same medical benefits as ordinary members, but free of costs.

In 1960 three types of scheme existed:

Sick Funds: This was the oldest type of scheme (and still the largest in 1960). Members paid fees or premiums in return for comprehensive cover for medical services and medicines. The member’s choice was limited to a panel of permanent and/or temporary medical practitioners who were remunerated on a capitation basis. Sick funds were crude fore-runners of pre-paid plans equivalent to health maintenance organisations (HMOs) and independent practitioner associations (IPAs).

Benefit Funds: (Not to be confused with benefit funds registered under either the old Friendly Societies Act or the Income Tax Act). Benefit funds evolved from sick funds which was how many had begun. Members paid a premium which permitted a free choice of practitioners who were reimbursed on the basis of a preferential tariff per service. The member had to pay a percentage of the bill with percentages varying according to membership fees and benefits. Funds were organised by groups of professions or medical practitioners. No clear distinction existed between sick funds and benefit funds with differences depending upon benefits and reimbursement methods.

Assurance Schemes: Assurance schemes originated as a means of supplementing the benefits of benefit funds. These schemes functioned on the basis of a third party taking the initiative to provide medical cover for profit.

Many medical practitioners shifted into sick and benefit funds due to the large number of patients they were able to access in this way. However, a significant number of medical practitioners also provided full-time and/or part-time health services according to Government or provincial tariffs.

The Advisory Council for Medical Schemes, and virtually all schemes, were of the view that there should be mandatory membership of medical schemes. The basis of the view was that this would ensure the effective spread of risk by admitting the ill and old, as well as the young and healthy, which would result in schemes becoming increasingly viable.

The Reinach Committee, however, did not recommend mandatory insurance on the basis that "mandatory membership by means of legislation was not necessary because most existing schemes had, on their own initiative, moved to mandatory membership." Although it was noted by the Reinach Committee that 55 schemes offered additional benefits such as sick leave payments, mortality benefits, funeral costs and disability insurance, it was felt that medical cover should be dealt with separately.

The Snyman Commission, which incorporated the recommendations of the Reinach Committee, reported in 1962. Some of the important comments and recommendations were:

There should be fixed tariffs for medical services for all groups of patients. A number of benefits were expected: This would have the positive effect that households could, to a greater extent, budget timeously for medical expenses. With fixed tariffs, expenses to suppliers would be stabilised and the spread of cost maximised. This system would also limit the State’s need to further enter the field of medical services. It would preserve the patient-doctor relationship and trust.
With the imposition of fixed tariffs, high income patients should pay higher premiums to cross-subsidise low-income patients.
To preserve personal initiative, own choice and variety, and to prevent possible disruption of the present system, the institution of a single (national) scheme should be avoided. The development and stabilising of present institutions and the founding of new benefit funds was preferred.
A Central Fund should be instituted to which each scheme should contribute in order to make provision for extraordinary expenses not covered by the schemes.
At least half the managing bodies of schemes should consist of members of the scheme which were elected by the scheme.
No scheme must be allowed to offer additional benefits that are not of a medical nature.
No scheme must be registered if it does not make provision for dependants of members.
Pensioners and widows of deceased members with their dependants must be allowed to continue their membership against premiums that are the same as other member’s premiums in the same group.
If a person was a member of a scheme and has changed job, they must be allowed as a member by the scheme of the new employer without preconditions such as waiting periods, age restrictions or medical reports.
There is a need for cover for smaller groups or individuals who are not in a position to make arrangements for medical insurance.
If an assurance company undertakes medical insurance, it must be subjected to the same legislation as other independent schemes.
New legislation must be created in which the relevant financial and other provisions of the Friendly Societies’ Act, No. 25 of 1956, are to be absorbed. Such legislation must also make provision for the establishment of a central council for schemes and the necessary machinery to give effect to the legislation under the authority of the Department of Health.
The reasons for the rising cost of medical expenses are so inherently connected with the quality and quantity of service that they will, to a great extent, prevail in the future. The final conclusion is that global medical expenses will remain high and might even rise further.

These recommendations, and subsequent debate, resulted in the Medical Schemes Act, No. 72 of 1967.

Supervision under the first Medical Schemes Act: 1967

The intentions of the Medical Schemes Act, No. 72 of 1967 ("the 1967 Act") were (Hansard, 1967):

To invent an insurance type of scheme to distribute the costs of medical expenses over a period of years;
To retain the costs of medical expenses at a low level; and
To co-ordinate and control the functioning of medical benefit and medical aid funds and to develop and propagate these schemes.

The initial Medical Schemes Act resulted in the creation of two important bodies. The first was the Central Council for Medical Schemes, the functions of which were to:

Control, promote, encourage and co-ordinate the establishment, development and functioning of medical schemes;
Investigate complaints and settle disputes in relation to the affairs of registered medical schemes; and
Perform such other functions as may be prescribed.

In addition the Act allowed for the appointment, by the Minister of Health, of a Registrar of Medical Schemes who would perform the duties assigned to the position by the Minister or the Secretary for Health.(Original Medical Schemes Act, No. 72 of 1967).

^{The Regulation of Tariffs and Payments: 1968 to 1986

Until this time, much emphasis had been placed on the regulation of tariffs set with the medical profession. The setting of medical fees between medical schemes and the medical profession was always a problem and a source of conflict. The Medical Association often objected to the fees that were set and the arbitration mechanism. This resulted in many doctors choosing to opt out of the tariff of fees system. If a medical practitioner was contracted in, then payment of the account was guaranteed by law. This provided an incentive for doctors to remain contracted in.
In order to resolve this conflict, a Remuneration Committee was set up in terms of the Amendment Act, No. 95 of 1969, to investigate the tariff of fees at least every two years. The objective of this amendment was to improve the arbitration mechanism such that disputes would not result in further doctors choosing to opt out of the tariff of fees system which was regarded as damaging to doctor/patient relationships.
However, the medical profession eventually regarded the Remuneration Committee in a negative light. Allegations were made that the Act was being used to control the medical profession and that the inflexible provisions relating to the Remuneration Committee were financially prejudicial to medical practitioners and dentists. By 1978 the Dental Society and the Medical Association indicated that they were no longer prepared to participate in the activities of the Remuneration Committee. Consideration had been given to regulating against the free choice of doctors to contract out. However, publication of draft legislation to this effect resulted in a further 1,600 medical practitioners deciding to contract out. By this time 3,941 out of a total of around 14,000 medical practitioners had already contracted out.
As a consequence of these conflicts, the Amendment Act, No. 51 of 1978, abolished the Remuneration Committee and the Commission that made recommendations to the Council on fees. Provision was made for the Medical and Dental Council to determine fees. This was allowed on condition that it prevent further contracting out. If not successful the Minister would step in to regulate the ability of the medical profession to contract out.
The Amendment Act, No. 42 of 1980, made provision for contracted in doctors to send accounts directly to medical schemes. This issue had been a constant source of conflict between medical practitioners and government. The previous dispensation only allowed accounts to be sent to patients who had to pass them on to the medical scheme. Medical practitioners argued that this caused extensive delays and reduced the benefit for contracted-in doctors of guaranteed payment.
However, the Browne Commission (1986) recommended very strongly in its interim report that the provision allowing direct payment be scrapped and that the doctor send the first and second account to the patient and only the third directly to the medical scheme. Upon receipt of the account, the scheme was required by law to pay within six weeks.
The Amendment Act, No. 59 of 1984, effectively eliminated the principle of contracting-in and contracting-out. Any profession or supplier of a service was allowed to determine its own tariffs through their respective statutory control bodies. The Representative Association of Medical Schemes (RAMS) was allowed to determine a scale of fees after consultation with representatives of suppliers of services. If a service supplier were to charge fees equal to or less than the fees indicated on the scale of benefits, the medical scheme was required to pay the supplier of the service directly, provided the scheme offered that benefit.
The "Free Market" Reforms: 1984 to 1988}

By 1980 it began to be recognised that there were too many medical schemes with a consequent inadequate spread of risk.

The amendments to the Medical Schemes Act in the Amendment Act, No. 59 of 1984, had the following objectives:

To have a health service which the ordinary person will be able to afford;
To achieve optimal security of cover by medical schemes to save their members from a financial catastrophe in times of serious or lengthy illness;
To create and maintain, in the interest of medical care, the best possible doctor/patient relationship; and
To prevent the socialisation of health services.

The prevention of a "socialised health system" became a recurrent theme in many parliamentary debates over amendments to the Act toward the end of the 1970s and for most of the 1980s.The extension of the private sector was seen as an important mechanism for preventing the government from having to take on any direct additional burdens with respect to healthcare. By this stage government still regulated the minimum benefits which all schemes were required to provide.

The Browne Commission^,which reported in 1986, developed a free-market theme where it saw the public interest served through the gradual privatisation of the public health service. This was followed by a White Paper on the Commission which largely accepted its recommendations, and set the scene for the Amendment Act of 1988.

The key recommendations accepted by the government were:

Benefit Schemes should not be compelled to become aid schemes (7.1.1.(2)).
In-house schemes should not be compelled to dissolve or to join other or larger schemes (7.1.1.(3)).
Schemes exempted in terms of the provisions of section 3 of the Medical Schemes Act should not be compelled to register under the Medical Schemes Act (7.1.1.(4)).
The operation of the free market should determine the number of schemes (7.1.1.(5)).
That 25 percent of contributions may be an excessive level of reserve to cater for increase in tariffs, but that it should be maintained as a guide for the time being: Provided that the Central Council for Medical Schemes consider this matter. Additional reserves for pensioners should, however, be established according to the circumstances of the scheme (7.3.1 (1)).
Medical schemes should consider expanding their range of services by running their own private hospitals and dispensaries (7.5.1. (2)).
The establishment of an additional type of medical cover for insurance for members who exceeded the annual maximum benefit be considered. (7.6.1).
All employers who have not yet made provision for their employees to join some registered medical scheme should now seriously consider doing so. (7.9.1 (3)).
A major effort should be made to make suppliers of services aware of the fact that they can play an important role in curbing over-utilisation of services and have a great responsibility in this regard (7.12.1(2)).
The market for medical coverage should attempt to devise a voluntary insurance system with a view to covering additional medical costs. If a satisfactory system can be devised the necessary statutory amendments should be made to make the establishment of such a system possible. (7.15.1(2) and (3)).

Key recommendations not accepted by government were:

The government rejected proposals that compulsory minimum benefits be removed on the grounds that "otherwise those members who do not have minimum cover would simply turn to the State for assistance." (7.6.2)
The Government rejected the idea of a scheme being set up purely to cater for the aged on the grounds that "there is already sufficient provision for medical services for such aged persons. Owing to the high claims experienced in the case of aged persons and the fact that no employer’s contributions will be payable in respect of them, the Government believes that the proposed scheme would result in an escalation rather than a saving in costs."

The Browne Commissionmade various suggestions supporting the application of risk-rating and experience-rating within the Medical Schemes Act :

"Greater flexibility in contribution rate determination should be allowed, enabling schemes to charge different contribution rates for different classes of risk. Provision could also be made for allowing different levels of benefit to be chosen by groups or individuals to satisfy their needs. This will encourage merging of small schemes with larger ones, resulting in increased administrative efficiency. In some cases significant cost savings could be achieved if the member paid small claims himself and was only allowed to claim from the scheme after a specific amount had been paid by himself."

The Browne Commission supported a "free-market" approach to health service provision in South Africa, with the public sector only taking responsibility for indigent patients. It supported the development of health insurance as a complementary form of private health cover, as well as the development of group-related underwriting, i.e. risk-rating. However, it also supported the extension of compulsory medical aid cover through employers although no concrete recommendations were made in this regard.

Measures to curb provider induced cost/expenditure increases were very soft with no concrete measures recommended. Recommendations concerning the introduction of risk-rating were seen as a measure to encourage the merging of smaller into larger schemes.

The Commission regarded risk-rating as having the potential to achieve significant cost saving through making the consumer bear the cost of low-cost high-frequency claims (general practitioner visits). Although these recommendations were apparently not accepted in the White Paper, they were nevertheless introduced in 1989 amendments to regulation.

The "Freedom" to Risk Rate: 1989 to 1994

Prior to 1989, a medical scheme registered under the Act was only entitled to vary the rate of members’ contributions based on their income and their number of dependants. In 1989 a significant modification to the environment was introduced in the form of a change to Regulation 8 of the Medical Schemes Act.

According to the modified regulation, a member’s contributions could be based on: number of dependants; income level; age; geographic area; actual claims experience; extent of cover provided; period of membership; and size of group to which member belongs.

This regulation allowed medical schemes to introduce risk-rating into the management of medical schemes, i.e. schemes were free to eliminate existing cross-subsidies within schemes. Although this was of little immediate significance to employer-based medical schemes, open schemes were affected.

The rationale for the change in direction in thinking was clear : the authorities regarded moral hazard on the part of the consumer as the most important variable in achieving cost-containment. In addition the insurance industry was beginning to perpetuate the view that cross-subsidies within health insurance are unfair.

The Amendment Act, No. 23 of 1993. No Act in 1994 to the Medical Schemes Act introduced further far-reaching changes in legislation. Statutory guaranteed minimum benefits and guaranteed payment for claims were removed from the Act. Schemes would be able to exclude or limit cover for procedures, and risk-rate to a greater extent. However, the Act also gave medical schemes the ability to directly supply healthcare services for members. In essence medical schemes were allowed to own and run clinics and hospitals, employ doctors, nurses, physiotherapists and pharmacists.

Two alternative tendencies were being permitted by the regulatory structure. The first offered medical schemes, through their administrators, the opportunity to compete with insurance products on the basis of risk-rating. Here cost containment occurs through attempting to control consumer behaviour. The second allowed schemes the opportunity to contain costs in the longer term through gaining greater control over the supply of medical services.

An amendment was also introduced allowing schemes to "provide additional cover for members by way of insurance, reinsurance or in any other manner whatsoever or, subject to the provisions of any law relating to insurance, underwrite or provide for such cover". (20B(5)(d). The consequence of this amendment may have been that insurance products got channelled through medical schemes acting as fronts for insurance companies and removing their reserves as profits.

The history of the medical schemes movement and its regulation, shows a drift from solidarity principles which defined the original schemes, to individualising health cover. To some extent this drift was slowed due to the predominance of employer-based schemes which indirectly provided compulsory membership on all employees above a particular income level. Where schemes were able to protect their membership base in this way, insurance products lacked influence. However, at no stage in this development had expenditure trends shown any abatement. From 1989 real per capita expenditure grows more steeply than during the community rated period, while beneficiary increases slow down. The slow-down in beneficiary growth is related to contributions growing faster than real incomes. (Figure 2.2).

Figure 2.2: Registered medical schemes: Per capita real expenditure and changes in beneficiaries

Source: Council for Medical Schemes, Compiled by the Centre for Actuarial Research, UCT

In 1994 the Melamet Commission reported and recommended further deregulation on the basis that insurance products represented the best way of providing health cover. It also recommended that all health cover be governed by a single Act, and that the remaining legislation preventing insurance products from being registered as medical schemes be removed. These recommendations, if implemented, would represent the final stage of a shift from medical schemes to insurance as a way of providing health cover within the private sector.

The philosophy supporting this trend suggested that there was no market failure within the healthcare market, except that resulting from moral hazard on the part of the consumer of healthcare. It recognised no strong requirement for an agency function on the part of the third-party payer, and regarded market forces as the primary factor that would achieve a socially desirable outcome.

The trends that emerged from the 1993 deregulation, however, do not support these views. After 1993 a significant shift occurred within the medical schemes environment, with de-facto for-profit open schemes (medical schemes operating as conduits for insurance companies) strongly incentivising broker organisations to cannibalise restricted membership schemes. From 1993 to 1999 the percentage of people covered in open schemes changed from 49 percent to 71 percent. (Figure 2.3). At the same time substantial increases in administration and other non-medical expenditure begin to occur. Whereas in 1992 non-medical expenditure averaged less than 6 percent of scheme Gross Contribution Income (GCI), by 2000 it grew to more than 13 percent of GCI. (See figure 2.4). From 1992 to 1999 (8 years) there was a 243.5 percent real increase in non-health expenditure with only a 6.5 percent increase in beneficiaries.

Figure 2.3: Medical Scheme Beneficiary Changes for Open and Closed Medical Schemes, 1990 to 1999

Source: Council for Medical Schemes

Figure 2.4: Medical Scheme Beneficiary Changes for Open and Closed Medical Schemes, 1990 to 1999

Source: Council for Medical Schemes

The additional non-medical expenditure emerging within the medical schemes environment apparently reflect profit-taking from schemes (via reinsurance and higher administration fees) and very high and hidden commissions paid by administrators to broker organisations. Whereas in the past higher medical costs caused the increasing costs for schemes, from 1993 the ability of "commercial" open schemes to charge higher tariffs within what appears to be a largely price inflexible environment resulted in an additional layer of cost added to scheme contributions.

Overall, however, during this period benefits declined and the older and sicker membership were excluded from cover to a greater extent. By 1999 no open scheme was permitting anyone over the age of 55 to join as an individual member. Virtually all open schemes applied life-time exclusions for pre-existing conditions, and age-rated and/or experience rated their membership without restriction. As such, by 1999 the majority of medical scheme membership were in an environment which excluded vulnerable groups from cover (e.g. the old and those with chronic conditions), where medical costs continued to rise (due to the retention of fee-for-service reimbursement) and where non-medical costs were driven up (through profit taking and hidden commission costs).

Figure 2.5: Non-medical Cost Trends from 1993 to 2000 (Rands)

Source: Council for Medical Schemes

The net result of the 1989 and 1993 deregulation was a significant increase in cost, a general reduction in benefits within schemes, and the virtual elimination of cover for vulnerable groups within the open scheme environment. These trends are consistent with international experience with voluntary health insurance markets.

Returning to Social Solidarity: 1994 to 1999

The direction recommended by the Melamet Commission was rejected by the new Government and replaced by a strategic direction which emerged from the 1995 National Health Insurance Committee of Inquiry. Although the focus of this report was on a system of National Health Insurance, medical scheme reform featured prominently.

Policy directions that were supported by the analysis adhered to the following four objectives:

The regulatory structure should reinforce the agency function of the third-party payer. This was seen as a fundamental requirement for empowering the consumer of health insurance and healthcare.
In order to limit confusion in the market, the regulatory structure should reinforce uniformity in the benefit structure of medical schemes. This would enable people to make effective decisions in their own favour.
Schemes should operate on the basis of solidarity, i.e. that groups do not get treated differently within a scheme. This remains a structurally rational manner in which to provide coverage.

The overall system should create a rational system of risk-sharing between as large a group as possible and, in the longer-term, ensure the availability of a minimum level of cover for all within the public and private sectors.

The recommendations flowing from the analysis of industry issues were largely incorporated in the Medical Schemes Act, No. 131 of 1998. The 1998 Act came into effect on February 1999 and key Regulations under the Act came into effect on 1 January 2000.

One year after the introduction of the reforms real per capita medical costs showed virtually no real increase (0.5 percent), while non-medical costs increased by more than inflation (4 percent). Thus the trend in increased non-medical costs continued as before. The effect of the new reforms on medical cost increases therefore appears to be deflationary. However, savings on medical costs were not passed on to members. Increased profit taking, through administration fees and quota share reinsurance (used to move the underwriting surplus of a scheme to the administrator as profit), saw virtually all the improvements from cost-containment disappear as gains to intermediaries (administrators, brokers, managed care companies). (See figure 2.6).

Figure 2.6: Medical Scheme Real Cost per Beneficiary and Benefit Trends, 1993 to 2000

Source: Council for Medical Schemes

(Note: During 2000 certain previously exempted schemes were required to register as medical schemes. This amounted to an additional 500,000 beneficiaries becoming part of the reporting system affecting medical schemes. These were mostly closed schemes covering certain public sector groups or large parastatals, e.g. Transmed, Medcor, etc.. The analysis in this section excluded these new groups to allow comparison of comparable data through time.)

Concluding Remarks

Until 1994 the health system was splitting markedly into a public sector focused exclusively on the indigent or those without medical scheme cover and private sector focused on the young and healthy employed population. The trends were well established by 1994 with a need for substantial intervention to change direction.

The period of medical scheme risk-rating indicates suggests that underlying medical cost increases were not affected by cost-shifting onto individuals. In fact the reverse is true. Increased underwriting and risk-rating increased the growth in private expenditure, slowing down the growth in medical scheme membership.

Stakeholder Views

Overview

The views of a range of stakeholders were obtained on various aspects of the health system and possibility of some form of mandatory contribution for health cover. On the whole dissatisfaction was expressed on the current public health service. Many groups, including employers, supported the strengthening of the public sector, particularly access to hospital services, as important for the future. A willingness to contribute over-and-above existing contributions was expressed by many, but only on condition an improvement in the public service occurred first. Many supported the idea of some form of enhanced amenity within the public sector for contributors. There was however limited support for enhanced services offered on a differential basis.

Willingness to pay

Many groups are willing to pay a small fee (pre-paid) provided the public sector improves its services first. Opposition to payment of such a fee did not emanate from potential contributors, but rather from the national Treasury (blanket opposition to earmarked taxes) and certain (but not all) trade unions (near-poor should not cross-subsidise the very poor). The latter trade unions supported a universal earmarked tax provided they could choose to fund these services in the private sector. Other trade unions supported the idea of a contribution provided they received something in return. They were supportive of these services being in the public sector. Evidence of the functioning and benefits of a low cost contributory system is provided by the clothing workers.

A willingness and ability to pay survey (W&A study) conducted by the Department of Health sampled the views of around 1,000 individuals of varying incomes (figure 3.1) on various aspects of their willingness and ability to contribute to a social health insurance fund focusing on public hospital services.

Figure 3.1: Willingness and Ability to Pay Study – Distribution of Respondents by Income

Source: Department of Health, August 2001.

Over 94 percent of respondents in the W&A study felt it was appropriate to pay for public hospital services. Up to 45 percent felt that public hospital services would improve if there were some form of additional payment. Another 36 percent felt people should pay for what they use. (Table 3.1).

Table 3.1: Willingness and ability to pay study: Reason for willingness to pay for public hospital services

Reason	Resp	%
Public Hospitals provide value for money	40	4.7
I think people should pay for what they use	305	36.0
Services will improve at public hospitals if we all pay something	382	45.1
I believe in Masakhane (civic duty) so it is our duty to pay	77	9.1
You have to pay otherwise they send you a lawyers letter demanding payment.	14	1.7
Other (please specify)	30	3.5
Total	848	100.0

Source: Department of Health, August 2001.

Ability-to-pay

Interviewees and stakeholders were not able to give a clear comment on this issue as no contribution level was put to them. However, it was felt that if a pre-paid contributory system were created they would probably voluntarily pay if they had access to improved services.

The W&A study found that 77 percent of respondents were able to pay when last receiving care at a public hospital. Given that 61 percent of respondents come from the income groups R0-R2,000 per month, this indicates that a significant number of low-income groups are able to pay at least something toward their health care.

Table 3.2: Willingness and ability to pay study: Able-to-pay when last receiving care at a public hospital

Able To Pay	Resp	%
Yes	608	76.9
No	161	20.4
Cant Remember	22	2.8
Total	791	100.0

Source: Department of Health, August 2001.

The W&A study suggests that 73 percent of respondents are willing to use and pay for public hospitals.

Table 3.3: Willingness and ability to pay study: Willingness to use and pay for public hospital care

Opinion	Resp	%
Willing to use Public Hospital, not willing to pay	56	5.6
Not willing to use public hospitals	166	16.5
Willing to use Public Hospitals, willing to pay	732	72.6
Don’t Know	55	5.5

Source: Department of Health, August 2001.

Earmarked tax

There were differences among government officials as to how they would understand an earmarked tax for health services. The national Treasury saw these as part of the general tax system, and therefore any earmarked tax will have to be off-set by a reduction of budget. Health officials saw an earmarked tax as replacing some tax funding, but also providing new funding. The rationale for new funding arises from the following:

The willingness to make an additional contribution;
The recovery of funds that should have been raised from the point-of-service billing of existing users; and
The need for full-cost recovery for new users of the public system.

The W&A study found that a significant proportion of the population (90 percent) interviewed were willing to accept a compulsory system of public hospital cover if services were improved. (Table 3.4).

Table 3.4: Willingness and ability to pay study: Support for compulsory membership if Public Hospital Insurance

	No		Yes			Don’t Know
	Resp	%		Resp	%		Resp	%
Support if the public hospitals stay as they are	903	89.5		96	9.5		10	1.0
Support if public hospitals are improved	85	8.4		908	90.0		16	1.6
Support if scheme members get differential treatment	402	39.8		567	56.2		40	3.1

Source: Department of Health, August 2001.

If no services were improved only 9.51 percent were willing to contribute. The introduction of mandatory cover of any form must involve a discernable improvement in hospital services. The results are similar where a payroll deduction is proposed (table 3.5).

Table 3.5: Willingness and ability to pay study: Support for compulsory Payroll deduction for covering public hospital costs

	No		Yes			Don’t Know
	Resp	%		Resp	%		Resp	%
Support if the public hospitals stay as they are	878	87.0		110	10.9		21	2.1
Support if public hospitals are improved	104	10.3		874	86.6		31	3.1
Support if scheme members get differential treatment	418	41.4		533	52.8		58	5.8

Source: Department of Health, August 2001.

Overall 55.9 percent of respondents in the W&A study felt that members of medical schemes should be excluded from any mandatory payroll deduction versus 34.8 percent who thought they should. (Table 3.6).

Table 3.6: Willingness and ability to pay study: Support for payroll deduction with exemptions for Medical Aid members

	No		Yes		Don’t Know
	Resp	%	Resp	%	Resp	%
Support for payroll deduction with exemptions for Medical Aid members	351	34.8	564	55.9	94	9.3

Source: Department of Health, August 2001.

Funding of the Public Sector:

Apart from the national Treasury, there is a general consensus (employers, trade unions and workers) that the public sector is under-funded which encourages all who can pay to use private sector services.

Tiering (differential amenities or "buy-up options in public sector hospitals):

Although there was some variation in the responses from trade union members, there was a large degree of support for differential amenities. There was however no support for differential services. Table 3.7 reports the responses on the W&A study toward differential amenities.

Table 3.7: Willingness and ability to pay study: Attitude towards a differentiated public health service

	Strongly Agree		Agree			Unsure		Disagree			Strongly Disagree
	Resp	%		Resp	%	Resp	%		Resp	%		Resp	%
Payers should be treated First	185	18.3		204	20.2	47	4.7		422	41.8		151	15.0
Payers should get nicer Wards	163	16.2		287	28.4	40	4.0		424	42.0		95	9.4
Payers should be able to make appointments	1	0.1		218	21.6	365	36.2		53	5.3		296	29.3
Payers should have TV's in their rooms	133	13.2		316	31.3	102	10.1		369	36.6		89	8.8
Payers and non payers should get same care	358	35.5		294	29.1	109	10.8		162	16.1		86	8.5
Won't use public hospitals regardless of improvements	45	4.5		33	3.3	92	9.1		414	41.0		425	42.1

Source: Department of Health, August 2001.

Improvement of public sector services

Employers, union representatives and workers indicated that reasonable improvements in the public sector will probably result in their shifting away from expensive private cover. Some unions were adamant that improvements to the public system should precede any introduction of a contributory system. Key problems raised were:

Shortages of medicine;
Poor physical condition of facilities;
Facilities are not clean;
Poor service delivery;
Rude staff;
Lack of doctors at clinics; and
Insufficient staff.

Injection of funds

Quite a few respondents (including employers and trade unions) raised the issue of a one off injection of funds to provide a face-lift to public sector services to initiate a contributory system.

Phasing

There was universal support for a phased approach to implementation with an initial focus on creating a voluntary contributory environment. This could either be via a voluntary SHI or a low cost medical scheme. A few supported the idea that the process should begin with public sector employees.

Revenue retention at facility level

The inherent logic of revenue retention at the facility level was accepted by all groups, including the national Treasury. However, there was uncertainty amongst other government officials concerning the true position of the Treasury Department.

Benefits

Employers felt that contributions and not benefits should be defined. Certain trade unions felt they should be allowed to opt for private primary care and a contribution toward public hospital service. Certain trade unions did not want any restriction on their choice of service provider.

Findings and Concluding Comments

There is an understandable variation in what different stakeholders understand by social health insurance. There nevertheless appears to be a fair degree of consistency in how stakeholders interpret their preferences in relation to a number of alternative options.

The following arise from the results:

Willingness to pay: Many groups are willing to pay a small fee (pre-paid) provided the public sector improves its services first. Opposition to payment of such a fee did not emanate from potential contributors, but rather from the Treasury Department (blanket opposition to earmarked taxes) and certain (but not all) trade unions (near-poor should not cross-subsidise the very poor). The latter trade unions supported a universal earmarked tax provided they could choose to fund these services in the private sector. Other trade unions supported the idea of a contribution provided they received something in return. They were supportive of these service being in the public sector. Evidence of the functioning and benefits of a low cost contributory system is provided by the clothing workers.

Ability to pay: Interviewees and stakeholders were not able to give a clear comment on this issue as no contribution level was put to them. However, it was felt that if a pre-paid contributory system were created they would probably voluntarily pay if they had access to improved services.

Earmarked tax: There were differences among government officials as to how they would understand an earmarked tax for health services. The Treasury Department saw these as part of the general tax system, and therefore any earmarked tax will have to be off-set by a reduction of budget. Health officials saw an earmarked tax as replacing some tax funding, but also providing new funding. The rationale for new funding arises from the following:

The willingness to make an additional contribution;
The recovery of funds that should have been raised from the point-of-service billing of existing users; and
The need for full-cost recovery for new users of the public system.

Funding of the Public Sector: Apart from the Treasury Department, there is a general consensus (employers, trade unions and workers) that the public sector is under-funded which encourages all who can pay to use private sector services.

Tiering: Although there was some variation in the responses from trade union members, there was a large degree of support for differential amenities. There was however no support for differential services.

Improvement of public sector services: Employers, unions representatives and workers indicated that reasonable improvements in the public sector will probably result in their shifting away from expensive private cover. Some unions were adamant that improvements to the public system should precede any introduction of a contributory system. The key problems raised were:

Shortages of medicine;
Poor physical condition of facilities;
Facilities are not clean;
Poor service delivery;
Rude staff;
Lack of doctors at clinics;
Insufficient staff.

Injection of funds: Quite a few respondents (including employers and trade unions) raised the issue of a one off injection of funds to provide a face-lift to public sector services to initiate a contributory system.

Phasing: There was universal support for a phased approach to implementation with an initial focus on creating a voluntary contributory environment. This could either be via a voluntary SHI or a low cost medical scheme. A few supported the idea that the process should begin with public sector employees.

Revenue retention at facility: The inherent logic of revenue retention at the facility level was accepted by all groups, including the Treasury Department. However, there was uncertainty amongst other government officials concerning the true position of the Treasury Department.

Benefits: Employers felt that contributions and not benefits should be defined. Certain trade unions felt they should be allowed to opt for private primary care and a contribution toward public hospital service. Certain trade unions did not want any restriction on their choice of service provider.

The Achievement of Equity within the Health System

Overview

The goal of equity within the health system is regarded as fundamental and recognised universally as a cornerstone of health policy. However, equity needs to be clearly defined in order for it to have practical value in the determination of health policy.

The distribution of health resources within South Africa, both within the public sector and between the public and private sectors, demonstrate large variations that could be regarded as inequitable. Government policy has to determine when this variation acceptable and when unacceptable.

This section reviews the issue of equity and how an understanding of it can be used to underpin Government policy from a practical point of view.

Defining Equity

Review

Amongst the countries of Western Europe there is general agreement that the provision of health care services cannot be left to the unregulated market. If it were, health care would become very expensive resulting in significant and unacceptable gaps in insurance coverage. (Able Smith, 1992, p.217).

Able-Smith (1992, p.218) lists the following specific issues upon which there is consensus amongst the countries of Western Europe:

Nobody is denied any important health care because of inability to pay. Dentistry, other than emergency dentistry, and optical care are often regarded as less important services, at least for adults, for which they can self-fund.
With the possible exception of higher income groups, health insurance is prevented from developing risk-rating, either according to individual risks, or according to number of an insured person’s dependants. Health insurance deliberately avoids applying strict actuarial principles.

National health insurance is very different from private health insurance. With the possible exception of high-income groups, health services for the compulsorily insured are not left to the functioning of the unregulated market because three vital elements for the functioning of such a market are missing. The first is informed consumers, who know precisely what they want to buy. Secondly, the need for health care cannot be known in advance and, when it comes, it can be very expensive. The third is the lack of separation between the functions of authorising purchase and supplying it. Consequently, health services are prepaid by some mix of taxes and health insurance contributions.
A complex mix of regulation and control has emerged in each country with differences as to how each aspect is applied.

According to Roemer (1980, p.190), in spite of the counteracting pressures of an entrepreneurial ideology in all countries, the long-term trend appears to be in the direction of achieving greater equity through rational organisation. In other words, the entrepreneurial pressures for regarding health care as a commodity, i.e. a consumer good to be bought and sold in the market place, are gradually being overcome by the social pressures for the distribution of services according to concepts of equity, that is, according to each person’s human needs.

All 32-member countries in the World Health Organisation (WHO, 1985) European Region adopted a common health policy in 1980, followed by unanimous agreement on 38 regional targets in 1984. The first of the targets is concerned with equity:

"By the year 2000, the actual differences in health status between countries and between groups within countries should be reduced by at least 25 percent, by improving the level of health of disadvantaged nations and groups."

Whitehead (1992) produced a discussion paper on the "concepts of equity and health" as part of the programme on Equity in Health in WHO’s Regional Office for Europe. The paper represents an attempt to create a practical tool for policy-makers through distilling the collective wisdom obtained from the Equity in Health programme. The intention was to establish a working definition of equity as understood within the context of WHO’s Health for All Policy. The conclusions of this paper and the inter-disciplinary advisory group that reviewed the initial draft are summarised below.

According to Whitehead, the term "inequity" as used in WHO documents refers to differences in health that are not only unnecessary and unavoidable, but in addition, are considered unfair and unjust. An important criterion used to determine which situations are unfair is the degree of choice involved. Where people have little or no choice of living or working conditions, the resulting health differences are more likely to be considered unjust than those resulting from health risks that were chosen voluntarily. The sense of injustice increases for groups where disadvantages cluster together and reinforce each other, making them very vulnerable to ill-health. From this view Whitehead arrives at the following working definition:

Equity in health implies ideally everyone should have a fair opportunity to attain their full health potential and, more pragmatically, that none should be disadvantaged from achieving this potential, if it can be avoided.

Based on this definition, according to Whitehead, the aim of policy for equity and health is not to eliminate all health differences so that everyone has the same level and quantity of health, but rather to reduce or eliminate those which result from factors that are considered to be both avoidable and unfair. Consequently, equity is concerned with creating equal opportunities for health, and with bringing health differentials down to the lowest level possible.

Whitehead criticises some potential definitions of equity on the basis that they do not satisfy a common sense understanding of equity. For example health services could be based on equal per capita expenditure. However, if this objective were achieved it would make no allowance for differential needs for care in different age and social groups in each region. Consequently such a definition could not be considered equitable. At the other extreme equity could be defined as the achievement of equal health status between all communities, age groups and social groups. This view is criticised as variations in health care services are only one of the many factors contributing to health differences between communities.

Based on the earlier working definition, and accounting for the short-comings of alternative definitions, the following definition is adopted by Whitehead:

Equal access to available care for equal need: This implies equal entitlement to the available services for everyone, a fair distribution throughout the country based on the health care needs of access in each geographical area, and the removal of barriers to access.

Equal utilisation for equal need: Care needs to be taken in interpreting this goal. If differences are found in the rates of utilisation of certain services by different social groups, this does not automatically mean that the differences are inequitable. Rather it is an indication that further research is needed to ascertain why the variations exist.

Equal quality of care for all: It is important that every person has an equal opportunity of being selected for attention through a fair procedure based on need rather than social influence. This issue is most important when resources are scarce or are being cut back. In such a climate it will appear unfair to many if members of particular, social, ethnic, or racial groups were consistently regarded as the lowest priority when it came to medical treatment. Equal quality of care for everyone also implies that providers will strive to put the same commitment into the services they deliver for all sections of the community, so that everyone can expect the same standard of professional care.

Equity Subject to a Budget Constraint

The emerging consensus is for a substantial quantity of health care services to be provided on an equitable basis. However, all systems face a budget constraint. Constant improvements in medical technology may effectively increase the needs that can be satisfied using health care. The introduction of new technology often adds costs to a health system without making a major contribution to health outcomes. Furthermore, new interventions, being expensive, cannot be afforded for an entire population, especially in developing countries. In developing countries, given their budget constraint, this often obstructs the development of priority services, as the new technology drains financial resources, which could contribute to a deterioration in the population’s health.

Mahler (1975) draws attention to the increasing disparity between the tendency to expand health care coverage, often to universal access, and the restrictive application of high technologies to specialised curative services. In other words, the higher the peak becomes, the more difficult it is to provide a universal system. Furthermore, the creation of a peak directed towards the few is selected not so much by social class or wealth, but by medical technology itself. (Kleckowski, 1980, p.101).

Consequently, the achievement of equity within the context of a budget constraint implies the conscious application of a limit on the services that are made available on an equitable basis. In addition, the introduction of new services would have to be on the basis that they lower the costs and improve the outcomes of existing interventions. As the wealth of a country increases, it will become feasible to increase the amount of services provided on an equitable basis.

Understanding a Definition Within the South African Context

Creating a practical definition of equity for use in developing policy in South Africa has to focus on defining both principles and mechanisms for achieving the goal of equity taking account of substantial disparities in income and relative needs and demands for health care. On the one hand the system cannot unfairly deny services from groups unable to afford the cost of health care. On the other, certain services cannot be denied to groups who are able and willing to pay for them.

The following are therefore a set of guidelines for the purposes of this report that can aid in establishing a useful definition:

General policy principle:The guiding principle of policy needs to focus on providing an equal entitlement to the available services for everyone, a fair distribution throughout the country based on the health care needs of access in each geographical area, and the removal of barriers to access.

Constraints: There has to be a clear recognition in policy of the existence of resource constraints and how policy has to react to ensure that equity is not compromised through a failure to adequately prioritise services. Government will have failed in its task of achieving equity if it lacks a rational approach to defining what services must be available to all within existing financial constraints. Defining and costing universal minimum service requirements irrespective of income, funder or provider, is an effective way of ensuring that resource constraints are rationally and fairly accounted for in policy.

Income cross-subsidies: Income differentials are a fact of life and must become part of the rational design of health systems. Government has to establish a clear policy on how a system of income cross-subsidies is to be achieved, both in funding the non-contributory sector as well as the private sector. Where higher income groups diminish the availability of health services within the non-contributory public system, policy measures and instruments will have to protect the availability of services.

Health-related cross-subsidies: Different groups have different needs for health services based either on factors such as age or gender, and because of socio-economic differences. Policy responses have to define how it will balance these differences through explicit resource allocation within public sector systems, or ensure cross-subsidisation from healthy to sick within insurance systems.

Strategic Elements of the Health System Affecting Equity

Overall level of funding for the health system

The overall level of funding going to health care is determined partially by government policy and partly by voluntary contributions. Where funding is either tax-based or mandatory, services are largely shared. Where voluntary contributions occur services are provided on an exclusive basis. The overall level of funding within tax-based or mandatory systems can have a significant effect on whether services are accessed on a shared or on an exclusive basis. If funding levels are too low, more services will be available in the exclusive rather than the shared public system. Apart from the equity considerations, this could also result in significant organisational inefficiencies and additional costs.

Government can impact on this area by:

Directly taxing income earners more to fund an increased public service;
Create a mandatory contributory environment in which a greater degree of equity is achieved within specified income groups; and
Permit the use of funds from the voluntary contributory environment to promote the expansion of services in the shared service or public environment.

There is no apparent strategic focus by Government which attempts to understand the implications of higher or lower levels of funding for the public system.

Income-based cross subsidies

Income-based cross-subsidies are generally achieved through the tax system, or mandating insurance in a manner that closely follows normal tax principles. In essence people pay according to their means, but receive benefits according to their needs. The following instruments are important within the South African context:

The level of general tax funding for public services;
Subsidies to the private sector (tax subsidies versus on-balance sheet per capita subsidies);
Contributions to medical schemes (flat-rate versus income-based); and
Mandating contributions to either social health insurance or medical schemes.

The redesign of the income tax subsidy represents the only viable short- to medium-term measure for achieving minimum required income-based cross-subsidies across the entire health system, both public and private.

Health-related cross-subsidies

Health related cross-subsidies are achieved differently in the public sector settings where services are subject to physical planning process, and insurance environments where access is entitlement-based.

The objective of the public sector is to achieve an equitable distribution of services on a regional basis within budget constraints.

Within insurance-based systems there is a need to protect cross-subsidies from those who are healthier to those who are sicker to prevent their systematic exclusion from cover. Health funds (medical schemes) must also be protected from the consequences of having disproportionately sicker groups of people where this arises.

The public sector has the following instruments:

Inter-regional resource allocation: However, the current fiscal federal environment prevents this from being achieved on an inter-provincial basis through national policy. There is also no clear framework for dealing with local government and district services.

Minimum norms and standards: This instrument can be used to create and implement a national policy framework, or to impose conditions on provinces limiting their discretion to allocated funds elsewhere.

Conditional grants: Conditional grants can be used to ring-fence allocations consistent with policy objectives linked to the achievement of equity.

Within the contributory environments the following instruments are available:

Open enrolment: which prevents any individual or group from permanent exclusion from cover;
Community rating of schemes: which prevents exclusion on the basis of health risk status (as contributions are determined on the basis of the average cost of the group and not of the individual); and
Risk-equalisation between schemes: which balances out the implications of uneven distributions of sicker groups between schemes.

Basic essential service and benefits

In order for equity to have practical meaning it must be expressed in terms of actual services or conditions which must be provided on an equitable basis. Policy instruments may differ between public non-contributory and private sector settings. Nevertheless, the principles remain the same.

Public sector

The public sector will have to define minimum services primarily through the establishment of a minimum basic package of services. This can be expressed practically in terms of policy through the establishment of service norms and standards.

Private sector

The Medical Schemes Act No.131 of 1998 introduced prescribed minimum benefits as a policy instrument for defining minimum allowable levels of medical scheme cover. This involves a positive list of conditions and treatments.

Requirements for the future

There is no coherent approach as yet to defining the basic essential minimum services between the public and private sectors. Ultimately both systems will need to provide a minimum core set of services which are consistent with one another. Once rationally defined, Government will have to establish clear mechanisms for ensuring that the desired entitlements can be met in an equitable manner in both settings.

Concluding Remarks

Given that South Africa is a developing country, it has to confront great income disparities, and resource constraints. The set of required instruments for achieving a coherent and integrated system of subsidies needs to cater for complex relationships between and within the public and private sector settings. The nature of health care provision is such that it naturally diverges from equity irrespective of whether publicly or privately funded.

Financial Framework of the Public Health System

Introduction

Although a proportion of revenue raised for government as a whole is not based on general taxes, no consistent set of principles has been established in South Africa as to how these funds should be raised, managed, and related to general tax revenue. There is furthermore no consistent set of principles underlying the manner in which funds are subject to provincial discretion.

The principles underlying the allocation of the budget arising from general tax revenue as well as that arising from alternative sources is consequently reviewed to determine appropriate principles to guide policy with respect to existing and potential health system environments.

Allocation of Funds arising from General Taxes

Roughly 80 percent of the health budget is allocated by provincial governments from an unconditional grant allocated from central government. The other 20 percent is a conditional allocation from central government to cater for the "spill-over" problem resulting from the concentration of secondary, tertiary and teaching services within only a few provinces.

The size of the spill-over effect is a significant proportion of the entire budget. It would be expected that a spill-lover problem should be a fairly small percentage of the total budget. The larger the spill-over the more consideration needs to be given to a consolidation of the jurisdictional reach of the health budget.

Gildenhuys (1993, pp.194-196) comments as follows:

There are few government services, such as defence, which can be rendered at one government level. Because of their nature most services can be rendered at any government level. The question is what factor determines the most suitable government level for the rendering of any service, and how should its financing be arranged?

The primary and most important factor to be considered with the allocation of functions is the extent of the benefits and costs created by the spill-over effect. Because the sole reason of government is the supply of collective and particular public services, as far as possible it is logical to match the extent of the spill-over with the jurisdictional scope of the government making the decisions about that service. This structural idea is called the correspondence principle.

Matching the jurisdictional area with the effective distance of the spill-over excludes the need for a complicated intergovernmental financial relations policy. A mismatch of the spill-over with the jurisdictional area of the government may result in serious misallocations of financial resources.

There are virtually no collective services without any externalities. Neither is there a tax system which can ensure that its impact is limited to the jurisdictional area of the taxing authority. Therefore, a measure of fiscal inequality will always exist. The ideal remains, however, the elimination as far as possible of any fiscal inequalities with a policy of intergovernmental fiscal relations.

Vertical fiscal equality means that all governments at whatever level have sufficient flexible revenue resources at their disposal to pay for the full costs of their services. This inequality is usually the result of an unscientific allocation of functions and revenue sources as provided for in the constitutional legislation of a country. Such an unequal allocation is usually the result of political expedience rather than the result of scientific rationality.

Despite the allocation of an equitable share of the unconditional grant to all provinces equity has not been achieved in the provision of health services. Provincial allocations also fail to keep pace with population increases. The budget allocations to health departments show no consistent correlation with underlying population and equity considerations both of which are central to health policy.

Consistent with theoretical arguments, in most countries the budget for redistributive public services are more centralized than for all public services. There has been a trend since the 1930s for central governments to take additional responsibility for redistributive programmes and to expand their scope and magnitude, with Australia, Canada, Denmark, England, and Sweden all joining the United States as illustrations. (Fisher, 1996, p.591).

Responsibility for social security, welfare, and housing is quite centralized, with federal expenditures accounting for at least two thirds of the total in the four major federal systems (Australia, Canada, Germany, and the United States). In all cases, federal expenditures are a greater share of the total for the broad category of social security, welfare and housing than they are for government purchases in general. Education expenditures are the least centralized of the group, although it is much more centralized in Australia than for the other three countries. (Fisher, 1996,p.592).

The ability provincial governments have to undermine allocations to health services arising either from conditional grants or user fees has been identified by the national Department of Health as a problem. Additional revenue from these sources, which should result in a net increase in revenue over budget, are offset through reductions in the general budget allocations at a provincial level.

To the extent that these reductions are consistent with national health policy no problem arises. However, in reaching these allocational decisions provinces are not required to defer to national health policy. This results in a misallocation of resources from what would occur if national policy were to prevail.

Based on the information reviewed strong consideration needs to be given to a greater degree of centralisation of the health budget. No evidence or rationale appears to exist suggesting the budgets be programmed at a central level. However, the ring-fencing of a significant portion of the provincial allocations after determination at a national level appears consistent with both international practice and the current and future needs of the health system.

Allocation of Funds arising from User Fees

User fees raised by public hospitals are currently not differentiated from general tax revenue. This is inconsistent with the normal treatment of user charges. Typically where user fees have a strong cost-recovery purpose, they are recovered and utilised at source and are not regarded as part of the redistributable income of government. The non-redistributable nature of user charges relates to fact that general taxes have not made financial provision for the service being sold. As such, fee recovery must cover the costs.

The following are recommendations regarding the principles that should be applied to user charges:

In all instances where user charges, consumer tariffs, or levies are charged, separate operational accounts should be maintained by the relevant institution or authority.
Financial accountability should be delegated to the lowest appropriate level where separate operational budgets exist.
Surpluses on all charges should not occur or be accumulated for redistributional purposes. Appropriate mechanisms should be put in place to ensure that surpluses and deficits even out over time.
As far as possible, specific redistributional goals should be achieved through general tax and budget allocations and not via the revenue obtained from dedicated taxes. It would not be inconsistent, however, for certain redistributional goals to be achieved amongst contributors (as opposed to that between contributors and non-contributors). Keeping to these guidelines should ensure that redistributional goals and objectives are transparent and based on clear and rational policy objectives.

Allocation of Funds arising from Earmarked Taxes

Earmarked taxes are important with respect to proposals for a mandatory contributory environment based on a specific contribution to be made to a public fund for the reimbursement of benefits obtained from public hospitals. Although such a proposal clearly does not take the form of general tax it nevertheless has many of the characteristics of a tax. This is related to two key features:

It is mandatory; and
There is a redistribution of income involved.

The justification for an earmarked tax often lies in the application of the exclusion and benefit principles. In exchange for payment, contributors gain access to the services so funded. Non-contributors would be excluded. The application of the exclusion and benefit principles in conjunction with an earmarked tax enhances the willingness-to-pay and improves tax compliance. However, where a new tax is introduced which replaces the funding from a general tax, an offset from general revenues could be considered. Any net improvement in financing would in all circumstances be an explicit policy decision of national Government.

Principles that should be applied with respect to earmarked taxes are:

Earmarked taxes should not be considered as an alternative to the general budget but rather be used only in specific instances where the quasi-public nature of the good or service requires a direct relationship to be established between the contributor and the good or service to be provided. Insurance of one form or another and retirement contributions, where compelled by the state, would fall into this category.
Where earmarked taxes are considered, separate operational budgets are required to ensure consistency between the funds raised and the entitlements to be funded.

Alternative Options for Reform and their Implications

Various alternative directions for reform of the public sector financing framework are possible. For simplicity they are broken down into four types that broadly reflect directions that can be taken.

Option 1 - Budget programmed at the national level

If programmed at national level the health budget will be more easily protected from inappropriate cuts and there will be more influence over provincial administrations. However, the over-centralised approach to programming would diminish some of the effective control over policy implementation and resource allocation. This option is generally weak in respect of decentralised service delivery, although it is possible that this could be overcome. On the whole this option is compatible with both the reforms to the medical schemes environment and any potential future social health insurance option: the centralised allocation of health funds allow a single administrative system for allocating the budget; protect the base-line allocations the health system from being undermined by provincial treasuries; and compatibility between the allocation of funds in respect of the contributory and non-contributory environment can be achieved.

Option 2 - Budget ring-fenced but not programmed at the national level

With the budget ring-fenced but not programmed a high degree of provincial discretion is permitted within any nationally determined policy framework. Operational decisions, including the programming of budgets are fully dependant on provincial, regional and local governments. This approach is compatible with decentralised models of service delivery and greater autonomy at facility level. It is also compatible with the medical schemes reforms and any potential social health insurance approach directed at public hospitals. As with option 1 the development of a single administrative framework for allocating general budget and social health insurance. It also protects the base-line budgets from being undermined by provincial treasuries when increased revenue occurs from medical schemes.

Option 3 - Provincial discretion limited through use of national norms and standards

Where an attempt to ring-fence provincial allocations occurs through the use of national norms and standards a weak form of option 2 results. It does become possible for a national policy framework to be implemented, but its potential effectiveness as a lever is subject to certain difficulties. These relate to changes in norms over time (which now become budget decisions) and enforcement. The risk of unfunded mandates is a potential but avoidable possibility. Compatibility with a policy framework incorporating the contributory system the only key objective achieved is the protection of the base-line allocation to the health system. Options for a unified allocation mechanism for both general tax revenue and social health insurance contributions are not possible.

Option 4 - Provincial allocations with full discretion

This option largely reflects the status quo. A key defect of this option relates to the need for ten health departments to separately motivate for budget allocations. The resulting consolidated allocation is less a result of national policy than it is of the adding up of ten individual bargaining processes tenuously linked to any national framework.

When provinces have full discretion over the allocation of budget they have virtually full discretion over health policy in their region. Links to all areas relating to a national policy framework are weak as there are no associated financial flows. This budget framework shows little compatibility with either existing or future policy environments.

Table 5.1: Evaluation of Alternative Options for Allocating the Health Budget

	Budget programmed at national level	Budget ring-fenced (but not programmed) at national level	Provincial allocations subject to national norms and standards	Provincial allocations with full discretion
Ability to prioritise national resources toward health care	High	High	Medium	Weak
National influence over provincial health policy	Medium/High	High	Medium	Weak
Inter- provincial resource allocation	Medium/High	High	Medium	Weak
Consistency with decentralised service delivery options	Weak	High	High	High
Compatibility with reforms to the voluntary contributory environment (medical schemes)	High	High	Medium/Weak	Weak
Compatibility with the introduction of social health insurance directed at public hospitals	High	High	Medium/Weak	Weak

Overall the option 2 is most consistent with both current and future policy directions of the health system. It is the least disruptive to the current organisational structure and can be introduced in a phased manner. Although there is a need for improved capacity at the national Department of Health, the focus is on strategic allocations linked to policy rather than interference in operational matters. As such the short-term need does not place an onerous burden on the national department. In the medium- to long-term the national Department of Health would have to develop a more coherent institutional framework around provincial financing linked to strategic policy objectives and any conditional allocations linked to public sector social health insurance options.

Option 4 reflects the status quo and is incompatible with both current and future policy directions. It is the weakest of the four approaches. Without significant changes to the current framework, linking provincial policy to national policy in key areas the public health system will probably diminish in importance over time. Although certain social insurance options will be possible despite these arrangements, i.e. mandating medical scheme cover, low cost private sector facilities will probably develop instead of public sector options. Where public sector services are sold into the contributory environment, differential amenities will inevitably become differential services, as private sector funding will be stable while public sector funding will vary.

Recommendation 1

It is recommended that option 2, or some variation thereof, be considered in the short- to medium-term. Such an option could be phased over time with the development of other reforms dependant on the restructuring. Although the implementation of this option is an important pre-requisite for social health insurance and related options incorporating the public hospital system, it is just as important for optimising existing policy objectives.

Recommendation 2 (alternative to recommendation 1)

An alternative approach to the full adoption of option 2 is the use of a mix of options 2 and 3.This would involve a reasonable increase in the value of the existing conditional grant going to public hospitals to a level sufficient to achieve base-line budget protection for public hospital services. This would include the use of variable matching conditional grants.

The matching could vary by province depending upon service needs and national policy. Here the Province is required to match a grant allocation with an allocation of their own. Funds would be allocated only if the matching occurred. This approach would prove important when any central allocation occurred from any social health insurance or related fund. Such an approach would allow a single mechanism to be used to allocate both public sector budgets and social health insurance budgets for public hospitals.

The conditional grant system should be combined with the development of a coherent approach to setting minimum norms and standards for provincial health services. Although in these instances no direct control over the budget allocations will exist, provinces will be required to adhere to minimum levels of service delivery. Provinces would nevertheless be free to offer services in excess of the minimum.

Together these approaches should achieve the objectives of recommendation 1 without full ring-fencing at a national level.

Reform of the Tax Regime and Subsidies for Medical Scheme Cover

Overview

According to Price et al (1995), the debate about the tax deduction arose primarily as a consequence of the funding crisis in the public health sector, and perceived inequality between the public and private sectors.

Employees currently contribute a certain portion of their salary to a medical scheme, with employers also making a contribution on their behalf. The Income Tax Act allows the employer’s contribution to be deducted as an expense before tax. On the employee’s side, a deduction is available only where an individuals medical expenses exceed 5 percent of income or R5,000. For pensioners, all medical expenses are tax deductible.

The Melamet Commission wrote that the tax deduction "encourages consumption of health care beyond the point where the costs of obtaining extra cover equate to the value of the marginal benefits received. Price signals are badly muffled. Medical cost inflation is thus encouraged." (Melamet Commission, 1994, p.44).

Price et al (1995) concluded in their evaluation that given "… the scarcity of health care resources in any country, the prime responsibility of government with respect to funding should be to improve the health care of the poorest in society. The very structure of the private health sector in South Africa goes against this principle, since it distributes health care resources predominantly according to ability to pay. The subsidisation of this sector by the government is not consistent with the principles of health care funding by the state. The current specific concessions allowed in South Africa are furthermore inequitable across income groups with high earners receiving a greater subsidy than low earners on medical aid, while self-employed individuals (including the whole informal sector) receive almost no subsidy at all."

Value of the Tax Deduction

A 1995 study (Price et al, 1995) estimated the net impact on the central revenue through the removal of the employer tax deduction. The analysis took account of the various offsetting influences. The study also assessed the extent to which an inflexible demand for medical scheme cover will impact differently on the Central Revenue Fund. The least flexible is 0 percent where the individual reduces private health consumption by the value of the lost subsidy. The assumed marginal tax rate used is 32 percent. Scenario 1 assumed the employer contribution is 50 percent while scenario 2 puts it at 75 percent. The analysis does not include the tax subsidy for out-of-pocket expenses, exempted medical schemes and pensioners. The analysis suggests that the net impact on government revenue of any reduction in the tax subsidy would range from R3,2 billion to R5,9 billion. Given the inelastic demand for medical scheme cover, the true impact could be of the order of R4 to R6 billion based on the extent of the employer subsidy.

Table 6.1: Impact on the Government Revenue from a removal of the employer tax deduction (R’billion) (based on 1999 registered medical scheme expenditure and 2000 prices)

	Reduced consumption of private medical care
	100%	75%	50%	25%	0%
Scenario 1	3,284	3,443	3,602	3,760	3,919
Scenario 2	4,927	5,166	5,403	5,641	5,880
Per capita value of the tax subsidy
Scenario 1	549	576	603	629	656
Scenario 2	824	864	904	944	984

Based on Price et al, 1995)

A microsimulation run by the National Institute of Economic Planning (NIEP) was performed for the Committee to estimate the total medical deduction allowable under income tax. The value was estimated at R7,9 billion and included both the individual’s contribution as well as the employer contribution. The amount was broken down according to the following family types:

Single individuals: R5, 056 billion;
Couples with no children (1 or 2 taxpayers): R2, 072 billion;
Couples with children (2 taxpayers): R15,576 billion;
Single parents: R12,662 billion.

The per capita value of coverage in the public sector ranges from just over R300 (2000 prices) in provinces such as Mpumalanga and Northern Province to around R500 in Gauteng and Western Cape excluding conditional grant allocations. When conditional grants are taken into account, in 2000 public sector per capita expenditure averages just over R700.

According to the evidence the value of tax subsidies in respect of private health care expenses exceed per capita expenditure in the public sector. In certain provinces this amount is significantly less than the estimated R1,127 available as a subsidy in the private sector. In fact the total value of the subsidy is higher than the total budget spent by the Gauteng Health Department, which effectively covers in excess of 7 million people.

National Health Insurance Committee proposals

The NHI Committee (1995) identified serious problems with the existing tax regime. "The Committee recognises serious inequity and distortions resulting from present tax policies regarding medical scheme contributions. These disproportionately reduce the price of high-cost packages, encouraging inefficient use and allocation of medical resources. In addition, if mandatory cover is extended to all employees, the current tax treatment of contributions would result in decreases in employees’ after tax income, and would affect disproportionately on the self-employed."

Price et al (1995) recommended that tax concession be restructured as follows:

All contributions, whether by employer or employee should be considered part of an employee’s taxable income.
A fixed absolute amount (not percentage) of all medical scheme expenditure, including contributions to approved medical schemes, should be allowed as a deduction from table income before tax.
This fixed amount should ideally be set at a level so that the per capita subsidy (including dependants) is not greater than what the state spends on each individual in the public sector for personal care (i.e. individual medical care, excluding community level interventions). The amount should also not be set so that the net income of people earning less than a specified figure does not increase.
Consideration should be given to allowing that portion of total medical expenses that exceeds 15 percent of income to be deductible before tax. This would provide disaster relief for households hit by an unexpected catastrophe.
There should be further discussion and research regarding expenditure by employers on in-house medical services that benefit individuals but are not a necessary part of the occupational health service. Our provisional view is that, where possible, these should be considered benefits taxable in the hands of employees.
The policy could be implemented over a few years by increasing the proportion of the employer’s contribution which becomes taxable each year.
The Department of Health should attempt to negotiate a once-off increase in public health spending to absorb the tax windfall from removing the concession, in order to keep total health expenditure (public and private) constant. The new level of expenditure should be pegged as a percentage of total government spending.

Taking note of its findings and the above recommendations the NHI Committee proposed the following measures:

All contributions, whether by employer or employee should be considered part of an employee’s taxable income.
A fixed amount of all medical expenditure, including contributions to approved medical schemes, should be allowed as a deduction from taxable income before tax.
Consideration should be given to increasing the current threshold above which medical expenses are tax deductible.

Assessment of the Tax Subsidy Framework

The existing subsidy framework has to date been considered within a fairly narrow policy framework. Furthermore, the outcomes of the policy have drifted away from the achievement of any rational public policy objectives. It is fairly clear that the subsidy policy has had an impact on the way in which the private health system has evolved. It is just as clear, however, that the concession in its existing form has had little impact on the fundamental problems of the private health system and the health system as a whole. Although it may have initially played an important role in supporting social solidarity goals within the system of private medical scheme cover, these have been substantially eroded. The subsidy in its current form promotes inefficiency and inequity rather than countering these trends.

The problems can be summarised as follows:

Very little of the tax concession genuinely benefits the final consumers of health care services. Much of the intended cost reduction impact is lost to inefficiency in health care service provision and excessive administration costs.
The tax concession results in a misallocation of publicly directed health resources in favour of higher income earners and private sector service providers.
The subsidy system is an off-balance-sheet transfer to income earners and is therefore not transparent. Approximately R4 billion to R6 billion lies outside of a clear health policy framework.
The per capita value of the tax concession appears to exceed the value of per capita expenditure in the public sector.
There are no clear policy principles and objectives underlying the current subsidy framework.

Reform Options

Consideration needs to be given to bringing the tax concession policy into a consistent overall strategic health policy framework. This would imply that it ceases to be an implicit policy area within the domain of tax policy. Health care is functionally related to both population and income in a stable way. Revenue insecurity only creates instability in this framework and promotes inefficiency. A restructuring of the tax concession should therefore promote transparency and certainty in revenue flows. It should also comply with public health policy in relation to equity.

A revised strategic framework should take consideration of the following:

Consideration should be given to a reconsideration of the tax subsidy within the context of strategic health policy and not tax policy.
The tax subsidy be reconsidered in favour of an explicit on balance-sheet subsidy provided to medical scheme members. The level of the subsidy should be related in some functional and rational way to the value of cover available through the public sector.
Within a broader and longer-term reform process consideration should be given to raising the subsidy through an earmarked tax in line with proposals to introduce a universal contribution of one form or another. Initially the subsidy should be funded from the increased general government revenue resulting from the withdrawal of the tax subsidy.
The allocation mechanism, whether the funds are raised from general or earmarked taxes, would need to comply with standard equity principles. Consideration should therefore be given to allocating the funds via the proposed system of risk-equalisation for medical schemes discussed in section 7.

Prioritisation

Given that certain reforms would need to be phased in and integrated with other reforms sequencing and prioritisation is important. The following lists reforms that could be considered initially and those that would emerge in conjunction with a broader more integrated reform process:

The tax subsidy should be withdrawn for all contributions to medical schemes.
Simultaneously, the estimated increased revenues should be budgeted from general taxes, through the Department of Health budget, as a per capita subsidy to medical schemes based on the number of beneficiaries covered. The subsidy should be set per beneficiary covered and not per member.
A temporary mechanism would need to be considered for making the subsidy allocation. Ultimately the allocation would be made to a risk-equalisation fund and allocated to schemes on the basis of an equity formula (see section 7).
The subsidy should ultimately be raised as part of the revenue obtained from a universal mandatory contribution toward a national health insurance fund. Both the collection and distribution of funds would become incorporated within an integrated framework.

Risk-Equalisation

Overview

Policy Relevance

During the 1990s the policy relevance of risk-adjustment mechanisms has increased as many countries seek to make their health insurance markets more competitive and to ensure high risk individuals and groups have access to cover. Countries that have taken this route include Belgium, Columbia, the Czech Republic, Germany, Ireland, Israel, the Netherlands, Poland, Russia, Switzerland and the United States. (van der Venn et al, March 1999).

South Africa is one of the only countries in the world with a community-rated open enrolment environment that lacks a system of risk-equalisation. This lack is however related more to the recent introduction of the introduction of community rating rather than a policy oversight. It is a policy reform therefore that must be placed high on the policy agenda.

NHI Committee Recommendation

As noted earlier in this Report the NHI Committee recommended that a risk-equalisation mechanism be introduced as part of a system requiring the mandatory membership of medical schemes. It was also recommended that medical scheme contributions be income-based, thus resulting in an automatic income-based cross-subsidy, provided a risk-equalisation mechanism was in place. The risk-equalisation mechanism effectively creates a much larger risk pool out of a number of smaller independent risk pools. However, the NHI Committee proposals did not make technical recommendations on how to provide for an income-based cross-subsidy mechanism in the absence of mandating income-based contributions, if this proved not to be feasible in the short-term.

Need for Review

As the medical schemes environment will remain a central feature of the health system, there is a need to ensure that key objectives of a national health system can be realised through the private system. These are:

Ensuring that the funding of essential health services are done on a pre-paid basis;
Preventing any groups or individuals from being excluded from access to essential health services;
Ensuring that risk pools are as large as possible;
Ensuring risk-related cross subsidies for essential health services are environment-wide (from healthy to sick);
Ensuring that income-based cross-subsidies for essential health services are environment-wide; and
As far as possible removing perverse incentives to drive up costs.

Whereas tax-based health systems provide very broad risk-sharing and income-based cross-subsidisation, individual medical schemes reduce the risk-pooling effects quite dramatically. The only approach capable of achieving the protection of key cross-subsidies between schemes involves the use of a risk-equalisation fund into which contributions are paid by below average risk schemes and from which funds are paid to above average risk schemes.

In the absence of risk-equalisation, certain schemes will obtain windfall gains from a below average risk pool, creating incentives to risk-select. As risk-selection ultimately results in the systematic exclusion from cover of vulnerable risk groups, this cannot serve the final objectives of the health system.

Purpose of Risk-equalisation

Risk-equalisation is a mechanism for achieving equity and efficiency in regulated private insurance markets. Its purpose is to prevent competition from occurring on the basis of risk selection. In doing so it serves to foster competition on the basis of healthier criteria such as cost and quality of health care services.

There are a number of risk-equalisation models proposed and operating internationally. Each country has a unique system of delivery and consequently different forms of risk-equalisation are used that suit the country in question. These range from public sector formula-based resource allocation systems to risk-equalisation between competing health funds or insurers.

Within private markets mandatory community rating and open enrolment is usually required to protect cover within voluntary and mandatory contributory environments with multiple funds or insurers. However, these measures are unstable on their own and risk-equalisation is regarded as essential to protect the environment.

Risk-equalisation also become central to any government instituted income-based cross-subsidies. This is either offered as a direct subsidy or through the impact of mandating income-based contributions to health insurers. Which option is preferable would depend on the circumstances prevailing in any particular country.

Risk-equalisation should improve efficiency and reward those with lower costs. To achieve this risk-equalisation models must be based on objective risk factors or diagnostic information, not actual treatment, utilization or expenses incurred.

According to van den Ven et al (March 1999, p.3) risk-adjusted premiums are the norm, not the exception, in competitive markets, and in the absence of regulation, health plans will tend to charge premiums that differ across both observable risk factors and benefit packages designed to attract specific risk types.

"This raised the question: is this fair? … risk-adjustment premiums can easily differ by a factor of ten or more for demographic risk factors such as age, and factors of 100 or more once health status is also taken into account. Almost universally, people agree that premiums which reflect such large differences are not fair, and that cross-subsidies are needed." (van de Ven et al, March 1999, p.3)

Van de Ven et al (March 1999, p.13) raise the following problems with permitting cream skimming in voluntary open health insurance markets.

The larger the predictable profits resulting from cream skimming, the greater the disincentive to for health plans to respond to the preferences of high-risk consumers.

The larger the predictable profits arising from cream skimming, the greater the chance that cream skimming will be more profitable than improving efficiency. At least in the short-run, when a health plan has limited resources available to invest in cost-reducing activities, it may prefer to invest in cream skimming rather than in improving efficiency. Efficient health plans, who do not cream skim applicants, may lose market share to inefficient health plans who do, resulting in a welfare loss to society.

While an individual health plan can gain by cream skimming, for society as a whole, cream skimming gains nothing. Thus any resources used for cream skimming represent a welfare loss to society.

Therefore, according to van de Ven et al (March 1999, p.14) regulations that are intended to increase access to coverage for high-risk individuals may instead induce selection efforts with the following unintended effects:

Problems with financial access to coverage for high-risk individuals;
Reductions in the quality of certain kinds of care;
Reductions of allocative efficiency and efficiency in the production of care.
"Given a system of imperfectly risk-adjusted subsidies, there is a trade-off between access to coverage and the adverse effects of selection. A relevant question therefore is: How can we prevent selection?" (van de Ven et al, March 1999, p.14).

Definition of Risk-Adjustment

"Risk-adjustment" can be used to mean different things in different contexts. There is therefore a need for a definition. Van de Ven et al (1999) define risk-adjustment to mean "the use of information to calculate the expected health expenditures of individual consumers over a fixed interval of time (e.g. month, quarter, or year) and set subsidies to consumers or health plans to improve efficiency and equity."

Risk-equalisation is a zero sum game and it is important that stakeholders recognise this: there will be some winners and some losers. As such the initial implementation of a risk-adjustment model needs a carefully planned transition. It is essential that stakeholders have a clear understanding of the objectives and structure of the model.

International Review of Risk-Equalisation Mechanisms

Criteria for the Selection of an Appropriate Risk-Equalisation Mechanism

The criteria for the selection of an appropriate risk-equalisation mechanism from the Briefing Paper on Health Insurance Regulatory Framework in Ireland published by the Department of Health in July 1994 are as follows:

Equalisation of Risk Profiles: The system should provide a stable environment for community rating and open enrolment, and should eliminate the incentives for health insurers to select preferred risks, by ensuring that each health insurer must bear the cost of a risk profile equal to the risk profile of all insured lives.

Equity: The system should be perceived to be equitable between health insurers and should not result in any health insurer having to share profits which it has made as a result of its own efficiencies and cost controls.

Cost Containment: The system should not contain any inherent disincentives for health insurers to seek to maximise efficiency and control costs.

Non-equalisation of benefit levels: The system should not equalise different levels of benefit paid by different health insurance schemes

Practicality: The system should be understandable and practical to operate

Predictability: The System should produce results which are as predictable as possible, in order to allow health insurers to cost their policies appropriately.

The adjustment procedure should also be reliable (minimum error) and not vulnerable to manipulation. It should further not compromise the right for privacy of insurers and the insureds.

Criteria Used to Establish Risk-Equalisation

In developing or implementing risk-adjustment it must be decided how the information will be collected and used. Payments that are calculated at the beginning of the prediction period will use only prior information. Prospective systems estimate risk premiums for each insurer’s portfolio, based on risk factors or on prior utilisation for that insurer’s portfolio. (Ellis et al, March 1999).

Advantages for Prospective system (Society of Actuaries, June 1995):

Greater degree of certainty for health insurers
Cash flow problems removed for those insurers with poor risk profiles

Disadvantages (van Vliet et al, 1992)

Significant problems with devising a satisfactory set of risk adjusters. Global risk adjusters such as age and sex are, on their own, poor predictors of future health care costs for any one individual. Data may be difficult to obtain to use other predictors.

Van Vliet et al (1992) suggest that the following risk adjusters should be included in a per capita risk-adjustment formula:

Age and sex;
Level of insurance coverage;
Region;
Employment status;
Family size;
Socio-economic status;
Height/weight ration (BMI);
Degree of urbanisation;
Supply of health care facilities;
Chronic conditions;
Physical impairments; and
Self-rated general health status.

Other factors can, of course, be included: family history, lifestyle factors (smoker, non-smoker status, sporting activity, race). However there is a trade off between an improved prediction and complexity. The greater the complexity, the higher the administration costs, and possibly a difficulty in understanding the process. (Wilson et al, Summer 1998).

It is not necessary to predict all the variation in costs for a medical scheme. A majority of the cost-variation is random and unpredictable (hence not a basis for risk selection). Thus, the adjustment procedure must be such that the marginal benefit to the insurer of identifying individuals to risk select is less than the marginal cost of obtaining the necessary information. (Wilson et al, Summer 1998).

Payments can be calculated retrospectively, at the end of the period using information that becomes know during that period. Such a retrospective system involves the redistribution of the observed risk in terms of the actual claims costs experienced by insurers over the relevant period. Prior utilisation patterns will be a key factor in the process. Retrospective and prospective systems can however also be used in combination.

Age and gender provide a good starting point for risk-equalisation but are insufficient as much scope for risk selection remains. To improve on age and sex, prior utilization could be considered as a risk factor. This, however, tends to reward past spending and will undermine efficiency.

A further type of model is a Health status model, which is based on indicators of the insured’s health, depending primarily on medical records and past history. This, of course, might raise privacy questions. The rationale for diagnosis-based risk-adjustment models is that certain diagnoses predict future health care expenditures.

Health Status models are better predictors of costs than pure demographic models. Examples include:

The Diagnostic Cost Group (DCG) model. The DCG models use information recorded on medical claims to classify individuals into categories based on their clinical similarity.
Other well known diagnostic based models include the Ambulatory Care Group (ACG) system, and the Disability Payment System (DPS).

A further approach is to consider "self-reported measures" from surveys. The advantages of this route are the information is not dependent on medical providers; no history of claims is needed; and socio-economic variables (lifestyle, taste, employment) can be taken into account. However these surveys are often costly, response rates can be low, and there are confidentiality and accuracy concerns. The most common type of information collected in this manner is perceived health status. (NERA, 1995).

Mortality has been suggested as a risk adjuster because of the high health care expenditures prior to death. Views differ on its importance. One argument raised is that the excess costs associated with the high costs of dying are unpredictable. Another view suggests that a dummy variable indicating death during the observation period should be included in any subsidy formula. Here health plans should be retrospectively compensated with a prospectively determined payment per death. Belgium currently uses death as a retrospective adjuster based on the average number of deaths per 1000 enrolees in prior years at the health plan level. (van de Venn et al, March 1999, p.31).

Disability and functional health status are also good predictors of future health expenditures. Indicators of functional health status reflect someone’s ability to perform various activities of daily living and the degree of infirmity. Disabled and functionally impaired persons appear to have around twice the health care expenditures of those who are unimpaired. (van de Venn et al, March 1999, p.31).

Countries with Risk-Equalisation

Table 7.1 provides a list of 10 countries with risk-equalisation funds excluding the United States which has a further 10. In addition to the 10, Australia has a risk-equalisation system operated by its health regulatory authority. There is relatively significant variation in the institutional set up between each of these countries. Local conditions are therefore important in establishing the ultimate shape and form of such a mechanism.

The Netherlands have a very well developed risk-equalisation system initiated in 1991. It incorporates both risk and income cross-subsidies. Figure 7.1 provides an illustration the various inter-relationships.

Figure 7.1: Illustration of the Netherlands Risk-Equalisation Fund

Table 7.1: Risk-adjustment systems in 10 countries

	Belgium	Columbia	Czech Republic	Germany	Ireland	Israel	Netherlands	Russia	Switzerland	United Kingdom
Risk adjusters	age/gender region disability unemployment mortality	age/gender	age	age/gender disability	age/gender hospitalisation both weighted with current expenses	age	age/gender region disability	many different regional experiments	age/gender region	age/gender prior utilisation local factors
Restrictions on premium contributions	Community rating	Zero premium contribution	Community rating	Community rating	Community rating	Zero premium contribution	Community rating	Zero premium contribution	Community rating per region	Zero premium contribution
Risk-sharing	Proportional risk sharing, at least 85%	no	no	no	See risk adjusters above	Severe diseases (6% of expenses)	Outlier risk sharing and proportional risk sharing	Many different regional experiments	no	Outlier risk sharing
Number of Health Plans	6	24	26	1,200	2 (until 1997:1)	4	25	100s	166	2,500 (early 1996)
Modality A or B	A	B	B	`B	B	A	A	A	B	A
Open entry for new health plans? (subject to certain conditions)	No	Yes	yes	yes	yes	yes	yes	yes	yes	Yes
Open enrolment every month/…/year	quarter	year	year	year	year	half year	year	year	half year	no open enrolment guarantee
Is long-term care included in benefits package	yes	no	no	no	no	no	no	no	no	no
Mandatory or voluntary membership	M	V	M	M	V	M	M	M	M	V
Year of implementation	1995	1994	1993	1994	1996	1995	1991	1993	1993	1991

Source van de Ven et al, 31 March 1999

Table 7.2: The practice of risk-adjustment in the United States

	Medicare programme, HMOs in 1997	Medicare, proposed for HMOs in year 2000	Federal Employees Health Benefits’ Programme (FEHBP)	New York State	Health Insurance Plan of California (HIPC)	Minnesota Buyers Health Care Action Group	Washington Health Care Authority
Risk adjusters	Age/gender region (county) institutional status welfare status	Age/gender region (county) Welfare status Principal Inpatient Diagnostic Cost Groups (PIPDCDGs)	No risk-adjusters Each consumer’s subsidy is based on 60% of the average premium of the six largest plans	Age/gender region	Gender Number of children 120 marker diagnoses risk-adjustment only applies if plan scores deviate from one by around 5%	ACGs	Age, gender employee status since 1989 DCGs Announced for 2000
Restrictions on premium contributions	Community rating	Community rating	Community rating	Community rating	Premium contribution depends on age, region and family/single within a rate band (! 10%)	Premium contributions set by competitive bidding	Premium contributions set by competitive bids
Risk-sharing	no	no	no	Condition-specific risk sharing	no	Stop loss for catastrophic individuals	yes
Number of Health Plans	100s	100s	100s	?	28	15	3
Modality A or B	A	A	A	B	B	A	A
Open entry for new health plans? (subject to certain conditions)	yes	yes	yes	yes	yes	yes	yes
Open enrolment every month/…/year	month	Month, with proposed transition to year	year	?	year	year	Year
Is long-term care included in benefits package	no	no	no	no	no	no	no
Mandatory or voluntary membership	V	V	V	V	V	V	V
Year of implementation	1972	2000	1960	1993	1992	1997	1989

Source van de Ven et al, 31 March 1999

Review of Recommendations and Comments

Advisory Group on the Risk-Equalisation Scheme which evaluated risk-selection and the need for risk-equalisation within Ireland provides the most recent formal review internationally. Various of the comments from their Report are provided below.

"The Advisory Group concludes, based on its own deliberations and on the basis of the arguments made and evidence presented to it, that risk-equalisation is essential to underpin community rating" (Advisory Group on the Risk-equalisation Scheme, 1998, p.30)

"The Advisory Group agrees, therefore, that a Risk-equalisation Scheme is a necessary feature of the private health insurance market. It has arrived at this conclusion because of:

The very high public policy priority given to preserving the stability of community rating; and
The fact that the facilitation of competition is to be subject to the preservation of the stability of community rating." (Advisory Group on the Risk-equalisation Scheme, 1998, p.30)

"Without risk-adjustment methods, rating structures being considered in state and national reform proposals are likely to provide incentives to carriers to avoid high-risk individuals in order to maintain the most competitive premiums, and individuals will continue to face premium or contribution choices that reflect risk selection rather than medical and administrative efficiency. The Academy considers risk-adjustment a necessity if rating restrictions do not allow up-front matching of premiums or contributions with the relative risk factors of the purchasers." (American Academy of Actuaries, May 1993).

"… if a government imposes community rating on a competitive industry (health insurance or otherwise), it has an obligation to support community rating by some form of equalisation." (Walter Neuhaus, Laboratory of Actuarial Mathematics, University of Copenhagen. In Advisory Group on the Risk-equalisation Scheme, 1998, p.30).

"… an effective prevention of cream skimming is a necessary condition in order to reap the fruits of a competitive health insurance market with a regulated premium structure." (Prof. van de Ven, Department of Health Policy and management, Erasmus University, The Netherlands. In Advisory Group on the Risk-equalisation Scheme, 1998, p.30).

"As a precondition between the sickness funds it was found necessary to implement a risk structure compensation. By that financial equalisation the different kinds of funds should get equal opportunities in the coming competition. In addition, cream skimming by selecting good risks should be prevented." (Dr Doris Pfeiffer of Verband der Angestellten Krankenkassen of health care reforms in Germany.

"Recommendation: A central fund, or re-insurance scheme, should be set up to provide the insurance funds with a risk-adjustment service … Even if funds are not allowed to risk-select, but are required to accept all comers, the distribution of high- and low-cost individuals … will be uneven across funds. The function of the central fund is therefore to compensate funds with a large proportion of high-cost individuals by transferring money to them from funds with a low proportion of high-cost individuals." (National Economic Research Associates, "The Economics of Health Care Reform: A Prototype", May 1993. in Advisory Group on the Risk-equalisation Scheme, 1998, p.30).

"The Advisory Group’s initial consideration, therefore, was whether a Risk-equalisation Scheme is necessary. The overwhelming majority of respondents felt that some form of Risk-equalisation was necessary to underpin community rating. The Society of Actuaries in Ireland and the Department of Health and Children, in particular, produced an impressive range of technical support for this view.

"The Society is firmly of the view that risk-equalisation is fundamentally necessary where health insurance is community-rated. Where community rating and competition co-exist, community rating may be undermined if some or all insurers in the market practice preferred risk selection (sometimes referred to as ‘cherry-picking’ or ‘cream-skimming’). (The Society of Actuaries in Ireland" in Advisory Group on the Risk-equalisation Scheme, 1998, p.32)

"A risk-equalisation mechanism is necessary to protect insurers who are required to operate in the community rated open enrolment environment from the potentially catastrophic effects of acquiring a portfolio of disproportionately poor risks. The risk-equalisation mechanism reduces the incentive for insurers to practice preferred risk selection (this incentive is, in fact, significantly greater for community rated than for risk rated insurance, since a portfolio of good risks does not necessarily imply any reduction in premium revenues)." (VHI in Advisory Group on the Risk-equalisation Scheme, 1998, p.30).

"The Advisory Group firstly considered whether a scheme of risk-equalisation based only on age and gender might be appropriate. … It has the benefit of being totally objective, is very simple to apply and requires minimal data (in the form of number of policyholders and claims costs analysed by age and gender) to be applied. If there are differences in the claims management efficiencies of different insurers, it ensures that that no health insurer will have to share profits which it has made as a result of its own claims management efficiencies and cost controls with any other insurers. It would, therefore, provide significant encouragement to competition. However, the Advisory Group accepts that age and gender are not sufficient to account for differences in health risk, and recognises that there is a substantial level of actuarial research in existence demonstrating the limited extent to which age and gender alone can predict health care costs, when non-randomly selected populations are studied. (Advisory Group on the Risk-equalisation Scheme, 1998, p.36)

Evaluation of Residual Risk Selection in the South African Medical Schemes Environment

Figure 7.2 provides results of analysis carried out for the Committee which show substantial variation in risk pools within the open scheme environment in 1999 representing 90 percent of the total number of beneficiaries. The cost-weighted demographic profile of the individual medical schemes are compared with the cost-weighted profile of all open schemes, closed schemes, and all schemes (market). Initial indications are that the various risk profiles have not changed significantly in 2000.

The variation in risk profile implies substantial cost differences for schemes unrelated to their efficiency in managing costs. As this analysis only measures the age and sex cost variation more subtle measures such as chronic members by age could exacerbate the variations shown. Whether by chance or design, the South African medical scheme market indicates an unfair distribution of risk between schemes, which has implications for both equity and efficiency.

There is a clear advantage for open relative the closed medical scheme environment where a higher percentage of pensioners exist. Thus risk selection targeted at the closed scheme market will provide a profitable short-term strategy for commercially oriented open schemes. The advantage ends, however, once closed schemes have been eliminated.

Figure 7.2: Price advantage/disadvantage for schemes representing 90 percent of the open scheme membership

Source: analysis performed for the Committee of Inquiry based on the Statutory Returns of Registered Medical Schemes for the 1999 financial year

Assessment of a Risk-Equalisation System for South Africa

Introduction

Based on both international practice and the evidence available on the South African health market serious consideration has to be given to the implementation of a system of risk-equalisation amongst medical schemes. Without such a system inefficient schemes will be in a position to undermine efficient schemes. In order to assess the viability of such an option for South Africa some analysis was done for the purposes of this report. This cannot however be regarded as a complete assessment.

This assessment looks at five areas central to a risk-equalisation process:

Risk criteria that could be applied in a South African context;
Options for income cross-subsidisation;
Legislative requirements;
Institutional requirements; and
Expected impact on the medical schemes environment

Risk Criteria Evaluated for South Africa

Based on international evidence the following criteria have been considered for the South African situation:

Age and sex;
Members with chronic conditions;
Benefit levels;
Mortality.

As the validity of using age, sex and chronic conditions is generally well accepted internationally the value of using death as a method of refining the calculation was assessed analytically.

The issue of benefit levels is also important, as these can vary by scheme. The risk-equalisation process must therefore create a rational link between the benefits to be subject to an equalisation process and those that can be excluded.

Options for income cross-subsidisation

The South African medical schemes environment is predominantly made up of open schemes. Open schemes typically charge flat-rate contributions, i.e. they are not income-based. There is therefore no income redistribution possible via the contribution. Although it can be mandated that schemes charge income-based contributions in South Africa, this will substantially destabilise the existing market.

Income-based cross-subsidies can however be achieved through allocations from an earmarked or general tax into a risk-equalisation fund. The risk-equalisation fund therefore allocates both the income- and the risk-based cross-subsidies. An earmarked tax for this fund is more appropriate than a general tax contribution, as it establishes a clear link between a shared risk-pool and the contributory environment.

Within the South African context this option should be considered in conjunction with the removal of the current tax subsidy. (See section 6). The following steps could be considered in converting from the existing subsidisation of the private sector to an approach more consistent with health policy:

Remove the existing tax subsidy: In removing the existing tax subsidy government revenue should rise.

Reduce general taxes: A reduction in general taxes should occur, equivalent to the revenue raised from removing the tax deduction.

Implement an income-based earmarked tax: A payroll tax equivalent to the value of the desired subsidy should be raised from all income earners. The funds should be paid into the risk-equalisation fund.

Distribute the funds to scheme according to the risk-equalisation formula: Both the funds raised for-risk-equalisation and the funds raised from the earmarked tax should be distributed according to the risk-adjustment formula.

Non-medical scheme members should be subsidised for public hospital cover via a public hospital fund: Non-medical scheme members, and their immediate family, forced to contribute should become entitled to free public hospital cover in a differential amenity.

Legislative Requirements

In many countries the regulator of the private medical scheme environment operates and runs the risk-equalisation fund. A separate statutory authority is also possible. However, due to the close relationship between the regulation of solvency, community rating, open enrolment and prescribed minimum benefits to the operational requirements of such a fund.

Legislation can be created separately or part of the Medical Schemes Act No. 131 of 1998 which establishes the governance structure and operational requirements of the fund. The legislation would therefore include the following:

The governance structure;
The mechanism and calculation according to which medical schemes pay in funds;
The mechanism via which earmarked tax contributions are made to the fund;
The mechanism and formula according to which general tax contributions are made to the fund;
The prospective or retrospective nature of the assessment of relative risk between schemes;
The formula according to which funds are to be distributed to individual medical schemes;
The formula and mechanism according to which funds are distributed to any public statutory fund for non-medical scheme contributors;
The timing of receipts and payments (e.g. quarterly or annually); and
Confidentiality requirements.

Institutional Requirements

The central feature of any risk-adjustment system is a risk-adjusted premium subsidy from the risk-equalisation fund to each consumer or to high-risk consumers only. In most countries the risk-equalisation fund pays the subsidy directly to the consumer’s health plan (medical scheme) and thereby lowers the consumer’s premium contribution.

The institutional implications for a South African proposal require consideration of the following institutional issues:

A statutory organisation needs to be established which will receive and disburse funds according to established criteria.
There should be a board overseeing an executive who will directly administer the fund.
A chief executive officer will need to be appointed, answerable to the board, who will directly manage the fund and be the accounting officer.
The risk-equalisation fund will need to be able to manage the following in respect of contributions it receives:

A contributor database in respect of contributions received from medical schemes;
A beneficiary database in respect of medical scheme members;
A contributor and beneficiary database in respect of any earmarked tax (this will permit the differentiation of medical scheme from non-medical scheme contributors); and
The expertise required to formulate and manage a risk-equalisation formula.

The sources of finance will include:

Direct contributions from an earmarked tax contribution;
Contributions from individual medical schemes; and
A formula-based subsidy from general taxes. (This would be important initially during the development phase of the fund, prior to it receiving contributions directly from any earmarked contribution. This could be phased out over time, with its function reduced to that of short-term solvency support).

In its final form the risk-equalisation fund would look fairly similar to that of the Netherlands with some important differences. These differences include the possibility that: it will take some but not all contributions directly from the covered population; and that it may receive some subsidy initially from general taxes.

Figure 7.3: Possible Institutional Framework for a Central Risk-Equalisation Fund for South Africa

Implementation

The strategy adopted by the Washington State Health Care Authority (HCA) was to initially adopt a demographic model based on age, gender and member status (main member, spouse, or dependent) in 1996. This was later refined to include health-status in 1998. In 2000, the HCA moved to a more complex health status model including more health-status factors. From the HCA’s experience it is also suggested that roles and time lines are clearly defined, so that issues can be addressed early and continuously. The HCA’s time line was as follows (Wilson et al, 1998):

Decide on the goal; assign roles and responsibilities, and develop guiding principles;
Narrow efforts to a finite of risk assessment variables that are of interest;
Assess data availability and quality; develop data disclosure mechanisms;
Decide what type of risk assessment model is feasible;
Build the model;
Determine how data will be collected and processed;
Define specifics of adjustment process, including the mathematics;
Conduct a dry run;
Implement.

It is proposed that if a risk-equalisation fund is established in South Africa that a similar process be adopted. Criteria could initially be based on demographic information and be improved over time to include more information.

Priority for Implementation

The longer the absence of a risk-equalisation mechanism the longer pricing instability will exist within the medical schemes environment. There are no reasons for delaying the implementation of this important instrument. It should therefore be prioritised for immediate development and implementation.

Late Joiner Penalties and Mandatory Medical Scheme Cover

Overview

The introduction of the Medical Schemes Act No.131 of 1998 included provision for a system prescribed maximum penalties that may be applied to scheme applicants who join a scheme for the first time only late in life. These penalties are applied as a surcharge on the scheme contribution and are intended to encourage earlier and continuous membership of a medical scheme.

From January to March 2000 an amnesty on the application of these penalties was in operation. The amnesty had the intention of permitting people formerly excluded from cover from joining without penalty. Those therefore who had been excluded from cover potentially against their will were therefore given a chance to join.

As it became clear that many administrators and brokers were deliberately barring and delaying access for members during the amnesty period, the amnesty period was extended, through an amendment to the regulations, until April 2001. At the end of the amnesty period the prescribed maximum penalties were applied.

Various concerns have however arisen concerning the application of the late joiner penalties and their potential negative implications for unfairly excluding people from cover and whether the penalties are sufficiently well designed to achieve their objective.

Evaluation of the Regulations

Regulation 13 (1) provides that a "medical scheme may apply premium penalties to an applicant or dependant of a late joiner and such penalties must be applied only to the portion of the contribution related to the member or any adult dependant who qualifies for late joiner penalties."

Table 8.1: Premium penalties for late joiners

Number of years an applicant was not a member of a medical scheme after age 30	Maximum penalty
5-9 years 10-19 years 20-29 years 30+ years	1.05 x contribution 1.25 x contribution 1.5 x contribution 1.75 x contribution

A number of concerns relating to the regulations occur and are as follows:

The regulations are unclear as to whether the "maximum penalty" refers to the overall contribution or the penalty.
The definition of creditable coverage does not include instances where cover was not obtainable for valid reasons such as:

Exclusion from scheme membership;
Unemployment;
Affordability of medical scheme premiums; and
Period of residence in a foreign country.

Schemes are currently permitted to re-assess a member on transfer, i.e. they are not compelled to accept the assessment of creditable coverage from a previous scheme. This can result in unfair assessments being done just to deter older applicants.
The reference to "may" in regulation 13(1) could also result in differential treatment of preferred applicants.

International Experience

Various countries including Australia and Ireland, both of whom have similarly regulated voluntary health insurance environments, have introduced late joiner penalties. These developments are fairly recent and appear quite successful in encourage early and continuous membership within voluntary environments.

The Australian approach has been termed "unfunded lifetime community rating" (ULCR). The objective of the ULCR system are limited and is "aimed at attracting greater numbers of younger people into private health funds." (Trowbridge Consulting, November 1997, p.i).

"The structural framework essentially comprises a central age at entry and an age step prescribed for all health funds. Each fund would set its own base rate for the central age. For each individual member, the age step would be applied to the base rate for a number of years by which the certified age at entry is below or above the central age." (Trowbridge Consulting, November 1997, p.iii).

The feasibility and desirability of a system of ULCR should comply with the following:

It must have a reasonably simple design structure;
It should be administratively simple and should avoid excessive reliance on centralised control;
It should not require extensive and complex regulations.

The following was recommended as the appropriate framework for Australia (Trowbridge Consulting, November 1997):

Certified age at entry: All contributors to be assigned a certified age at entry for rating purposes.

Central age: The system to operate with a central age of 35.

Base rate: The base rate to be an individual health fund’s price for standard cover for one year for a single person joining the fund at the central age of 35.

Age step: The age step to be 2.5 percent, representing the adjustment to the base rate for each year by which the certified age at entry is below or above the central age.

Minimum and maximum certified ages at entry: The minimum and maximum certified ages to be 20 and 75 respectively (thereby attracting a maximum discount of 37.5 percent and a maximum loading of 100 percent respectively on the base rate).

Grace period: A grace period to be specified during which people who are not contributors can join a health fund and be treated as existing contributors.

Category factor: Each fund to have a category factor for couples, for couples with children and for single parents, to specify the ratio of the price for the category relative to the price for a single (e.g. for a couple with no children, a fund must nominate a factor of 0.90, meaning that such a couple would pay 10 percent less than if each partner insured as a single).

Determining the certified age at entry: For new contributors, the attained age on the first day of the calendar month of joining a private health fund. Subsequent to entry this will be modified according to periods of absence.

Selection of funds: All funds to be obliged to accept all applicants for health insurance on standard terms and conditions.

Portability: All funds to be obliged to recognise the certified age at entry of any contributor wanting to transfer from another fund.
Periods of absence: Certified age at entry to be increased retrospectively by one year to recognise each aggregate period of non-payment of contributions of 12 months.

Record keeping and monitoring: Health funds to retain audible records of date of birth, months of contribution and coverage of every member. The Commonwealth to establish a central agency for monitoring or recording the data necessary for effective administration of the ULCR system including entitlements of contributors.

Issues for South Africa

Late joiner penalties are required in instances where membership of health insurance is voluntary rather than mandatory. Within a mandatory environment all people become members if they fall within specified qualifying criteria such as income or employer size.

The introduction of a mandatory environment is superior to a system of late joiner penalties as it is administratively simpler and fair from a life-cycle point of view. However, a legislated mandate may not be feasible for one or other reason. Under such circumstances adequate measures need to be in place to encourage long-term membership.

Prior to the introduction of a mandatory environment for membership of medical schemes the problem of adverse selection needs to be appropriately managed. However, the use of these measures should not be permitted to lead to unfair discrimination against particular individuals and groups or to permanent barriers to entry for late joiners.

It is suggested that the following be taken into account in revising the system of late joiner penalties where voluntary membership predominates:

The existing regulations need to be revised to ensure that the maximum penalty refers instead to the maximum contribution.
An expanded definition of creditable coverage needs to be considered to include individuals who were unable to be members of a medical scheme for a valid reason. Consideration should also be given to accommodating those who were excluded from cover within the previous regulatory environment, i.e. prior to 1999.
Consideration should be given to an annual rather than a five-year adjustment in the penalty. This creates a much greater incentive for younger people to join early and remain within a scheme.
Consideration should be given to the introduction on an ad hoc basis of further amnesty periods. A clear indication must also be given as to what rights are conferred by any amnesty. A regular amnesty period every few years may however undermine the effectiveness of the penalty system.
Given the high cost of medical schemes, the size of the penalty could be reduced to acknowledge that not everyone will be able to afford continuous medical scheme membership throughout their lives.
It is proposed that consideration be given to a replacement of existing provisions with a framework more consistent with that in place in Australia and Ireland.
In order to reduce dependence on the system of late joiner penalties, the medical scheme environment needs to move to a system of legislated mandates. As groups get brought into this framework the applicable penalties have to be removed.

The introduction of a risk-equalisation system will also significantly reduce the systemic risk faced by an individual scheme from any late joiner. Once a risk-equalisation mechanism is put in place, the system of penalties could be re-assessed downward.

The medical scheme environment should move ultimately to legislated mandatory membership where feasible. To the extent that this can be achieved, the need for a system of late-joiner penalties will fall away.

Cost Containment in the Private Sector

Introduction

The private sector is characterised by systemic cost increases significantly in excess of general inflation and economic growth. This is primarily due to the combination of third-party payment with fee-for-service reimbursement.

Excluding administrative costs, real costs have increased by 249.7 percent from 1974 to 1999. The largest increases have been in hospital services (560.3 percent) and medicines (302.1 percent). Medical scheme expenses unrelated to actual medical services have increased by 444.8 percent in real terms from 1974 to 1999 and have increased faster than medicines.

Systemic cost increases in private markets for health are a universal phenomenon where voluntary health insurance predominates. The only long-term viable approach for containing cost increases, while simultaneously guaranteeing quality, is through global budget options and capitation.

Given these cost increases serious policy consideration has to be given to assisting in the development of a market less susceptible to systemic cost increases. Although internationally many such measures are standard, in South Africa very little effort has been put into considering a domestic framework for cost containment.

Measures typically range from supply-side interventions, such as limitations on the number of private hospital beds, services to the introduction of new technology. This section focuses on key cost drivers in the private sector and proposes measures to counter them.

Cost Drivers

Factors causing cost increases in the private sector result primarily from an imbalance of power and information between consumers and suppliers of health care. This imbalance is caused both by the need to purchase health care through risk-sharing mechanisms, and because health care goods and services are very personal and options for choice of supplier are limited.

Three broad cost drivers can be identified which result in systemic cost increases within the context of health insurance.

Demand for health insurance cover

The demand for health insurance/medical scheme cover is linked to the price (premium/contribution) charged. However, the full price is rarely faced by the direct purchaser. Many employers subsidise the premium/contribution which implies that year-on-year changes in price (premium) appear less onerous than they actually are. In addition, the tax subsidy for employers can serve to diminish the full impact of price changes. This provides greater scope for medical schemes to increase contributions higher than would be the case if members were more price sensitive.

Consumer demand for services

As a consequence of the insurance for medical services, medical scheme members face zero point-of-service costs. The result is over-consumption of medical services in cases where the consumer of health care services has significant discretion. This primarily affects primary care services, and over-the-counter purchases of medicines. Access to other services will be screened first by service providers.

Supplier induced demand

Supplier induced demand is the most important contributor to health cost increases. Service suppliers, within fee-for-service systems, have strong financial incentives to: over-supply services (increase the volume of visits, load scripts, etc.); over-charge for services (due to the weak bargaining position of patients); and to introduce new technology at high cost. The weak position of the purchaser of health services (medical scheme) to intervene in the supply of medical services effectively results in the supply of services close to what the market will bear. Within South Africa the degree of market concentration in the provider market (hospitals in particular) exacerbates this imbalance.

Interventions aimed at achieving cost containment in the private market for health care have to address these key cost drivers in order to materially impact on costs in the market. Interventions can be broken down as follows:

Supply-side controls: These relate to direct restrictions on the entry of new medical technology; price controls on medicines; restrictions on the expansion of private hospital beds, and regulating the location of medical services.

Medical schemes and intermediaries: Excessive contribution increases and administration fees can be directly regulated.

Figure 9.1: Cost drivers in the private health sector

Constraints on Developing Low-cost Healthcare Service Providers

Certain of the regulations governing health professions are preventing the development of staff model health services within the private sector. The following review is based on information provided by external stakeholders:

The Health Professions Council (HPC) Ethical Rules:

In terms of the ethical rules (see below) medical practitioners and other registered allied health professionals (such as physiotherapists, occupational therapists etc.) are prohibited to take up employment in a full time capacity and allow the employer to render an account for specific services rendered by such health care professionals. These professionals are held liable for unprofessional conduct should they allow themselves to be "abused" in such a way and are thus subject to disciplinary procedures enforced by the HPC. This specific ethical rule means that the services provided by these health care professionals can only be charged for via private practice .The result of this arrangement is the delivery of health care at the highest price.

This ethical rule does not apply to full time health professionals employed by the state, as they are specifically exempt from such rules by legislation. Professionals working in full time employment in the staff model mining system have unofficially enjoyed similar exemption, although no regulation is in place providing for such exemption. These health professionals do not directly bill. However their services are paid for by the relevant entity providing health care for the mines that in turn recover the cost from the mine owner. This still, theoretically, constitutes a breach of the ethical rule.

If health care is provided to person falling outside the employment of the mining industry at the facilities operated within the mine health care sector (such as hospitals) health professionals of all disciplines in private practice have lodged complaints with the HPC (and pharmacy council) against full-time staff. The complaints are necessarily about not about professional practice per se; but about the potential competitive threat posed. The net result is that full time employed professionals are at the risk of being disciplined by the HPC or other professional bodies. No statutory regulation exists to protect this practice.

To allow for the development of a staff model which will support national health objectives the ethical rules of the HPC have to be adjusted specifically to protect the practice of the full time employed health professional in the private sector.

The ethical rules exist as part of the Medical, Dental and Supplementary Health Service Professions Act 56 of 1974 ("The Act"). The rules which impact on the staff model system are:

Ethical Rule 4:

The use of a name of a practice of –

any name or expression, except the name of the practitioner....[or partnership or juristic person]

the expression ‘hospital’ or ‘clinic’ or any other words that could create the impression that such practice forms part of, or is in association with, a hospital, clinic or similar institution [such as AHS].

Ethical Rule 8:

Sharing fees with any person or practitioner who has not taken a commensurate part in the services for which the fees are charged.

Ethical Rule 9:

Charging or receiving fees for services not personally rendered, except for services rendered by another practitioner with whom he is associated as a partner, shareholder or locum tenens.

Ethical Rule 10:

Practising in partnership or association with a person not registered in terms of the Act

Ethical Rule 11:

Practising in or as a juristic person not exempted in terms of S54A of the Act or registered in terms 54A of the Act but not complying with such conditions of exemption.

It is therefore recommended that the above rules be fully reviewed so as to permit staff model private service providers to operate. Without an adjustment to these rules, it will prove very difficult to contain costs in private healthcare, or to develop low-cost options.

These rules are also questionable from a Constitutional perspective as they restrict the trading of health professionals without a clear rational public purpose. As such, irrespective of whether or not these provisions exist in legislation, they could in any case face a legal challenge if any attempt is made to enforce them.

Competition Commission

Containing costs within the private sector requires the creation of closed networks and a degree of vertical and horizontal co-ordination in the forming of contracts. Given the specific problems peculiar to private markets for healthcare, competition regulation needs to become more specialised for the area. If this is not possible in terms of general legislative provisions, competition may need to be specifically regulated in terms of the revised Health Act.

Supply Controls

Supply-side controls are an effective method of limiting cost increases resulting from supply-induced demand. Measures include:

Limitations on the number of new hospital beds;
Limitations on the introduction of new technology;
Reducing the over-concentration of health facilities and providers within particular areas;
Regulation of perverse incentives: kick-backs to suppliers, percentage-based mark-ups, etc.; and
Direct controls on prices of goods and services offered.

Supply-side measures can lead to unintended inefficiencies (refusal of needed services) or inappropriate price rigidities. There is also the possibility that different areas of the health system may be regulated differently and inconsistently when it comes to supply-side measures (e.g. pharmaceutical manufacturers, pharmacists, medical schemes, private hospitals). Supply-side measures may also fall foul of the Constitution if carried out in a manner that is arbitrary and unfair.

To be affective, supply-side measures, where they control the introduction of new facilities, or the location of service providers, requires both clear and uniform criteria in combination with a regional approach. In the absence of uniformity, the application of policy in this area may be regarded as arbitrary.

To achieve the required consistency a clear policy framework on supply-side cost containment must be implemented. Such a framework will require co-ordination between all areas of health policy and the health system that have a role to play or will be affected. The role of provincial and local government, as well as various regulatory authorities will need to be defined.

Regulation of Medical Schemes and Intermediaries

9.4.1 Background

The power imbalance between purchasers and providers of health services provide the underlying scope for over-servicing, over-pricing and inefficiency. Medical scheme related cost drivers include the following:

Risk-selection as the basis for price competition: Here medical schemes pay little attention to the underlying service provider costs and focus exclusively on the manipulation of risk pools to achieve market-related advantages. Year-on-year cost increases are regarded as less relevant as only the relative position is important for capturing market share.

Intermediary costs: There is clear evidence of a massive increase in the administration costs of schemes within open medical schemes. The increases now exceed those, on a percentage basis for medicines and all other services except hospital services. Although medical schemes are regulated as not for profit administrators, who operate on a for-profit basis, extract substantial sums of money taking advantage of weak corporate governance. Excessive administration fees and quota share reinsurance agreements play a significant role in driving these costs up. Although the cost of managed care may play a role, many managed care interventions have merely become an additional means to increase the administration fee of a scheme with no genuine value-added.

Broker activity: The substantial increase in broker activity within the medical schemes market has resulted in administration companies increasing the commissions, both legal and illegal, to attempt rapid increases in market share. This becomes possible when brokers effectively "control" blocks of members and can shift them based on the highest bidder. This activity has the dual impact of reducing the efficiency of scheme selection in the market and the creation of an additional layer of cost. As brokers are largely shifting members between schemes, this additional layer of cost occurs without any real value being added.

Risk-selection

Solutions to the cost-containment problems resulting from risk-selection are largely the domain of measures designed to contain discrimination on the basis of health status. Once these have largely been removed, competition between schemes can only occur through more competitive pricing of the medical services covered. This includes the measures introduced thus far: community rating, open enrolment, and prescribed minimum benefits. Additional measures include: risk-equalisation and broadening the range of prescribed minimum benefits.

Intermediaries

The cost increases occurring here are a combination of weak corporate governance and market-incentives resulting from risk-selection described above. Dealing with this problem can include the implementation direct limits on administration fees which now exceed 25 percent of the contribution, excluding reinsurance. Consideration however has to be given to a package of related interventions including:

Implementation of standardised administration agreements, which include standard performance criteria;
Improved corporate governance requirements which break up the undue influence of related parties;
Direct limits to administration fees, as well as the development of benchmarks for the industry;
Implementation of the accreditation requirements for administrators;
Routine inspections of administrators;
Comparative analyses of administration fees on an industry basis made available to the public.

Brokers

Broker activity needs to become more transparent with balancing of the powers between medical scheme members and the broker. Areas of reform which can improve the functioning of the broker market are:

Movement away from commission-based to fee-based payment where appropriate. This is particularly important where ongoing services are involved. Here brokers should not receive commission, paid directly or indirectly by the medical scheme, which is a shared cost of all members of the scheme. Where brokers offer specific services to particular members or groups of members, the fee needs to be paid and negotiated by the broker with the direct beneficiaries of the services. These members will therefore pay more for their medical aid cover than those who do not make use of a broker.

Cap commissions in the limited instances where they are appropriate: To prevent administrators and schemes from being blackmailed into paying over- or under-the-table commissions, they must be capped, with the cap uniformly applied and strongly enforced.

Transparency: Full disclosure of all vested interests of the broker, as well as all fees charged is a minimum, but not sufficient, requirement for the effective functioning of the broker market. Transparency without genuine choice, which occurs through broker collusion, will reduce the value of this intervention.

Choice: The practice in the insurance market is to force applicants to use a broker. Although direct selling is possible applicants receive no price advantage, despite the resulting lower costs. This puts brokers in a strong position to collude on pricing to the extent that there is no price advantage received from shopping around. Similar tendencies in the medical scheme market are occurring, with some medical schemes effectively channelling all applicants through brokers. Medical scheme members and applicants need to be able to shop around for the best priced broker as well as the best priced scheme. This will occur where schemes are required to accept direct applications from potential members, free of commission costs. Commissions should be negotiated directly with members and not with the scheme. In the case of ongoing services, if commission-based, this must be for the cost of the person receiving the service and not shared by the scheme. The introduction of a fee-based alternative, negotiated between the broker and the scheme applicant or member needs to be an option subject to the discretion of the applicant or member and not the broker or administrator (i.e. the case at present).

Improving the Operation of the Market

The rational use of highly specialised services or expensive diagnostic services can be encouraged through the use of more market-sensitive interventions in addition to direct controls. An important option open to government arises through making strategic use of public sector services and facilities.

Enhanced amenities (buy-up options)

The reform proposals concerning differentiated amenities within public hospitals (see section 10) will have the effect of creating an alternative provider to the private sector for general hospital services. This intervention, if implemented would guarantee the availability of reasonably priced service, and stimulate competition between private hospital.

Removal of Constraints on the Development of Staff Model Hospitals and Provider Services

The "ethical" constraints placed by the Health Professions Council rules need to be fully reviewed to permit the creation of lower cost staff model healthcare service providers in the private sector. Unless these options are permitted, only highly inefficient fee-for-service approaches, or variants thereof, will be allowable in the private market.

Intensive care and high care units

Intensive care units are under-supplied in the public sector and over-supplied in the private sector. A consequence of this is increased pressure by private hospitals to keep medical scheme members for unnecessary periods in high care or intensive care. If the public sector expanded its intensive care capability with a view to offering this capacity to medical schemes at reasonable cost, services can be improved for both public and private patients within public hospitals. Competitive pressure would also be placed on the private hospital system to more appropriately price and utilise these services to maintain market share.

Highly specialised services

Many academic hospitals are in a position to provide very specialised services not normally or easily provided in a private sector setting to medical schemes. Economies of scale can then be achieved in running of specialised units with benefits for both public and private sector patients.

Radiology

An area of significant abuse in the private sector involves over-servicing and over-pricing radiology services. Expensive equipment such as magnetic resonance imaging equipment (MRIs) and CAT scans are over-used. Consideration can be given to the strategic purchasing of this equipment for the public sector and offered to both public and private sector patients. These arrangements could lead to a reduction in current over-servicing, and a significant reduction in cost for medical schemes.

Dialysis

Dialysis and related services are heavily rationed in the public sector. Such services are also very expensively available in the private sector. With an expansion in dialysis services offered by the state and sold to medical schemes on a cost-recovery basis, with some surplus built in for cross-subsidising public patients, will result in an expansion of the overall; availability of dialysis in the country. This could be achieved without increasing the overall cost of dialysis.

Home-based care services

A significant need is developing for home-based care of people suffering from HIV/AIDS. Such services are rare in South Africa at present. Consideration needs to be given to an expansion of such services within the public sector for the use of both public and private sector patients. This will reduce the cost of treatment at public and private hospitals and improve the quality of life of many sufferers.

Palliative care

Palliative care services are not offered easily in the for-profit market. However, the need is great generally. Given that palliative care involves the lower cost treatment of terminal patients, many people could benefit from continuous and appropriate care in a lower cost setting. This will both reduce costs for public and private patients, but also improve the quality of life of patients and family members.

Step-down facilities

Inadequate attention has been given by both the public and private sectors to the advantages of good quality step-down facilities. As such many people in both environments reside longer in very high-cost setting. Within the private sector these costs can be out of all proportion to the marginal treatment cost of the patient. The introduction of both private and public sector step-down facilities, making use of significant spare capacity in the public sector, needs to be seriously considered. Expansion of public sector step-down services will find a ready market from medical schemes. The issue of step-down facilities needs to take account of new licensing requirements, accreditation of such facilities and the maintenance of minimum quality of service standards. Such services can be used for rehabilitation and palliative care.

Essential drugs

Currently drug prices vary considerably between the public and private sectors. Public sector prices are kept low through the bulk purchasing of the government. Consideration needs to be given to making certain drugs available nationally, either free, where this is merited, or on a cost-recovery basis. This can be considered in the areas of chronic medication, HIV/AIDS, and medications on the essential drug list. This intervention can serve not only to dramatically bring down the cost of treating certain conditions, but eliminate the incentive for medical schemes to discriminate against certain classes of patient (e.g. chronic sufferers, people with HIV/AIDS). Such an intervention will be both market sensitive and have significant implications for the achievement of public health objectives.

Implementation

The implementation of a coherent strategy aimed at achieving cost-containment while simultaneously achieving key public health objectives needs to become a priority area of government. Certain of the interventions need great care and a high degree of co-ordination to be successful. For this reason a strategic task team should be considered to be established which can establish a national policy framework and implementation plan.

Concluding Remarks

Cost containment within the private sector needs to be seen within the context of broader structural reforms making the health system more accessible and appropriately prioritised in terms of services. Cost increases are resulting in incentives to discriminate against poor health risks and to exclude people from insurance. An increased and unfounded burden results in the public sector. The careful and strategic use of public sector services can turn this situation into an advantage. Direct controls on the supply of services, the correcting of perverse private sector arrangements, and the intelligent expansion of appropriate services will together improve the functioning of the health system. Partial interventions, however, will easily be circumvented. Interventions must therefore be well planned, co-ordinated and implemented in combination, to achieve their full effect.

Public Hospital Reform

Purpose

The public hospital plays a central role in South Africa’s health system. Public hospitals serve the needs of the vast majority of the population. They also take responsibility for the teaching and training needs of the county’s health service personnel. Public hospitals have typically operated within a bureaucratic environment with key operational decisions made centrally by provincial or national health departments.

This environment appears to have diminished the responsiveness of public hospitals to environmental change both within the public and private health arenas. The evidence points strongly to the fact that public hospitals are not well placed to take advantage of alternative revenue sources or to benefit from any future mandatory contributory environment.

The purpose of this section is to provide a specific proposal concerning the future reform of public hospitals. The intention is to propose a design that would be consistent with both current and future policy needs. The proposal is based on a review of policy options to date, discussions with key role-players in the Department of Health, selected hospital CEOs, external consultant support, and the Treasury Department.

Overview of Policy Process and Trends Since 1994

During 1995 and 1996 a project to evaluate the reform of public hospitals was commissioned and made a number of recommendations relating to:

Decentralisation;
Revenue retention;
Differentiated amenities;
Hospital tariffs;
Governance; and
Human resources.

The central thrust of the final report was that public hospital management should be significantly decentralised. It further recommended that the policy of revenue retention would be enhanced by this decentralisation. An important recommendation involved the establishment of differential amenities within public hospitals to enhance fee payment. Revenue retention was seen as likely to be enhanced through both the introduction of revenue retention and differential amenities.

The Hospital Strategy Project found that provincial governments were not regularly updating or managing their tariff schedule adequately. They consequently recommended that a better, and possibly simpler, tariff system be adopted and regularly updated.

Attempts have been made by the national Department of Health to introduce this framework. However, implementation remains minimal and ineffectual at this stage. The following processes are noted:

Hospital Decentralisation and the development of Performance Management Agreements (PMAs): Attempts were made to delegate certain operational responsibilities to hospital managers in accordance with a contract to be entered into with the provincial Department’s of Health. The process appears to have faltered. Problems exist with the support for managing contractual arrangements within provincial Department’s of Health, and in the insufficient degrees of freedom given to hospital managers to cope with the obligations flowing from the PMAs.

The development of a Uniform Public Hospital Fee Schedule (UPFS): An all inclusive fee schedule was developed by the national Department of Health. Implementation has been slow due to the weak and non-standardisation of billing systems within public hospitals.

Retention of Revenue pilot projects: Two pilots were implemented in the 1998/99 financial year in Gauteng and Western Cape to test revenue retention options within the existing financial framework of the public sector. The pilots demonstrated clearly that the success of revenue retention was too dependent on the discretion of provincial treasuries. It also indicated that the existing regulatory framework was inadequate and there is consequently a need for a dedicated regulatory framework for public hospitals. The pilots did indicate that even minimal revenue retention changed the behaviour of hospital managers. Despite the obvious advantages to revenue retention, the failure to progress in this area suggests strongly that confusion exists as to how this should be achieved within the given financial framework of government. Insufficient attention has however been given to the creation of an entirely new framework.

Appointment of Chief Executive Officers (CEOs) as heads of hospitals: The appointment of CEOs has done little to achieve the improved management of hospitals. Problems in hospital management stem largely from the inability of hospital managers to directly manage their institutions.

Hospital rehabilitation: Attempts to compensate for backlogs in capital expenditure on public hospitals has proven to be inadequate. Funds for this purpose are allocated via conditional grants to provinces after approval of project proposals. The funds themselves are small relative to the capital needs of the hospital system and difficult to access. The result is an under-utilisation of the available funds due primarily to the inherent structural inefficiencies inherent in the over-centralised and onerous allocation process.

Review

In conjunction with the Department of Health, the Committee of Inquiry held a number of workshops on public hospital reform. Various specific problem areas with public hospitals were highlighted and reported on below.

Incentives to identify private patients and bill

There are currently no incentives for public hospitals to identify private patients. As the fee revenue goes to provincial treasuries public hospital managers find it logical to turn away private patients where possible. Private patients result in uncompensated costs for public hospitals.

Flexibility to negotiate alternative forms of reimbursement

Increasingly medical schemes have different tariff schedules. It is becoming important for them to negotiate differential tariffs to use their market power more effectively.

Opportunities for making specialist units and services available to the paying market

Specialist services can be maintained for both public and private patients. Renal units, cardiology, burns, etc. are examples of services which medical schemes would wish to contract for without any need for a differential amenity. The broader patient and funding base will serve to preserve these services for public sector patients.

Treasuries and their approach to separate operational accounts for public hospitals

Most treasuries appear unhappy about separate trading accounts for public hospital own revenue. It appears treasuries would like to redistribute these funds away from health services. This would be the only logical explanation for disallowing separate trading accounts.

Willingness to pay for public hospital services

According to results from the Willingness and Ability-to-Pay study, most people are only prepared to pay for public hospital services, whether directly or through any form of prepayment, if there is an improvement in the public hospital system.

Consequences of a lack of revenue retention at public hospitals

In the absence of a coherent revenue retention system many informal arrangements outside of the health system are starting to occur. The University of the Witwatersrand has established a PTY Ltd company (Wits Consortium) which is a wholly owned tax exempt company. Wits Consortium has purchased its own dialysis facility and is negotiating directly with medical schemes. Funds are then distributed to doctors on the Wits staff establishment. This option currently only benefits Wits staff, but does not benefit the public health system in any explicit way. Pharmaceutical trials are also being operated through Wits Consortium on public hospitals without adequate compensation for staff time used or patient expenses.

Consequences of a lack of public hospital autonomy

The lack of a formal system around revenue retention and proper hospital autonomy is resulting in backdoor mechanisms which are damaging the performance and morale of public sector staff. Public hospitals are "privatising" to the detriment of public patients. This implies that many public hospitals are effectively cross-subsidising the private sector – where publicly employed staff who serve private patients and are reimbursed by medical schemes. There are signs that this is occurring on a large scale.

Redistribution of retained revenue

The redistribution of revenue has to be balanced against the incentive to collect it in the first place as well as the need to cover any costs incurred.

Opportunities for offering public sector services to private medical schemes

The public sector is well positioned to provide services to a substantial portion of the private market, at reasonable cost if they were able to engage more flexibly with the private medical schemes. It is near to impossible at the moment for a hospital to negotiate and implement contracts with medical schemes. Too much has to happen at the provincial level – and they don’t have the time or the inclination.

Risks that public hospitals could lose their public character

Increased autonomy and interaction with the private sector could change the public sector character of public hospitals. This could be avoided, however, if a coherent policy environment were created.

Cost implications for the private sector of public hospitals selling services to private medical schemes

Being able to sell public sector hospital services to private medical schemes should result in a dramatic decline in costs for the private sector.

Public/private partnerships

There is currently no clear policy framework within which public/private partnerships (PPPs) are set up. As a consequence PPPs are largely ad hoc.

Equity and resource allocation

There is a general concern about equity in the allocation of resources, but there is no clarity as to what the redistributional objectives are. They are not explicit.

Public health system incompatibility with the medical scheme reforms

The public health system has not engaged strategically with the intentions and opportunities raised by the Medical Schemes Act reforms.

Arbitrary nature of relationships between provincial health departments and Treasuries

There are general concerns about the arbitrary nature of engagements with provincial treasuries. No clear guidelines or principles relating to budgeting or to revenue retention are evident. They largely appear to make them up as they go along.

Revenue generation undermined by treasuries

All revenue generation objectives can currently be undermined without much difficulty by provincial treasuries. This applies to any future social health insurance proposals as well as to conditional grants.

Need for hospital autonomy

The ability to manage must be decentralised to the hospital level. This needs to be a prior consideration to revenue retention. The inability to generate revenue from private sources is a symptom of over-centralisation.

Lack of political will with respect to decentralisation process

There was a feeling that the decentralisation process utilising European Union consultants could not get the job done due to a lack of political will. The problem is centralised authorities appear unwilling to relinquish direct controls.

Problems with dividing up management responsibility in public hospitals

Various PPP proposals involve the dual management of public faculties with the private sector directly managing a portion of the hospital. Such systems should not be permitted, as the hospital CEO should have full control over the entire facility. This would not preclude options which recognise this.

Problems with the centralisation of capital budgets for public hospitals

The centralisation of the capital budgets for public hospitals results in low levels of maintenance expenditure. Public Works departments have their own agendas and time scales for delivering. Their procurement processes are also dubious. If the funds were allocated directly to the hospital, and the hospital were in control of the procurement process, public hospitals would be maintained far better than at present. Capital backlogs are occurring primarily because the public hospitals themselves have little direct control over capital and maintenance expenditure.

Global budgets for public hospitals

There is a need for comprehensive global budgets for public hospitals. This coupled with greater autonomy will substantially change the performance of public hospitals. Many of the existing problems will resolve themselves.

Relationship between hospital autonomy and social health insurance

There appears to be no point in introducing a system of social health insurance in the absence of hospital autonomy and where treasuries can undermine health department budgets.

Link between the funding of public hospitals and national policy

There is currently no direct link between national health policy and the funding of public hospitals. If the current system is maintained there is little chance of this situation changing.

Centralisation of health budget coupled with decentralisation of operational control

The option of centralising health budgets coupled with the decentralisation of operational control to the lowest level possible should be considered. The centralisation/ring-fencing of the budget protects the health budget within the confines of national health policy, while the decentralisation of operational control ensures the efficient management of those funds.

State of public hospitals

The current state of public hospitals affects revenue generation. Basic amenities need to be improved generally. This needs to be coupled with an up-front investment in differential amenities.

Differential amenities versus differential services in public hospitals

Although differential amenities are supported, differential services should not be introduced into the public sector. The latter should not preclude the possibility of permitting a private hospital patient to opt for their own doctor. Here the public hospital will be reimbursed for the facility fee, while the doctor would be reimbursed separately by the patient or medical scheme. This approach is consistent with most major industrial country systems, such as Australia, United Kingdom and France.

Hospital Decentralisation

Comprehensive Approach

The need for public hospital decentralisation has been long acknowledged as an essential part of the reform process. However, elements of hospital reform are happening outside of a broader reform context. Decentralisation cannot be seen separately from revenue retention, differential amenity options and certain reforms in the financial and budgeting framework. Without this reform process, it is unlikely that consistency with the Medical Schemes Act and potential mandatory contributory environments can be achieved. Although these latter reforms will impact on coverage and the direction of the overall health system, without public hospital reform, the public hospital system will merely not participate in the evolving system.

It will also be important for this new environment to be established in national rather than provincial legislation, with areas of provincial discretion clearly specified in the Act. The Act should also provide for transitional arrangements with some flexibility built in to permit the national Department of Health to implement the process in a reasoned manner. A co-operative process with the Treasury Department would be essential.

Differential Amenities

Based on previous recommendations and the current review it appears unlikely that a mandatory or voluntary contributory system will want to make use of public hospitals unless some degree of differentiation occurs. Based on the consultation process engaged in by the Committee the following guidelines and preconditions are suggested for differential amenities:

Differential amenities should not amount to a differentiation in services;
The public hospital should at all times retain its public character;
Consideration needs to be given to an injection of funds to improve basic services and amenities up front;
The management of a differential amenity should at all times remain the direct responsibility of the hospital manger or CEO;
Differential amenities should not be created as a separate independent provider system, but should always remain part of an existing public sector hospital;
Staff should serve both the general and differential amenities in accordance with normal procedures and rotations’
Over time, with improved funding and management the underlying need for differential amenities should diminish.

Financial Framework

The financial framework consistent with a decentralised public hospital model needs to ensure that the CEO or hospital manager has full control over financial matters. Although consideration will have to be given to transitional measures, an ultimate structure is presented here which gives the appropriate degree of autonomy:

Public hospitals should operate their own bank accounts and receive budget allocations from the provincial government on an agreed draw-down schedule;
The budget allocations to public hospitals will be represented in the White Books as global transfer payments;
The public sector Financial Management System will only record the transfer payment, thereafter bank account records provided in the reporting process will be used to assess expenditure trends;
The CEO or hospital manager needs to become the accounting officer;
The hospital should provide detailed financial reports to the province in an agreed format;
Revenue taken in from private patients would be retained directly by the hospital and will be reported on but not appropriated by the legislature;
The public hospital would report to the province on all revenue, including that from private patients;

The budget allocation needs to include funds for both major and minor public works;
All procurement should be subject to an internal tender process;
Equipment purchases will occur in accordance with policy decisions made by the provincial Department of Health – but not subject to review for financial purposes;
The implementation of differential amenities or expansions of the hospital should be in accordance with an agreed policy framework developed and monitored by the provincial Department of Health;
The Hospital Board must have fiduciary responsibility and held individually accountable for the financial position of the hospital;
The hospital will need to be audited by the Auditor General’s Office;
Hospital deficits will result in overdrafts at the bank, and as such a clear policy will need to be formulated to deal with liquidity problems faced by any particular hospital;
Each hospital operating a separate bank account, irrespective of size, will require a senior Chief Financial Officer (CFO);
All procurement of drugs and consumables, normally part of a central tender process, can be ordered via any central procurement process, and payment made to the provincial treasury within an agreed period of time;
The employment of staff on a temporary or full-time basis will be at the discretion of the hospital within an agreed policy framework based on an agreed staff establishment;
The staff establishment will be generated by the hospital in consultation with the province Department of Health (this is a reversal of the current approach where the province determines the staff establishment without much consultation with the hospital).

The emphasis of the above framework is to shift the key operational workload and decisions onto the hospital. The key role of the provincial Department of Health shifts exclusively to policy development and monitoring of compliance. Decisions concerning staff, equipment and capital programmes will then be based on health service priorities and not budget control. Budget control now becomes the clear responsibility of the CEO, as the accounting officer, and the hospital board who have a fiduciary responsibility to ensure the solvency of the institution.

Governance Structure

The decentralisation of hospital accountability and responsibility suggests a far more meaningful role for the hospital board and CEO. The following provide a range of the roles and responsibilities that should ensure appropriate governance if introduced:

The Provincial Department of Health:

The provincial Department of Health will liase directly with the hospital concerning all matters affecting health policy. To ensure a direct relationship with the hospital, at least one or two department officials could sit on the hospital board of major hospitals. The Department of Health will take responsibility for negotiating the medium-term contracts and reporting process for all public hospitals. The department would need to develop policy in the following key areas:

Regional service norms;
Staff norms;
Equipment;
Framework for differential amenities and private hospital patients;
Tariffs;
Quality standards and reporting;
Borrowing guidelines and oversight (as with the new structure hospitals will be in a position to borrow funds directly for capital programmes);
Solvency requirements;
Equity requirements.

The Hospital Board:

The Board will be appointed by the MEC for Health in a particular province;
Chairperson will report to the MEC for Health;
The Board will have fiduciary responsibility for the hospital;
A range of appropriate skills and individuals of good standing should sit on the Board;
The Board will appoint the CEO;
The CEO will report to the Board;
The functions of the board will be to provide effective oversight, but not to become involved in the day-to-day running of the hospital;
Examples of the functions of the Board would include the following:

oversee and approve the policy directions adopted by the CEO within the framework of provincial policy;
oversee and approve the contract with the provincial Department of Health;
approve and monitor the human resource strategy;
approve and monitor the equipment plan;
approve and monitor the financial plan; and
review and evaluate expenditure reports.

The CEO:

The CEO is appointed by the Board;
The CEO will be the accounting officer;
The CEO will manage, appoint and determine the contracts of the senior management team including the Chief Financial Officer;
Responsibilities will include:
The day-to-day management of the hospital;
Report to the Board on a regular basis;
Develop and implement approved policies with respect to finance and service provision;
Develop plans consistent with provincial policy;
Enter into contracts with all external parties;
Develop and enter into agreements and contracts with the provincial Department of Health; and
Develop and manage the finances.

Human Resources

A key performance factor in the health system involves the quality and motivation of personnel. This requires appropriate remuneration policies, career advancement opportunities and adequate oversight and discipline. To achieve the right mix the hospital should be given as much scope as possible to determine its human resource strategy. This will include the ability to determine flexible contract arrangements for medical and nursing staff who may wish to work part-time in the private wards or in the private sector. However, the establishment of staff privileges should at all times be at the discretion of the CEO. If contract arrangements are abused or conduct inappropriate, the privileges could be withdrawn.

Relationship to Private Sector

As the CEO will have the ability to enter directly into contracts with the private sector, the development of direct contracts with medical schemes for specialised services or differential amenity wards will be greatly improved. The contract will be entered into with the CEO and the funds will be paid directly into the hospital bank account.

Special arrangements in terms of billing, tariffs and reporting can be negotiated directly. If up-front capital is required to develop a new service, in terms of such a contract, the funds could be obtained either through borrowing against future revenue or through a grant from the provincial health department. The latter will usually occur if there is a consequential improvement of services for non-paying patients.

The flexibility of these arrangements should not impact on the public character of the hospital provided a consistent policy environment has been created to ensure this. Over time, the more flexible operational and revenue environment could lead to a basic improvement in the quality of public hospitals, removing the necessity for explicitly separate amenities.

Findings and Recommendations

This report finds that the current regulation and governance structure of public hospitals is inconsistent with existing reforms such as those relating to the medical schemes, and all future options relating to a potential social or national health insurance system. It also finds the current structure harmful to effective service delivery irrespective of the revenue source. review process looking at public hospital consistency with the Medical Schemes Act and potential Social Health Insurance options came to the conclusion that:

Public hospital reform is the pivotal reform element holding back overall health systems reform.
Hospital decentralisation cannot advance without consideration of a completely revised governance structure.
Hospitals should ultimately operate their own bank accounts.
Hospital boards should be given greater accountability in future, to the extent of their having fiduciary responsibility for the hospital.
The CEO should be the accounting officer of the hospital rather than the Head of the provincial health department.
The new structure should be established in national rather than provincial legislation.
Transitional issues should be allowed for in the legislation.
Implementation should occur with the co-operation of the Treasury Department.

State-Sponsored Medical Scheme

Overview

The 1995 NHI Committee recommended that mandatory contributions "would not necessarily have to go to an existing medical scheme, but may be channelled via a new state sponsored hospital plan…". The manner in which this proposal is stated suggests an option more along the lines of the Public Hospital Fund proposed in 1997. However, the idea of a State Sponsored Medical Scheme has been proposed in various submissions and by the Central Bargaining Chamber of Government. Such schemes have been implemented in various regulated private insurance environments such as Australia and Ireland to bolster the not-for-profit community-rated open-enrolment environments. In both these countries, once the health insurance environment had fully matured, these state-sponsored schemes have been privatised. South Africa consequently needs to consider the opportunities that one or more state-sponsored scheme could offer to the consolidation of health current health policy.

Purpose of a State-Sponsored Low-Cost Scheme

A state-sponsored medical scheme would be in a position to achieve a number of basic health policy objectives. These are:

A scheme would be available which is not burdened by excessive and unnecessary administration and marketing fees.
A scheme of last resort would always be available for anyone of low-income able and willing to join a medical scheme;
A benchmark scheme will be available in the market;
The scheme would be established as the lowest cost scheme in the market, setting a minimum benchmark price against a set of minimum essential benefits;
The cost level of the scheme would provide an indication of the income group for whom mandatory membership of a medical scheme could be set;
The Scheme would provide a basis for the determination of any potential subsidy for medical scheme members;
An opportunity will be created for establishing and taking advantage of contracts with the public hospital provider system;
A state sponsored scheme could be one of the key schemes used for public sector employees when membership of a medical scheme becomes mandated in that environment.

Target Group for Cover

The group targeted for cover would be low-income groups employed in the formal sector. With the conversion of the tax subsidy into an explicit per capita subsidy, low-income groups would benefit most. As such the size of the target group will be strongly influenced by any subsidy policy introduced.

Benefits

The benefits offered would be as follows:

Hospitalisation offered in differential amenity in a public hospital;
Specialist services in a public hospital;
Primary care offered primarily in private sector capitated networks.

Contributions

The estimated contributions for a family of four will be around R500 per month. Current low-cost medical scheme options affordable to families of four with a monthly income of less than R4,000 per month range from just over R400 to around R800 per month.

Relationship to Public Hospitals

Great difficulty is experienced in developing contracts between medical schemes and public hospitals. As discussed elsewhere in this Report, much of this difficulty arises from inflexibility in the public sector system, and the lack of a specific regulatory dispensation for public hospitals. Correcting for this inflexibility should create the opportunity for contractual arrangements between a state sponsored scheme and other medical schemes.

The development of these options in conjunction with a state-sponsored scheme should have spin-off benefits for other medical schemes in two areas:

Private hospitals will be compelled to look for competitive contracts along similar lines to public hospitals.
The state sponsored scheme will offer opportunities for the development of options whereby specialised services are shared between the public and private sectors.

Civil Service Medical Scheme Cover

Background

There are roughly 400,000 civil servants covered by medical schemes out of a total of 1 million. Medical scheme cover is available to civil servants on a voluntary basis and they are not compelled to take up membership. For those that do, two-thirds of their contribution is subsidised. Those without cover fall typically into lower income categories.

The group in cover makes use of open medical schemes, where substantial intermediary costs (brokers) and excessive administration fees are paid. A risk pool with 1 million principal members is so large, however, that consideration has to be given to options where the combined purchasing power of such a group can be maximised to obtain more cost-effective cover. Substantial savings are however possible, through:

More efficient purchasing of health services;
The lowering of administration costs; and
The elimination of unnecessary intermediary (broker) commission-related expenses.

Any group-related solution to bringing all civil servants into medical scheme cover will invariably impact on potential models of service provision in the private and public sectors. Depending on its structure it could also serve to dampen upward trends in administration and other intermediary costs.

Concerns with the Status Quo

Medical scheme coverage for civil servants has emerged from a period in which Government as an employer played only a small role in actively planning health benefits for members. Consequently, coverage is incomplete, expensive and increases each year by around twice the general inflation rate. The public sector as an employer is experiencing significant cost increases that it has no influence over.

Discussion of Options

The reform of medical scheme cover for civil servants presents an opportunity to Government to generate socially acceptable and viable forms of health cover and provision generally within the medical schemes environment. The management and reform of cover for such a large group will inevitably expose inefficient and over-priced private sector business models; both in terms of health service provision, administration fees and markets for intermediary services.

Co-ordination of Civil Service Access to Health Cover and Services

Coverage of a group as large 1 million principal members and a further 2 million dependants requires strong oversight by Government as an employer, strong representation by employees, as well as strong governance of whatever form of cover is eventually chosen. A designated structure should therefore be considered that serves this purpose.

This structure would need to be representative (employer and employee) and potentially have, inter alia, the following responsibilities:

Design and implementation of medical scheme cover for employees;
Make recommendations on subsidies to employees and those who have retired;
Mandate cover for civil servants;
Oversee the accreditation of medical schemes; and
Design and implement a restricted membership scheme for public sector employees.

Mandating Cover for Civil Servants

As voluntary membership of medical schemes results in adverse selection, it is appropriate to move toward mandating cover for all civil servants. However, in introducing such a mandate, acceptable cover would have to be created for lower-income civil servants (i.e. those presently without significant cover). It would be inappropriate generally to compel civil servants to join poorly managed and over-priced medical schemes.

Restricted Medical Scheme: Proposal

Existing open medical schemes do not have the specific interests of public sector employees in mind when decisions are made concerning benefit costs, payment of intermediaries, and the type and quality of administration.

As the majority of civil servants presently not covered are low-income earners, the cost and benefit management of their cover will be essential to creating acceptable and affordable options for them. The most feasible method for ensuring that cover can be obtained at reasonable cost for this group, and even civil servants in general, is to establish a dedicated low-cost restricted membership scheme for public servants.

This scheme would be registered with the Council for Medical Schemes and focus on the needs of public servants and their dependants. In this way the interests of civil servants will be reflected in the decisions of scheme management.

The Medical Schemes Act makes provision for the governance structures of all medical schemes. Schemes must have at least 50 percent of the board of trustees elected from the scheme membership. This permits the appointment of government and employee representatives (who must be scheme members) as well as generally elected membership onto the board of trustees.

The establishment of a restricted membership scheme of this type will ensure that that the buying power of upward of 500,000 principal members can be used to purchase cost-effective administration services and high quality, low cost medical benefits.

This scheme could eventually be opened up to general enrolment and even become the "state-sponsored" scheme discussed in section 11.

Restricted Medical Scheme: Benefit Options

Low-cost medical scheme options can be defined in various reports as those offering an "essential" package of benefits (as discussed in the monograph), and costing less than R1,000 per month for a family of four with earnings of R4,000 per month or less. Note that in 2001, 53.7 percent of medical scheme members earned less than R4,000 per month.

An analysis which considered 166 options from 32 open schemes, identified 41 options as fulfilling the low-cost criteria. (Ranchod et al, 2001a and 2001b).

The most important way low-cost options improved affordability in recent years was to use capitated primary care.

The industry will probably need to break through the R500 per month barrier in product design in order to satisfy the goal of affordable healthcare.

It is in the area of hospitalisation benefits that most work needs to be done in the development of low-cost options. It is our opinion that a key element of contracting with either public or private sector hospitals will be to enter into risk-sharing arrangements, rather than traditional fee-for-service.

Our recommendation for low-cost option design is to consider the following:

Hospitalisation offered in differential amenities in a public hospital.
Specialist services in a public hospital.
Chronic medicine offered either in the public hospital or with a strict formulary by the primary care providers.
Primary care offered in private sector capitated networks.

Restricted Medical Scheme: Administration and Intermediary Costs

The selection of an administrator is by law determined by the independent management of the medical scheme and not any other party. A scheme with upward of 2 million beneficiaries will be strongly placed to negotiate reasonable administration costs.

One important benefit of a restricted membership scheme will be the removal of any need to pay commission-related fees to brokers operating within the open scheme environment. Many administrators are paying at least 6 percent of Gross Contribution Income (GCI) just to prevent brokers from removing members from their scheme. This is equivalent to the value of a full administration service.

Accredited Medical Schemes and Limitation of Choice

Many open schemes are expensive and do not provide reasonable cover for health benefits. The reduced cover is often difficult for the general public to see, something that is not necessarily accidental. It is therefore recommended that the co-ordinating structure discussed in section 12.3.1 establish accreditation criteria to qualify a set number of open medical schemes that can serve as alternatives for the restricted membership scheme.

Employees could either be restricted to the accredited schemes. Alternatively the subsidy could be limited to only those schemes which have been accredited. Such accreditation should occur on a provincial basis, to take advantage of schemes that have established lower-cost relationships with hospital and primary care network providers.

Equitable Subsidy System

A contribution subsidy is presently paid by the employer as a fixed 2/3 of the gross contribution payable, irrespective of the scheme chosen. This subsidy should be capped based upon the general desirability of the scheme chosen and the circumstances under which it is chosen. Consideration could therefore be given to subsidising member contributions taking account of the following:

The income of the member.
Whether or not the scheme is accredited (where members are not limited to a set number of schemes).
The subsidy system could be used to counter adverse selection where a number of scheme options are available to members.

Funding the Post-retirement Subsidy

The post-retirement subsidy is contingent upon the subsidy provided to current employees. The primary question for Government is whether this post-retirement liability needs to be fully funded or dealt with on a pay-as-you go basis. Pre-funding the liability does not appear a logical route to follow, given that this approach will not in any way alter the underlying risks associated with liability. Such an approach will merely attract intermediary charges, and administration fees. This approach is however different to approaches required by employers in the private sector, where the liability is reflected on their balance sheet, and where some uncertainty may arise concerning the ability of the employer to fulfil its obligations.

Regionalisation

Schemes that wish to implement capitation options, or negotiate cost saving approaches with service provider networks, are best able to do so if their membership is strongly concentrated within designated regions.

As scheme membership could be made available to civil servants via a limited accredited group of open schemes, and a specific restricted membership scheme, consideration needs to be given to a regional approach to both strategies.

Where schemes are accredited, these could be specific to each province. Thus the three accredited schemes in the Western Cape may be different to the three schemes selected in Free State. This would allow Government to utilise those open schemes that are able to incorporate regional approaches in their product design. This could also form part of the accreditation criteria.

The restricted membership for civil servants could also be regionalised with options created that are province specific. Thus the scheme remains a national scheme, but civil servants join the option appropriate for their province. It will therefore become a condition of employment that members must join the option designed for their province.

Regional representation in the formulation of medical scheme strategies will be very important to ensure legitimacy and to make recommendations on alternative strategic approaches. This should probably not occur within the scheme(s) but occur instead through the central co-ordinating structures set up to oversee the civil service policy as a whole.

Figure 12.1: Regional Structures for Civil Service Medical Scheme Strategy

Relationship to a Risk-Equalisation Fund

Removing the artificial cost advantage one scheme may have over another due to their demographic profile is best managed through a risk-equalisation fund/mechanism. The medical scheme arrangements created for the civil service should become subject to the risk-equalisation approach discussed in section 11.

Options in Relation to an Open State Sponsored Scheme Option

The development of a low-cost restricted membership scheme for civil servants will create a sound basis for the general development of low-cost medical scheme cover. It will achieve this through allowing the development and testing of low-cost primary care and hospital options that would prove difficult to achieve within the existing open scheme market.

The civil service restricted membership scheme could eventually be made available to general enrolment. This scheme could either become the state-sponsored medical scheme discussed in section 11, or operate alongside it. The extension of this scheme into the open market will provide the opportunity for members of traditional schemes to take advantage of pre-negotiated regional network arrangements and benefit options.

Strategic Direction

Overall Framework

The strategic approach would see the establishment of a government co-ordinating structure, with provincial representation from both the employer and employees. This structure would determine the subsidy system for all civil servants. It would also set up and see to the registration of the restricted membership scheme and develop its benefit options. The co-ordinating structure would also develop, negotiate and implement the accreditation mechanism for a limited number of medical schemes per province. (See figure 12.2).

Figure 12.2: Framework for Universal Contributory Cover for Civil Servants

Potential Timelines

The implementation of such a reform would require fairly detailed planning. However, it should be possible to achieve this in a period of roughly three years. (See figure 12.3).

First phase: This would see the implementation of the government co-ordinating structure, and the establishment of a limited number of accredited medical schemes per province. It would also see the negotiation and design of the restricted membership scheme.

Second phase: This would see the implementation of the restricted membership scheme and the introduction of mandatory universal cover for all civil servants. The restricted membership scheme would have 9 options, one for each province. The overall framework could be expanded to all public and semi-public institutions.

Third phase: The restricted membership scheme could be initially expanded to incorporate all public and semi-public institutions. This could ultimately be lead to the opening up of the scheme to general open enrolment. The scheme could also become the proposed state-sponsored scheme discussed in section 11.

Figure 12.3: Timelines for Implementation of a Strategy for Universal Medical Scheme Benefits for Civil Servants

Concluding Remarks

The development of universal medical scheme cover for civil servants is probably one of the most important immediate health system opportunities within the short-term. It will provide the impetus for the creation of low-cost models of health service provision in the private market, and help develop opportunities for public hospitals to make surplus capacity available to medical schemes.

Within a broader reform context, the proposed restricted membership scheme for civil servants could ultimately be made available for general enrolment and take on the role of the state-sponsored medical scheme proposed in section 11.

Medical Savings Accounts within Medical Schemes

Overview

Medical savings accounts became a feature of the medical scheme environment during the 1990s. Although the regulatory environment did not make provision for them, many commercially oriented schemes began introducing them as a means of risk-rating groups and providing incentives for the out-of-pocket purchases of a range of health services where there is greater consumer discretion.

With the introduction of the Medical Schemes Act No.131 of 1998, medical savings accounts were permitted, but subject to a limitation on the value of the contribution that could go to one.

Most health goods and services are not subject to a high degree of consumer discretion as they are provided on the recommendation and supply of a doctor or other medical professional. For this reason the funding via a risk pool is justified, as services are demanded on the basis of medical necessity rather than personal preference.

Some medical scheme administrators have motivated for the use of medical savings accounts on the basis that they encourage cost savings.

Medical savings accounts do not occur internationally on any significant scale and are generally not supported by government policy. Shifting the risk onto consumers has also not been shown to result in any fundamental shifts in health care services. On the whole they are regarded as measures to reduce the life-cycle and risk-related cross-subsidies typically occurring within large risk pool arrangements.

Within the United States attempts by commercial lobby groups to get federal support for medical savings accounts was partially blocked by the Democrats in 1996. Democrats opposed medical savings accounts fearing they would only appeal to the healthy and wealthy, leaving those with less money and more health problems behind in an increasingly costly risk pool. The net result of the debate was the limitation of medical savings accounts to a demonstration programme. (Families USA, August 1996).

There is no evidence internationally that shifting costs onto consumers encourages more rational purchasing of needed medical services. Consumers will however respond to medical goods and services that are preferred and not needed. However, by their very nature such goods and services are not insurable, and should be purchased on an out-of-pocket basis.

Discussion

The policy-related justification for medical savings accounts is very limited, and all evidence suggests that it is counter-productive. Information from the Council for Medical Schemes also indicates that medical savings accounts are not being used in accordance with the provisions of the Act. (This is based on information supplied by the Finance section of the Council for Medical Schemes).

A further concern arises from the potential substitution of a large portion of the overall medical scheme contribution from risk-contributions into non-risk contributions (savings accounts), resulting in reduced risk-pooling within medical schemes. Hidden benefit reductions have most probably occurred for essential medical services through the substitution of a portion of contribution from necessary toward unnecessary health care services.

No objective evidence could be found that self-insurance reduces cost trends of necessary medical services. Cost reductions only occur through strategic and selective purchasing of health goods and services on a collective basis, i.e. public health systems and health insurers, where alternatives to fee-for-service are used. Collective purchasing logically only occurs for services within a risk pool. Individual purchasing of needed health services fragments purchasing power as well as access to services.

Administration companies charge unusually high administration fees for managing medical savings accounts, sometimes upward of 10 percent of contribution. The reconciliation of individual entitlements and interest accruing and charged is essentially unregulated at present. Given that medical savings accounts are essentially the personal savings of an individual, many individuals are likely to be financially worse off than if they placed the funds in their own personal bank account.

Industry Commission (Australia)

The Industry Commission in Australia evaluated medical savings accounts within their private health insurance industry. They listed the following concerns with such an approach (Industry Commission, 1996).

The most problematic deficiency was that unlike other adaptations to community rating systems, "there is no pooling of risks between individuals. Medical savings accounts are based on the premise that much ‘insurance’ is really inter-temporal smoothing. But how much is ‘much’? The Commission sought, but was unable to obtain, information on the amount of savings needed to meet most people’s lifetime health costs."

"Some individuals’ lifetime health costs will be low and others high. The latter’s savings may be insufficient to meet their health costs. Moreover, a person might get chronically ill when young, before sufficient savings had been amassed."

"MSA’s do not take into account the fact that persons needing more than average care would be grossly under funded … Savings schemes are no substitute for the creation of risk pools which allow individuals to share their risk exposure (although they may assist in creating sufficient savings to buy insurance in retirement)."

"Another problem is posed by those people with low lifetime health costs. They would leave an excess in their savings account at death. What is the appropriate policy for such undepleted savings?"

"Medical savings accounts raise issues about transitional arrangements, portability, prudential requirements and, to a much lesser extent, sovereign risks, similar to those posed by lifetime rating."

"Low income people are able to save less, and so could meet fewer exigencies than others. This is also true for people who opt in and out of the workforce (say because of child rearing)."

"Medical savings accounts would appear to require a completely different set of skills than those possessed by current insurers."

Recommendations

Medical savings accounts are clearly problematic in a number of important areas of policy and consumer protection. It is therefore recommended that the current policy be revisited with a view to phasing them out of medical schemes, or at the least substantially diminishing their impact on risk pools and contribution costs. The focus of health policy needs to be on risk-sharing and cost containment. None of these key health policy objectives can be achieved through medical savings accounts.

Key Strategic Challenges

Introduction

The strategic challenges facing the South African health system provide the context for reform. They highlight deficiencies from a holistic perspective rather than as problems affecting the public or the private sector. The relationship between the public and private sectors is not neutral with strong feedback effects operating between the two.

Context for Reform

When contrasted with the key reform objectives of a health system, many cannot adequately be met by the current policy framework in South Africa. Important areas raised are summarised below based on views expressed by the World Health Organisation (WHO, 2000).

Many countries are still making inadequate efforts in terms of responsiveness and fairness of health contribution with respect to the provision of health services. The impact is most severe on the poor who are driven deeper into poverty by the lack of financial protection against ill health.

The ultimate responsibility for the overall performance of a country’s health system lies with government, which in turn should involve all sectors of society in its stewardship.

The careful and responsible management of the well-being of the population is the very essence of good government. For every country it means establishing the best and fairest health system possible with available resources.

Publicly financed healthcare systems remain the backbone of health care in most countries.

The route of prioritising only primary health care as the route to achieving universal coverage is now under severe criticism. The alternative approach calls for an understanding of "basic" health care which requires the delivery of essential health care, defined mostly by criteria of effectiveness, cost and social acceptance.

Defining "basic" essential health care implies an explicit choice of priorities among interventions, respecting the ethical principle that it may be necessary to ration services, but that it is inadmissible to exclude whole groups of the population.

Health policy and strategies need to cover the private provision of services and private financing, as well as state funding and activities.

Oversight and regulation of private sector providers and insurers must be placed high on national policy agendas. At the same time it is crucial to adopt incentives that are sensitive to performance.

Incentives within unregulated private insurance markets for health care are so skewed that the normal rules of competition do not work.

Monopoly power on the part of service providers results in higher prices, lower output, and lower product quality.

When physicians are reimbursed on a fee-for-service basis they are given powerful incentives to provide more services than are necessary.

According to international experience no single payment system is optimal. For this reason most countries adopt mixed systems.

Pre-payment is regarded as the best form of revenue collection for health services, while out-of-pocket payments tend to be regressive and impede access to care.

The main challenge in revenue collection is to expand prepayment, in which public financing or mandatory insurance will play a central role.

In the case of revenue pooling, creating as wide a pool as possible is critical to spreading financial risk for health care, and thus reducing individual risk and the possibility of impoverishment from health expenses.

Achieving greater fairness in financing is only achievable through risk pooling – that is – those who are healthy subsidise those who are sick, and those who are rich subsidise those who are poor.

Insurance schemes designed to expand membership among the poor is an attractive way to channel external assistance to health, alongside governmental revenue.

Governments need to promote community rating (i.e. each member of a community pays the same premium), a common benefit package and portability of benefits among insurance schemes, and public funds could pay for the inclusion of poor people in such schemes.

In middle-income countries the policy route to fair prepaid systems is through strengthening the often substantial mandatory, income-based and risk-based insurance schemes, with increased public funding to include the poor.

Strategic purchasing needs to replace much of the traditional machinery linking budget holders to service providers. Selective contracting and the use of several payment mechanisms are needed to set incentives for better responsiveness and improved health outcomes.

Within insurance environments it also matters how revenues are combined so as to share risks: how large they are; whether competition exists between pools; and whether, in the case of competing pools, there are mechanisms to compensate for differences in risk and capacity to pay.

Evaluation of Current Policy Context

Public sector

Linkages between policy development and implementation:

The decentralisation of the health budget within the context of fiscal federalism implies the decentralisation of health policy. Although institutional provision is created for the development of national policy, there is very little direct connection with provincial health systems. As a consequence most national policy implemented at a provincial level only relates to relatively minor issues that can be informally agreed to by all provinces at a national level.

Decentralisation of operational responsibility and accountability:

The public health system combines a decentralised policy development process with highly centralised levels of operational responsibility. There is clear evidence that this division of responsibility between policy and operational responsibility is perverse and dysfunctional.

Raising revenue from voluntary and mandatory contributions:

The approach to financing public health services, and hospitals in particular, makes it virtually impossible for effective cost-recovery to occur when services are provided to higher income groups. In order to prevent the under-funding of public health services, when they serve people not provided for in the general budget, cost-recovery must occur. The current system of public finance makes no provision for efficient cost-recovery options within the public service.

The federalisation of health funding and policy creates a potentially fatal disjuncture preventing the serious consideration of mandatory contributory options (i.e. social or national insurance) designed around the use of public sector services. By their nature such systems would require centralised institutions which raise the funding and reimburse health services. Without a unified system allocating the health budget to provinces, the required centralisation of the contributory system would clash irreconcilably with the fiscal federal system.

As provinces would receive the money conditionally from the contributory system, but have discretion over the funding from general taxes, differences will occur between basic services and enhanced services provided through the contributory environment. Without resolving this issue a contributory system dedicated to the public sector will prove discriminatory and serve very few public policy objectives.

User fees:

The application of means tested user fees for hospital services to uninsured patients using public hospitals is both discriminatory and operationally flawed. Apart from the fundamental inability of public hospitals to apply the exclusion principle, which is a prerequisite for any system of user fees; and to do means test assessments at point-of-service, which no systems reform could make work. The sheer volume of patients seen makes individual billing of uninsured patients in all settings (public or private) administratively impossible.

The introduction of a revised policy on hospital tariffs (the uniform patient fee system or UPFS) although an improvement upon previous tariff systems is a palliative measure and will achieve little in the way of cost-recovery for the public hospital system and public policy in general. The tariffs do serve some purpose in charging medical schemes, or social insurance funds such as the Road Accident Fund (RAF). However, public hospitals themselves will see very little of the increased revenue, and structural flaws relating to billing out-of-pocket patients, the application of the exclusion principle, and the application of the means test cannot be overcome merely through adjusting the tariffs.

Budget Allocations:

The allocation to the health service is declining in real terms on a per capita basis. This results in staff reductions and capacity problems. The reduced quality of service available in the public sector creates a privatisation by default, with only the private health system as an alternative. The absence of any real choice of sector for higher income groups results in the monopoly pricing of both medical services and medical scheme contributions. Public sector budget cuts appear to be one of the most significant contributors to increases in overall health spending.

Equity:

The achievement of equity on an inter-provincial basis is virtually impossible due to the existence of the fiscal federal system. Furthermore, the strict division between the public and private sector disallow any coherent subsidy framework that can span both systems in a coherent fashion.

Human Resources:

The rigidity of the centralised system of human resource regulation has resulted in a significant deterioration in morale and capacity within all elements of the public sector. This has had a more severe impact on the health system which is already complex and multi-disciplinary. Staff retention in critical areas of the health service is now difficult both as a consequence of inadequate budget, remuneration and career opportunities. Options that allow staff to work in both environments simultaneously are currently very difficult to operationalise and control.

Private Sector

Cost increases:

The private sector is characterised by chronic cost increases linked to the fee-for-service reimbursement of providers, an oligopolistic service provider market (which prevents cost containment resulting from competition between service suppliers). Recent trends also show that people are in a weak bargaining position relative to open medical schemes. As a consequence consumers face an inelastic demand for medical scheme cover, which is abused. This takes the form of over-charging administration fees, the extraction of underwriting surpluses from schemes using quota share reinsurance agreements, and the paying of excessive commissions to brokers in competition for market share.

Links to the public sector:

It is likely that a market for lower cost public sector services would develop, given cost pressures driven by over charging in the private sector. However, the inability to contract due to public sector inflexibility is a key constraint despite a willing market for public hospitals.

Low-cost contributory environment:

The development of a low-cost market for medical scheme cover is hindered by the following:

An oligopolistic provider market;
The inability of medical schemes to formulate contracts for improved amenities at public hospitals, or for other relevant public health services, due to public sector inflexibility; and
The existing tax subsidy which only serves to reduce the cost of cover for higher income groups.

Risk-Selection:

There is evidence that a significant degree of residual risk-selection continues to exist in the medical schemes market. In the absence of any system of risk-equalisation, this will result in instability between medical schemes.

Tax Subsidy:

The value of the tax subsidy toward the private health system is substantial and is estimated at R7,8 billion. It currently lacks a clear public policy objectives with associated identifiable positive outcomes. The subsidy therefore needs to be reconsidered within a broader subsidy reform framework.

Demographic Structure of Medical Schemes:

The demographic structure of medical schemes imply a differently structured health system to that of the general population. This creates concerns about the resulting efficiency of the health system as a whole given the substantial resource allocation bias in favour of the medical scheme market.

Intermediaries:

Intermediaries do not always act in the best interests of scheme members and the public at large. This includes instances where administrators abuse their influence over schemes under their management; where brokers blackmail administrators into paying kickbacks to retain members; and where managed care arrangements are merely structured to extract additional fees from schemes. The shift of members between schemes is largely induced by broker activity, rather than active decisions of members. Thus schemes are incurring substantial increased costs, for no added value to the environment. Overall non-medical expense related expenditure, which includes administrative expenditure and broker fees, is the fastest growing cost-driver in the private health market.

Unfair Discrimination:

There is evidence of significant discrimination against people with chronic conditions in open medical schemes. Currently the prescribed minimum benefits do not protect members from this form of abuse. As most people who suffer from chronic conditions are in older age cohorts this amounts to unfair discrimination on the basis of age.

Mandatory Contributory System

The introduction of a mandatory contributory environment in addition to the non-contributory tax funded public health system has been the ultimate objective of health policy since 1995. Such a contributory system can take the form of dedicated social health insurance (SHI) fund for contributors only. It could also take the form of national health insurance (NHI) where both contributors and non-contributors benefit from a universal system.

National Health Insurance versus Social Health Insurance:

From an organisational point of view the implicit and explicit subsidies required within the overall health system remain identical irrespective of whether the regulated contributory and non-contributory systems remain separate.

National health insurance is not an option that emerges overnight as an alternative to social health insurance. Instead it becomes feasible within market economies where formal employment levels are high. Prior to this mixed systems are inevitable.

Future Paths for South Africa:

Regulated private insurance coupled with various social health insurance options and government subsidies represent the middle-income country route toward building a universal system.

National health insurance, or the complete nationalisation of the private sector, cannot be seriously considered as reasonable options for South Africa.

National health systems and insurance can be based upon single or multiple payer systems. The choice of system largely depends on historical developments and local conditions. Whichever system prevails makes little difference to the underlying equity principles and objectives.

Concluding Remarks

Although many of the elements of a unified and integrated health system exist in South Africa, at present they do not result in a functional and integrated framework. If these deficiencies are not addressed, the health system as a whole will continue to increase in cost, while simultaneously reducing and becoming increasingly unfair in the allocation of cover.

Role and scope of government involvement

The ultimate responsibility for the overall performance of a country’s health system lies with government, which in turn should involve all sectors of society. A government has the responsibility for establishing the best and fairest health system possible with available resources. Health policy and strategies need to cover the private provision of services and private financing, as well as state funding and activities. The oversight and regulation of private sectors has to form part of the overall government response and must be high on the policy agenda.

Central Objectives

Increased risk pooling: Risk pooling needs to be encouraged through the use of a combination of instruments. These would include the tax system, subsidies to private regulated insurers, the creation of risk-equalisation mechanisms within both public and private sectors, government mandates, and the reinforcement of community rating.

Finance: Government policy needs to ensure that a universal minimum financial allocation is made available for all people resident in South Africa. It should however be possible to top-up this minimum allocation with medical scheme contributions.

Benefits: Government policy needs to provide a framework that results in cover for a minimum level of essential services, irrespective of whether it is provided in the public or the private sectors.

Service provision: Ensuring that a sustainable universally available service provider system is in place must underpin government’s strategy with respect to healthcare. Central to this strategy must be the strengthening of the public sector owned and controlled network of services.

Efficiency: Given the existence of perverse incentives in unregulated markets for health care, any regulation must pay careful attention to the incentives generated. The use of mixed systems for covering and providing health care combined with the correct elements of choice is the best approach to balancing health care objectives with the need for operational efficiency.

Role of the Public Sector

The public sector system must remain the backbone of the overall health system and should be protected from chronic under-funding.

Role of the Private Sector

The private sector can provide an effective environment for achieving increased levels of funding over and above tax-based allocations. However, as the private market for health care suffers from chronic market imperfections, public sector involvement is required to ensure that funding levels are socially optimal and not merely what the market will bear.

Integrated Strategy for Health Systems Reform

Overview

The reform strategy outlined in this section integrates the information from previous reform processes and the analyses of earlier sections in this report and encapsulates it within a rational reform path. To a large extent reforms recommended in earlier processes remain intact. This section provides a prioritisation of those reforms pivotal to the achievement of fundamental health systems reform.

Elements of the reform process which need to be prioritised in the short- to medium-term are identified and divided into four phases. The achievement of the final phase is seen as a fairly long-term objective which will not be fully realised within the next ten years. It is important however to reflect this final phase here to provide clarity on the ultimate direction of the health system.

Government policy is clearly complicated by the peculiarities of the health market which necessitates intervention if health goals are to be achieved. Further complicating Government policy is the fact that no clear formula exists to achieving its goals. The level of economic development and the maturity of existing institutions affect the available options and opportunities. There is therefore an ongoing obligation on Government to carefully assess policies on their merits.

Although the level of economic development and existing institutions influence reform options, they need not undermine the achievement of policy goals that address the underlying obligations of Government.

Within this section various institutional arrangements are suggested. These are given suggested names consistent with their function. Ultimately, if accepted, these functions could be arranged quite differently from an institutional point of view. This would however not diminish the need for the functions themselves.

Principles

Applying some of the key principles identified thus far the following obligations on the state can be identified with respect to social security (which includes health care) and health care in particular:

Basic necessities: All people living in South Africa are entitled to the basic necessities of life.

Comprehensive and co-ordinated social security programme: The State must adopt a comprehensive and co-ordinated social security programme in which responsibilities and tasks are clearly allocated to the different spheres of government, and appropriate financial and human resources are available for its implementation;

Responsibility: The national government and not merely a single department or authority has the overall responsibility of co-ordinating a social security programme and for co-ordinating and managing the health system.

Equity: Public sector resources must be allocated on an equitable basis. Within the context of health care, an equitable distribution of health services is also required.

Finance: Overall revenue allocated to the health system must involve a fair share of the overall resources of the country.

Implementation: Policies and programmes must be reasonable both in their conception and their implementation for them to be regarded as compliant with the qualification of progressive realisation provided for in the Constitution.

Prioritisation for those in desperate need: Government must identify those in desperate need within the context of health service provision and ensure that it has a valid programme in place to cater for this need.

Emergency Medical Care: The right of access to emergency treatment is not subject to the qualification of progressive realisation. However, the Constitution does not make provision for this right to be free of charge. The State is therefore obligated to provide an environment in which the exercise of this right can occur in a manner that does not undermine the sustainability of the health system, and minimises the financial distress of all concerned.

Children: Children have special rights provided for in the Constitution. The importance of children within the life-cycle, within families, and broadly in their relationship to effective social development is a clear underlying value of society. Families with children in distress also have a first call on state resources. The State is therefore obligated to ensure that access to basic health care services is unrestricted for children and related services where children may be affected.

Goals

Access to a basic set of services must be guaranteed to all and not be based on the ability of any individual or group to pay.

Although access to a basic set of services will be guaranteed for all, those able to contribute, whether in the form of taxes or mandatory contributions, should be required to contribute. As far as possible contributions need to be made over in a pre-paid form and, where essential basic services are concerned, with user fees and co-payments eliminated.

The central objectives of a final system are clearly those underpinning all of health policy. The objectives underlying the desired structure and institutional framework may be more specific and should include:

Financing:

Achieve the integration of the existing voluntary contributory system with mandatory contributions and tax-based finance;
Broaden the risk-pooling and thereby lower the costs of accessing the health system for all residents;
Eliminate all co-payments (user-fees) at point-of-service for all public facilities, and private facilities where these involve basic essential services;
Ensure that all those able to contribute on a pre-paid basis do so.

Service provision:

Ensure the existence of a strong public sector owned and controlled service as a provider of last resort for all residents;
Ensure an equitable distribution of public health services for the entire non-contributing population;
Provide the flexibility for public sector services to be available to medical schemes on a full cost-recovery basis;
Provide flexibility for medical, nursing and auxiliary staff to be able to contract for work in both the public and private sectors;
Provide a regulatory framework that ensures that private sector providers cannot abuse their potential monopoly power to increase fees, costs and utilisation in excess of socially desirable levels.

Institutional framework:

Responsibility for health policy should reside with the national Department of Health;
The administration key public sector services and functions, such as hospital services and district management, should be fully decentralised with improved financial accountability and governance;
Where national priorities must be met by a lower tier of government, as far as appropriate, allocations should be conditionally allocated by the national Department of Health.

Reform Strategy

It is the recommendation of this Report that in the medium- to long-term South Africa move toward a National Health Insurance system compatible with multiple funds and a public sector contributory environment as defined in the 1995 NHI Committee Report. Initially the environment should continue to be strictly differentiated between a private contributory environment and a general tax funded public sector environment. Over time this strict differentiation can diminish with a broader contributory environment emerging, replacing general taxes as a revenue source. The ultimate elimination of general taxes as a key revenue source is unlikely for a fairly long time, and may in fact not even be desirable as policy objective.

Four phases are envisaged defining important linked reform measures. The phases guide the evolution of health system toward the achievement of a universal contributory system.

Phase 1: Development of the enabling environment for greater integration:

The current health system is incompatible with the introduction of or integration with contributory environments. The overall system of cross-subsidies is fragmented and not structured in accordance with strategic policy goals. Furthermore, the strict partitioning between the public and private sector spheres negatively affects the operational effectiveness of both environments.

The priorities within phase 1 therefore need to focus on putting in place an enabling environment for more substantive and far reaching policy reforms.

Phase 2: Implement preparatory reforms:

In conjunction with the establishment of the enabling environment, a number of reforms of strategic importance should be implemented. These need to focus on the creation of regulated risk pools, and major enhancements to the regulation and subsidisation of the medical schemes environment. The objective is to improve the quality and cost-effectiveness of cover within the voluntary contributory environment (medical schemes).

The phase 2 reforms serve to enhance the voluntary contributory environment in order to facilitate the establishment of a mandatory environment emphasised in phases 3 and 4. The greater the degree of cover, and the acceptability of the contributory environment, the less the disruption involved in establishing any future mandatory environment.

Phase 3: Implementation of the initial mandates:

Once the preparatory reforms of phase 2 are in place, the groundwork would have been established for the implementation of the first statutory mandates. Given the income distribution in South Africa, the mandates should begin with higher income groups. Where lower income groups are concerned, this phase should focus on further development of the voluntary contributory environment.

Phase 2 would have seen the initiation of a state-sponsored medical scheme. Phase 3 should focus on the development of a contributory scheme for non-medical scheme members. This will help to establish the institutions in government that would ultimately manage a public sector contributory scheme within a National Health Insurance framework. Thus two contributory mechanisms will exist: the first based on medical schemes; and the second a dedicated public sector contributory fund.

Phase 4: Implementation of National Health Insurance:

The last phase envisages the implementation of a universal contributory system which would to a substantial degree replace general tax funding a source of revenue. General tax as a supplementary source of revenue may nevertheless prove desirable. The final phase essentially envisages the establishment of a contributory environment for all groups and individuals assessed to be in a position to contribute toward the health system. These contributions would not replace medical scheme contributions, but rather fund a proposed subsidy provided to medical schemes. All contributions and general tax allocations should ultimately be made directly to a Central Equity Fund (CEF) which would in turn allocate them to the public sector and medical schemes based on a risk-adjusted equity formula.

A Public Sector Contributory Fund (PSCF) should be considered to become the national funding authority for the public health system. This would enable the consolidated allocation of all funds directly allocated by the national Department of Health. Phases 1 through 3 would have seen the centralisation of key components of the health budget (including the establishment of a contributory system dedicated to funding public sector services), and the establishment of capacity to fund provinces via improvements in the capacity to manage the conditional grant system. The end phase of these enhancements should see the creation of the PSCF to take responsibility for and manage the allocation of funds from general tax revenues and contributions allocated through the CEF.
Figure 15.1 Reform Strategy and Approximate Timeline

Phase 1: Enabling Environment

Objectives

Public sector reform pre-requisites for the introduction of a contributory component to funding public sector services require the:

Centralisation of the health budget;
Creation of a dedicated unit to manage the system of conditional grants;
Implementation of differential amenities; and
Decentralisation of hospital management.

Private sector funding reforms required as pre-requisites for the introduction of a mandatory environment for medical scheme membership are the:

Revision of the subsidy system for medical schemes;
Implementation of a risk-equalisation fund;
Availability of contracting options for public sector services; and
Mandating of medical scheme membership for all public sector employees as a pilot.

Complementary measures to the above would include:

Developing an ongoing policy with respect to defining and implementing a basic essential package of services compatible with both the public and private sector approaches to service reimbursement and provision.

Preparation of the Public Sector Budget System

Centralisation (ring-fencing) of the key components of the Health budget:

Centralising the allocation of the health budget creates a coherent link between national policy objectives and provincial service delivery. It also establishes a required direct relationship between the budget allocation process and any future contributory system dedicated to using public health services. This could be phased in over a number of years as the capacity of the national Department of Health is improved to revise the current allocation system.

If no central ring-fencing of the health budget occurs the viability of any future social health insurance option dedicated to funding public health services will experience difficulties for the following reasons:

The creation of a social health insurance institution will require a parallel administration system resulting in inefficient duplication when the conventional budget system could be adapted instead.
Ensuring that funding levels for public hospital services will be compatible with the entitlements to those services for both contributors and non-contributors.
Ensuring compatibility between the funding of hospital and district level services will prove difficult.

However, it should be noted that the use of matching conditional grants in conjunction with minimum norms and standards could achieve many of the objectives without the full ring-fencing of the public health budget. This less onerous approach could therefore be considered in the short- to medium-term if full centralisation is perceived as too risky. Irrespective of which option is adopted, the establishment of a dedicated unit with the required expertise to allocated current and future conditional grants should nevertheless be established.

Establish an expert unit to manage the conditional grant system:

Ensuring the achievement of an equitable distribution of physical health services requires careful management of the conditional grant system. Both allocation criteria and conditions need to be established centrally. This role should not extend to the micro-management of provincial administrations, but should instead attempt to achieve designated broad service targets within a coherent national framework with mutual consent. This unit should ultimately be incorporated into the proposed PSCF.

Preparation of the Public Sector Hospital System

Decentralisation of public hospital management:

The current inflexible regulation of public hospitals affects both their management of revenue sourced from general taxes and their ability to access the private sector contributory (medical scheme) environment. Hospital decentralisation needs to extend to their having strong boards, a CEO as the accounting officer, and the holding of their own bank accounts. Hospitals need to be placed in a position to enter into and manage contracts directly, subject to the oversight of provincial health authorities and the relevant hospital board.

Implement a coherent uniform policy with respect to enhanced amenities:

In order to reinforce the contributory system, enhanced amenities need to be introduced into the public system. It is proposed that this focus on, but not be limited to, hospital services.

As access to particular services and amenities will depend on whether or not someone is part of the contributory environment, a smart card identification system is important, if not essential. If such a card is not produced at a public facility, however, no-one will be denied cover. They will only be denied access to the enhanced amenity.

Public facilities should at all times retain their public sector identity. No separate facilities offering only enhanced amenities should be created as this will impact on access and equity. Over time the system of enhanced amenities may fall away once the need for an inducement to pay ceases to be important.

Initially enhanced amenities will apply and be developed with respect to medical schemes. However, the development of a public sector contributory system for low-income groups would also develop around the enhanced amenity.

Financial injection to enhance the public sector services:

In order to implement a system partially reliant on voluntary contributions, basic improvements to the facilities and services are required. Otherwise contributions will not be forthcoming. These improvements must not be confined to the enhanced amenities.

Establish a process to develop and implement minimum service requirements for the public system:

To ensure the consistency of the conditional grants, a national policy framework for public services, integrating all levels of care, is required. As the conditional grant system is effectively the implementation arm of public health policy, this process should focus on the establishment of service requirements and norms and standards.

Human resources:

Staffing arrangements also need to come under tighter management, but allow for greater flexibility to serve both the public and private sectors. These contracts should be limited to scarce personnel only, and probably to specified institutions.

Consolidation of Medical Scheme Reforms

Remove residual risk-selection:

The removal of residual risk-selection requires a number of smaller reforms in conjunction with a risk-equalisation process. The risk-equalisation mechanism is discussed further below. The other reforms required include:

Prescribed minimum benefits: The expansion of prescribed minimum benefits to include chronic conditions, expanded HIV/AIDS cover and other essential services;

Benefit Options: The phasing out of separate options in schemes, or alternatively, limiting their number and the basis upon which benefits can be differentiated;

Medical Savings Accounts: The phasing out of medical savings accounts; and

Late Joiner Penalties: The introduction of an improved system of unfunded lifetime community rating.

Increase coverage generally:

This can occur through a number of measures:

Mandating membership in restricted membership schemes; and
Requiring all civil servants to become members of a medical scheme.

Effective regulatory environment for intermediaries:

A system which ensures appropriate economic rewards for intermediaries needs to be considered. A substantially improved approach to consumer protection is also required.

Development of a Policy Process on Basic Essential Services

The public and private sectors define their benefit entitlements differently. The public sector establishes services which provide comprehensive cover for an undefined range of conditions. The conditions covered may be qualified through the use of protocols.

The regulatory environment for private sector has moved toward the creation of positive lists of services for which cover is provided. The Medical Schemes Act No. 131 of 1998 now specifies a positive list of conditions and treatments which must be covered by schemes.

Government has to move toward defining what it regards as basic essential services which everyone must be covered for. Although these may be defined differently between the public and private sectors, there must be convergence on the approaches adopted in the two environments.

Ultimately both the public and private sectors need to provide a minimum core set of services. Within medical schemes these would be regulated as prescribed minimum benefits. Within the public sector a similar process would occur and be framed as minimum norms and standards.

Development of Integrated Subsidy System

Revise the subsidy system:

The tax system is currently the most important vehicle for achieving most of the risk-pooling required to generate income-related cross-subsidies. Consideration has to be given to using the system of general taxes to achieve income-related cross-subsidies in the medical schemes environment.

The Medical Schemes Act No.131 of 1998 provided for the introduction of a limited number of risk-related cross-subsidies within the medical schemes environment. However, there are various structural deficiencies which need revision and include:

Inequity in the allocation of public health services;
Tax subsidy to medical schemes;
The lack of risk-equalisation between schemes; and
Unfair penalties.

Revision of the Budget System:

The budget system needs to be revised to ensure that the regional allocation of health services is equitable. Furthermore, the value of subsidies given to the private sector should not exceed that provided for people covered through the public sector.

Tax subsidy:

The tax subsidy currently contradicts health system objectives. A revision of the overall system along the lines begun through the Medical Schemes Act must be considered, which achieves greater risk pooling and converts the tax subsidy into an explicit risk- and income- adjusted subsidy. The achievement of these cross-subsidies will require the creation of a risk-equalisation fund.

The existing tax subsidy should be phased out and an explicit on-balance sheet subsidy phased in. The subsidy would initially be funded from general taxes but could gradually replaced by an earmarked tax. The subsidy would take the form of a risk-adjusted per capita allocation in respect of everyone in a medical scheme, or who participates in any public sector voluntary or mandatory contributory environment.

Introduce a system of risk-equalisation:

A system of risk-equalisation is required both to balance the uneven risk pools (i.e. schemes with above or below average proportions of older or sicker people) between medical schemes and to distribute the explicit subsidy that replaces the tax subsidy.

Implementation of measures to Contain Private Sector Cost Increases

Various measures are possible to reduce service costs within the private sector using either direct limitations on the supply of technology, or through market oriented measures such as central purchasing by the public sector for both public and private sectors.

Supply controls:

Currently under-utilised or not consistently utilised are a number of measures that are essential to containing the ballooning of costs on the supply-side within the private sector. Internationally there is careful management of the introduction of new technology, and the geographical distribution of services. Over-concentration leads to supply-induced demand which private sector funders find difficult to control. Government has to implement a coherent framework which aims to directly limit excessive concentrations of providers and new technology.

Medical Scheme Administration Fees:

Open medical schemes are experiencing significant increases in non-medical expense related costs. Consideration needs to be given to capping administration costs at reasonable levels. Annual increases should also be limited to ordinary inflation.

Market-related measures:

The rational use of highly specialised services or expensive diagnostic services can be encouraged through making these services available to the private sector at cost through the public sector. In certain instances, services could be provided universally free of charge where appropriate. Tuberculosis and HIV/AIDS services would be examples.

Areas that can be seriously considered for sale to medical schemes on a cost-recovery basis to the mutual advantage of both environments are:

Enhanced amenities;
Intensive Care Units;
High Care;
Highly specialised services;
Radiology;
Dialysis;
Home-based care services;
Palliative care;
Step-down facilities; and
Essential drugs.

Removal of bottlenecks constraining the development of managed care:

The development of improved selective contracting within the private sector requires that schemes be permitted to determine the terms and conditions of service suppliers in contracts entered into on a willing buyer and seller basis.

Medical schemes must be placed in a position where they can include volume and price in the same contract. Only in this way can appropriate risk be shared between the scheme and the service provider.

The shifting of risk onto insured members or individuals has proven to be an inferior and discredited strategy. It is only when adequate risk-sharing between funders and providers exist that genuine efficiencies can be realised.

To prevent provider collusion from preventing the development of appropriate selective contracting a legislative framework that deals with competition needs to be introduced into the Medical Schemes Act and enforced by the Council for Medical Schemes.

Establishing a consistent competition dispensation for the health system needs to be developed in conjunction with the Competition Commission.

Phase 2: Implement Preparatory Reforms

Objectives

This phase needs to expand cover within the voluntary contributory system (medical schemes) and consolidate reforms to improve the system of cross-subsidies.

Implement Risk-equalisation Fund for Medical Schemes

The full implementation of a risk-equalisation system, begun in phase 1, is required. The risk-equalisation fund will be a statutory authority reporting to the Minister of Health. Its key functions will be two-fold:

Ensure that all medical schemes face the average demographic and health risk structure of the market as a whole; and
Distribute a risk-adjusted subsidy to medical schemes.

The risk-adjusted subsidy should replace the existing inefficient and inequitable tax subsidy provided to employers and individuals. Ultimately, the risk-equalisation fund will be expanded into the CEF implemented in phase 4.

Implement Risk-adjusted Subsidy to Medical Schemes

The tax subsidy, currently valued at R7,8 billion will need to be replaced by a more efficient and explicit subsidy to medical scheme members. This subsidy should initially be funded from general tax revenue. In phase 4 of the reform process the revenue source could move toward a universal mandatory contributory system.

The aggregate value of the subsidy will be determined according to the value of public sector services not utilised by individuals receiving cover in the private sector. The difference in the value of the original tax subsidy and the revised subsidy will be used to improve the public sector budget. The distribution mechanism for the subsidy should be the risk-equalisation fund.

Implement State-Sponsored Medical Schemes

The implementation of a state-sponsored medical scheme targeted at low-income groups and the informal sector would have the following objectives:

A scheme of last resort would always be available for anyone of low-income able and willing to join a medical scheme;
A benchmark scheme will be available in the market which can generate competitive pressure on inefficiently run open schemes;
The scheme would be established as the lowest cost scheme in the market, setting a minimum benchmark price against a set of minimum essential benefits;
The cost level of the scheme would provide an indication of the income group for whom mandatory membership of a medical scheme could be set;
An opportunity will be created for establishing and taking advantage of contracts with the public hospital provider system; and
This scheme could be one of the key schemes used for public sector employees when membership of a medical scheme becomes mandated in that environment.

Not only will such a scheme create downward pressure on costs within the private market, it will assist in the development of a low-income contributory environment.

Implement Mandatory Environment for Civil Servants

Currently just under 50 percent of civil servants are on a medical scheme. The first phase of any move toward a mandatory contributory environment should therefore begin with moves to bring civil servants under some form of cover. This can be achieved through an employer mandate requiring membership of one of a limited number of accredited medical schemes. Although this measure can take the form of an employer and employee bargaining process (as opposed to a statutory one) aspects of it can be reflected in legislation in accordance with a broader statutory framework.

Phase 3: Implement Statutory Mandates

Objectives

There are two major objectives to this phase:

Establish a mandatory environment for medical scheme membership for high-income groups; and
Initiate a voluntary public sector contributory environment outside of the medical schemes movement.

The first measure represents the final phase in the reform of the medical schemes environment, while the second establishes the basis for a subsidised low-income contributory environment.

Mandate Medical Scheme Membership

The mandatory medical scheme membership for high-income groups will reduce adverse selection and stabilise medical scheme membership. This move really only becomes socially acceptable once major efforts have been made to bring private sector cost increases under control and to ensure adequate risk-equalisation between schemes. This measure represents the final phase of the medical scheme reform process initiated through the introduction of the Medical Schemes Act No.131 of 1998.

The mandate should initially focus on groups that are largely in membership already. As such the mandate should involve limited disruption to employer costs and employee benefits. Qualifying groups for the mandate need to be based on a combination of employer size and minimum income levels.

By this phase a fairly sustainable and good quality low-income scheme environment could have evolved based on reforms introduced in phases 1 and 2. This would have been achieved through the following:

Supply-side measures: to limit supply-induced demand;

Replacement of the tax subsidy with an income-based risk-adjusted per capita subsidy: lower the cost of cover for low-income groups;

Public sector contracts with medical schemes: introduce effective competition with private health service suppliers; and

State-sponsored medical scheme: introduce competition with existing medical schemes.

The lower-cost environment should provide employers with greater choices for more affordable cover for good quality services. The mandatory environment also diminishes the need for late-joiner penalties.

Implement Voluntary Contributory Environment for Low-Income Groups

The existing system of user fees for higher income groups using public sector hospitals needs to be replaced by a form of pre-payment. For individuals within a medical scheme this issue will already have been resolved. A pre-payment system exclusively for public hospital utilisation for low-income groups outside of the mandatory framework is required for those who may not wish to, or be unable to afford, medical scheme cover.

Pre-payment would qualify individuals for access to enhanced amenities within public sector hospitals. For non-contributors access will only be permitted to basic amenities. In terms of this system, everyone will be entitled to a free state service whether or not they make a contribution. The pre-payment system merely provides a vehicle for "buying up" within the state system and thereby incentives payment.

To minimise the risk for the state system, and the sustainability of the medical schemes environment, the mandatory medical scheme membership for high-income groups needs to be implemented together with the creation of the low-income contributory system.

Phase 4: Final Implementation of National Health Insurance

Objectives

The final stages of the reform process will involve the implementation of a universal contributory environment. Despite the universality, a degree of flexibility will continue to exist between funding environments and provider choice. Choice of funding environment will become more limited the lower the income of an individual or family. However, the flexibility of the third-party purchaser (public and private) to contract with different providers will be high. This will increase price competition between providers and serve to preserve the sustainability and affordability of health care services generally.

Overall Framework

The final phase of the reform process would seek to combine the public and private sectors under a single universal contributory system. This system would directly fund the basic public health sector service, a differentiated public service for non-medical scheme members, and subsidise members of medical schemes who would be able to top-up their subsidies with their own contributions to their medical scheme.

The proposed environment would effectively integrate the regulated multi-fund (medical schemes) and a public sector managed and controlled system. The universal contribution would be distributed via a proposed Central Equity Fund (CEF) which would evolve from the risk-equalisation fund established in phase 2.

To cater for adverse selection issues within the medical schemes environment, higher income groups will be required to join a medical scheme. They would however be free to choose their scheme. This would already have been implemented in phase three.

Lower income groups, including the informal sector, would have the option to choose between a medical scheme or access the enhanced public sector amenities via the PSCF. Contributions for the low-income formal sector workers will be mandatory where they exceed a statutorily determined level.

Both the CEF and PSCF should operate within a statutory framework established by the Minister of Health.

In this way the public and private systems become integrated within a unifying framework that permits public policy to coherently influence equity and access to health services irrespective of where they are situated (i.e. public or private sector) or their revenue source.

The continued existence of the medical schemes environment will be essential, as these vehicles provide an established and functioning contributory environment which, if regulated to achieve minimum equity and access guarantees, can ensure that adequate resources are available within the overall health system.

Medical scheme revenue would be derived from two sources:

A risk-adjusted subsidy based on an equitable allocation from the CEF; and
Top-up contributions (also risk-adjusted via the CEF) derived from member contributions.

To prevent or contain perverse trends in the distribution of physical resources resulting from the medical schemes environment, supply-side management of any escalation or maldistribution of private sector resources needs to become an established regulatory function of national and provincial government.

Figure 15.2: Framework for a Universal Contributory System

Central Equity Fund (CEF)

The CEF should initially start out as the institution created to manage risk-equalisation between schemes as well as allocate a risk-adjusted subsidy to medical schemes. The initial subsidy would be an explicit allocation from general taxes (as described in phase 2) replacing the existing tax deduction. This system would:

Ensure equity in the raising and allocation of health resources;
Improve the mobilisation of revenue through combining contributory (which contains a strong willingness-to-pay element) and tax-based sources of revenue; and
Ensure a consistent link is maintained between the economy’s capacity to fund an acceptable health service and the funding made available.

The CEF would have the following functions:

The collection of income-based contributions from the public;
Alternatively to (a) the same effect could be achieved via a formula-based subsidy funded from general taxes;
The management of a contributor database and membership information;
The distribution of funds to:

The public sector basic amenity service for non-contributors;
The public sector enhanced amenity services for contributors; and
The private sector medical schemes.

The fund distributions will be based on an equity formula which would incorporate both income and risk-based cross-subsidies.

The health system could move entirely away from a general tax-funded system toward a universal contributory system. Within such a framework some allocations from general taxes to deal with any revenue instability would initially remain. If this occurs consideration could be given to the allocation of a fixed percentage of revenue from VAT and corporate taxes supplementing the contributory system.

Contributions would be required on a mandatory basis from specified income classes within qualifying employer groups. Certain employer groups may be excluded, i.e. those below a specified number of employees. For groups falling outside of the mandatory net, voluntary membership would nevertheless be possible. Non-contributors would be entitled to utilise the basic amenity service of the state free of charge. They will however not have access to enhanced amenities. Although some free-riding will occur it should not be financially significant.

The CEF would not attempt to manage provider reimbursement itself but merely fund institutions in the public and private sector that are specialised to perform this complex function. Within the public sector the responsible institution will be the national Department of Health, via the PSCF. In the private sector, medical schemes would directly fund services.

With respect to the private sector, the CEF would perform a risk-equalisation and income-cross-subsidisation function. The schemes would continue to directly manage the reimbursement of service providers. Funds involved in the risk-equalisation process would come directly from medical schemes.

Public Sector Contributory Fund (PSCF)

A Public Sector Contributory Fund (PSCF) should be established to manage the reimbursement of provincial health departments. This authority would not deviate from the equity allocation distributions established by the CEF, but focus on and be empowered to ensure the equitable regional distribution of physical health services.

The PSCF would flow from reforms initiated in phase 1 where specialised capacity in the form of a national Department of Health unit is created to manage and distribute the health allocations as conditional grants to provinces.

The PSCF would determine the allocations to provinces as conditional grants. It would also establish the associated conditions in conjunction with the national Department of Health. It would however not determine the actual provincial budgets, as these will be set by the provincial governments.

The determination of allocations to provinces would be a technical and not a policy exercise. Policy decisions would be the exclusive responsibility of the Department of Health.

The PSCF will need to take into account factors such as the physical distribution of health care service providers and personnel in the determination of allocations and conditions. Progressively achieving an equitable distribution of health resources would need to occur through negotiations with the provinces and the development of a consistent funding framework.

The PSCF would also need to determine the allocation of funds in respect of enhanced amenities via the conditional grant mechanism. The proposed system should form part of the existing budget system and cycles.

Medical schemes should be able to directly contract with the public health system where desired. They should also be in a position to pay funds directly to the PSCF in respect of pre-paid or utilised services in the public sector. Where such payments are made, the PSCF should adjust the provincial allocations and conditions accordingly.

Subsidy to Medical Scheme Members

All residents should be entitled to a subsidy equivalent to the risk-adjusted per capita average of all contributions received into the CEF. This subsidy should be available irrespective of whether the individuals are covered through the medical schemes environment or through the PSCF.

This subsidy system will evolve from the recommended changes to the tax subsidy system which forms part of the short- to medium-term reform process.

This approach will permit individuals to opt out of the public sector provider system without impacting on the degree of income- and risk-related (i.e. healthy to sick) cross-subsidisation underpinning the allocations of the CEF.

Financial Implications

The functioning of health systems are extremely sensitive to the financial framework within which they operate. Unstable revenue sources that have no relationship to changes in service demand are as problematic as revenue sources very sensitive to demand changes. To be effective health systems must operate with sustainable revenue sources, a reasonable degree of sensitivity to demand changes, and a variety of provider reimbursement mechanisms.

The South Africa public sector budget system does not provide sufficient flexibility to ensure the financial sustainability of the public health system. The current framework appears to have generated structural under-funding of the health services and their associated capital requirements. The public health system is also unable to respond to demand changes resulting from higher income groups (with and without medical scheme cover) using public hospitals due to rigidities inherent in the budgeting process.

The envisaged framework outlined in phases 1 through 4 requires that certain principles be established in implementing the reform process. The following are recommendations on a number of principles:

All revenue taken from user fees or funds received from medical schemes within public sector facilities should be retained at source and be used to cover the cost of services sold;
Allocations to public health services from the general taxes should as far as possible be determined nationally and allocated to provinces on an equitable basis;
Redistributional objectives within the public sector must be achieved with budget allocations emanating from general taxes and not from user fees and medical scheme payments;
The management of hospitals must be decentralised to permit the utlisation of more appropriate financial management approaches with respect to revenue from multiple sources;
The overall financial framework should be designed to maximise the effects of risk- pooling.

Coverage

Coverage changes over the four general phases with the gradual expansion of the contributory system. The public sector basic amenity is the non-contributory environment offered free to all below a certain income level. Higher income groups move from a voluntary contributory environment into mandatory options for both medical scheme membership and a final National Health Insurance (NHI) contribution.

By phase 3 the user fee system for public hospitals is eliminated and replaced by a combination of mandatory medical scheme membership and a voluntary contributory system for an enhanced differential amenity. Middle- and upper-income groups will largely be compelled to join a medical scheme during this phase. Public sector schemes will be able to contract for the differential (enhanced) amenity. Phase 4 creates a mandatory contributory environment which includes low-income groups. From that stage on, low-income contributors will access enhanced amenity services.

Figure 15.3: Implications for Coverage over all Phases

Table 15.1 Summary of coverage by broad income category

	Phase 1	Phase 2	Phase 3	Phase 4
Poor	Public sector: basic amenity (free)	Public sector: basic amenity (free)	Public sector: basic amenity (free)	Public sector: basic amenity (free)
Low-income	Public sector: basic amenity (user fee)	Public sector: basic amenity (user fee) Medical Scheme (voluntary)	Public sector: basic amenity (free) Public sector contributory fund (voluntary) Medical Scheme (voluntary)	Public sector: basic amenity (free) Public sector contributory fund via NHI contribution (mandatory) Medical Scheme (voluntary)
Middle-income	Public sector: basic amenity (user fee) Medical Scheme (voluntary)	Public sector: basic amenity (user fee) Medical Scheme (voluntary)	Medical Scheme (mandatory)	NHI contribution (mandatory) Medical Scheme (mandatory)
High-income	Public sector: basic amenity (user fee) Medical Scheme (voluntary)	Public sector: basic amenity (user fee) Medical Scheme (voluntary)	Medical Scheme (mandatory)	NHI contribution (mandatory) Medical Scheme (mandatory)

Concluding Remarks

The various phases outlined in this framework reflect the need for careful planning and prioritisation of interventions. The reform process is complex and multi-dimensional. Significant technical work and consultation will be required in virtually every phase and step of the process. This complexity should be recognised as inherent to health systems reform and a degree of openness and flexibility permitted to fully develop the reforms for implementation.

Concluding Remarks

A pivotal recommendation of this Report is that the reform direction and approach developed and proposed in the 1995 NHI Report remains valid and should continue to be the basis for further reforms. In the long-term this requires that South Africa move toward a National Health Insurance system over time making use of multiple funds in the form of regulated medical schemes, coupled to and compatible with a universal contributory system.

It is a finding of the report that the Medical Schemes reforms initiated in 1998 (Medical Schemes Act No.131 of 1998), based on the 1995 NHI Report, are an essential component of a stable health system. These reforms prevented the large-scale removal of high-risk groups from cover and have been instrumental in starting the drive toward lower cost medical service models in the private sector.

The reform process has to take into account the need to look at a phased approach whereby key enabling measures are implemented and the base established for the longer-term reforms. This Report has grouped the reforms into four phases:

Phase 1: Development of the enabling environment:

Reform of the public hospital system:

Decentralise public hospital management;
Centralise key aspects of the public health budget;
Implement a coherent uniform policy with respect to enhanced amenities;
Investigate the possibility of a financial injection to enhance public sector amenities;
Establish a process to develop and implement minimum service requirements for the public system;
Revise the human resource environment as it relates to health personnel to improve management and incentives to perform.

Consolidation of Medical Scheme reforms to remove any residual risk-selection and to increase coverage:

Expand prescribed minimum benefits to include chronic conditions and other essential services;
Phase out benefit options or, alternatively limit the degree to which they can be differentiated;
Phase out medical savings accounts from medical schemes;
Refine the late-joiner penalties;
Require all civil servants to become members of a medical scheme; and
Significantly improve the regulatory environment for intermediaries.

Development of an effective policy process on defining and implementing Basic Essential Services: Ultimately both the public and private sectors will need to ensure coverage for an equivalent minimum core set of services. Within medical schemes these would be regulated as prescribed minimum benefits. Within the public sector a similar process would occur and be framed as minimum norms and standards.
Development of an Integrated Subsidy System:

This process needs to focus on rectifying structural deficiencies within and between the existing risk-pooling mechanisms (i.e. medical schemes and any part of the system funded from general taxes). These should include dealing with:

Inequity in the allocation of public health services;
The tax subsidy to medical schemes;
Risk-equalisation between medical schemes; and
Unfair penalties applied within the medical schemes environment.

The public sector budget system needs to be revised to ensure that the regional allocation of health services is equitable. Furthermore, the subsidy provided to the private sector should at no time exceed that provided to people covered through the public sector.
It is essential that a system of risk-equalisation between medical schemes be introduced. This fund would also serve the function of allocating any appropriately structured risk-adjusted subsidy to medical schemes provided by Government.
The tax subsidy currently runs counter to the achievement of health policy objectives and must be reformed. It is recommended that it be converted into an explicit income- and risk-adjusted subsidy. This subsidy could ultimately be funded from an earmarked tax, although initially it should be funded from general tax revenue.

Measures to contain private sector cost increases need to be more explicitly targeted by Government policy. These should include the use of:

Direct controls on the supply of services;
Various market-related measures; and
Improved regulation of competition.

Phase 2: Implement preparatory reforms which include:

A risk-equalisation fund.
A risk-adjusted subsidy to medical schemes;
A state-sponsored medical scheme; and
A mandatory environment for civil servants.

Phase 3: Implement initial mandates and develop voluntary low-cost contributory options for low-income groups:

Mandate medical scheme membership for higher income groups; and
A voluntary contributory environment for low-income groups outside of the medical schemes environment.

Phase 4: Implementation of National Health Insurance:

Implement a universal contributory system which would be offset from general taxes.
Establish a Central Equity Fund which would have the following functions:

The collection of income-based contributions from the public;
Alternatively, a formula-based allocation funded from general taxes could be considered;
The management of a contributor database and membership information;
The distribution of funds to:

The public sector basic amenity service for non-contributors;
The public sector enhanced amenity for contributors;
The private sector medical schemes;
The fund distributions will be based on an equity formula which would incorporate both income- and risk-based cross-subsidies.

Establish a Public Sector Contributory Fund to manage the reimbursement of provincial health departments. This authority would not deviate from the equity allocation distributions established by the Central Equity Fund. This fund would however take into account the regional distribution of public health services and attempt to achieve equity.
All residents of South Africa should become entitled to a subsidy equivalent to the risk-adjusted per capita average of all contributions and revenue received into the CEF. This subsidy system should evolve from the reforms in phases 1 through 3.

Bibliography

Advisory Group on the Risk-equalisation Scheme, Report of the Advisory Group on the Risk-equalisation Scheme, The Minister for Health and Children’s independent review of the Risk-equalisation Scheme, 8 April 1998.

Adams SI, A Study of the changes in the Affordability of Medical Scheme Contributions, following changes to the Medical Schemes Act (Act 131 of 1998), Mini Thesis, Actuarial Science Department, University of Cape Town, 11 April 2001.

AFReC, Public Hearings on the MTEF Figures for 2001/2002, Submission to the Select Committee on Finance, 23 August 2001.

American Academy of Actuaries, "Health Risk Assessment and Health Risk-adjustment, Crucial Elements in Effective Health Care Reform", May 1993.

Arrow KJ, "Uncertainty and the welfare economics of medical care", American Economic Review, December 1963, pp.941 to 973.

Berman P, "Supply-side approaches to optimizing private health sector growth," Private health sector growth in Asia: Issues and implications. Edited by W. Newbrander. 1997. John Wiley & Sons, Ltd, pp.111-133, p.124.

Bradshaw D, Masiteng K, Nannan N, "Health Status and Determinants", Chapter 4, South African Health Review 2000, The Press Gang, Durban, December 2000, pp. 89-124.

Carstens, I Grobbelaar, N. Smit, A. van der Merwe, A. van der Walt, G. van der Walt, "Health Care", in Social Security Law, general principles, Eds. MP Olivier, MC Okpaluba, N Smit, M Thompson, Butterworths, Durban, 1999, pp.213-247.

Cichon M, Newbrander W, Yamabana H, Weber A, Normand C, Dror D, Preker A, Modelling in health care finance, A compendium of quantitative techniques for health care financing, ILO, Geneva, July 1999.

Department of Health, Draft Policy Document, Policy proposals concerning the financing of the private health industry in South Africa, 28 April 1997.

Department of Health, A Social Health Insurance Scheme for South Africa: Policy Document, September 1997.

Department of Health, Finance Committee Task Team Report, December 1994

Department of Health, Health Sector Strategic Framework, 1999-2004.

Department of Health, Smith MJ, Cornell J, Solanki G, A Study of Public priorities, Willingness and Ability to Pay for Social Health Insurance in South Africa, Progress Report, Strategy and Tactics, August 2001.

Department of Finance, 1998, Principled Approach to User Charges and Levies, Unpublished.

Donaldson C, Gerard K, Economics of Health Care Financing, The Visible Hand, Macmillan, 1993, pp.35-36.

Enthoven AC, "Why managed care has failed to contain health costs, Health Affairs, Fall 1993, pp.21 to 43.

Families USA, Bipartisan Consumer Health Bill Turned into Partisan Bill Harmfull to Consumers, Washington, 1996.

Families USA, Health Insurance Bill Clears Congress: Medical Savings Accounts Limited to Demonstration Programme, August 1996, http://epn.org/families.

Final Report of the Commission of Inquiry into Health Services, 1986, RP 67/1986. (Brown Commission).

Financial and Fiscal Commission, Recommendations 2001-2004 MTEF Cycle, May 2000.

Van der Merwe D, Social Transformation in South Africa by Means of Social Assistance: A legal Perspective (1998) 4; Devenish Commentary 73; Davis, Cheadle and Haysom Fundamental Rights 355. Also see Olivier M, Some Trends and Critical Issues in South African Social Security (30-31 March 1998)

Fischer, Ronald C., 1996, State and Local Public Finance, Second Edition, Irwin Chicago.

Gildenhuys, J.S.H., 1993, Public Financial Management, J.L. van Schaik, First Edition.

Gilson L, Social Health Insurance Stakeholder Analysis: Report of Preliminary Investigation, Prepared for the Directorate of Health Financing and Economics, Department of Health, Centre for Health Policy (University of the Witwatersrand) and Health Economics Unit (University of Cape Town), March 2001.

Ginwala K.N., Health Legislation in South Africa, Masters Thesis, University of Natal, 1981.

Health Care Working Party of the Society of Actuaries, Private Health Insurance in Ireland: Challenging Times, June 1995

Herber, Bernard P., 1975, Modern Public Finance, the Study of Public Sector Economics, Third Edition, Irwin Illinio.

Industry Commission, Private Health Insurance, Australia, 1996.

Jewkes R, Penn-Kekana L, Bradshaw D, Nannan N, Harrison A, Myer L, The Health of the Nation’s Youth, 2000, Contribution to the National Youth Commission Report on the State of the Nation’s Youth in the Year 2000.

Katz, M. 1995. Third Interim Report of the Commission of Inquiry into certain aspects of the Tax Structure of South Africa. Pretoria. Government Printers.

Kironde S, Tuberculosis, Chapter 17, South African Health Review 2000, The Press Gang, Durban, December 2000, pp. 335-349.

Marone JA, The ironic Flaw in Health Care Competition, the Politics of Markets, Competitive Approaches to Health Care Reform, eds. R.A. Arnould, R.F. Rich, W.D. White, The urban Institute press, Washington D.C.

Melamet Commission, the History and Development of Medical Schemes in South Africa, Unpublished, Paper prepared for the Committee, 1994.

Melamet Commission, Commission of Inquiry into the Manner of providing for Medical Expenses, 1994.

National Economic Research Associates, "The Economics of Health Care Reform: A Prototype", May 1993.

National Economic Research Associates, Risk-adjustment and its Implications for Efficiency and Equity in Health Care Systems, 1995.

Price M, Khunoane B, van den Heever AM, Should Tax Concessions on Medical Aid Contributions Be Removed?, Centre for Health Policy, 1995.

Ranchod S, McLeod, HD. Adams S, Low-Cost Options in Medical Schemes: The Need for Low-Cost Options and an Analysis of Benefit Designs Used in 2001, Centre for Actuarial Research, University of Cape Town, Monograph No 6: 1-77. December 2001a.

Ranchod S, McLeod, HD, Chronic Medicine Benefits in Medical Schemes: An Analysis of Benefit Designs in 2001 and Changes Since the Introduction of the Medical Schemes Act, 1998, Centre for Actuarial Research, University of Cape Town,Monograph No 7: 1-40. December 2001b.

Reinach Departmental Committee Regarding Medical Benefit, Friendly and Assurance Schemes, 1962.

Restructuring the National Health System for Universal Primary Health Care, Report of the Committee of Inquiry into a National Health Insurance System, Executive Summary, June 1995.

Shung-King M, Giese S, Hendricks M, Iriam J, Abrahams E, Guthrie T, Hussey G, Jacobs M, Proudlock P, Child Health, Chapter 19, South African Health Review 2000, The Press Gang, Durban, December 2000, pp. 365-391.

Teisberg EO, Porter ME, Brown GE, "Making Competition in Health Care Work," Harvard Business Review, July-August 1994.

Thomas S., Muirhead D., Doherty J., Muheki C., "Public Sector Financing", Chapter 5, South African Health Review 2000, The Press Gang, Durban, December 2000, pp. 127-147.

Trotman-Dickenson, D.I., 1996, Economics of the Public Sector, First Edition, Mackmillan London.

Trowbridge Consulting (internationally Woodrow Milliman), Unfunded Lifetime Community Rating, "preferred model for Australia", November 1997.

United Nations Economic and Social Council, Committee on Economic, Social and Cultural Rights, Consideration of Reports Submitted by States Parties Under Articles 16 and 17 of the Covenant: Concluding Observations of the Committee on Economic, Social and Cultural Rights (Canada), 10 December 1998, E/C.12.1.Add.31.

Van der Merwe D, Social Transformation in South Africa by Means of Social Assistance: A legal Perspective (1998) 4; Devenish Commentary 73; Davis, Cheadle and Haysom Fundamental Rights 355.

Van de Venn WPMM, Ellis RP, "Risk-adjustment in competitive health plan markets", Chapter 17, Handbook of Health Economics (eds. Culyer AJ and Newhouse JP, 31 March 1999.

van Vliet, van de Ven, Towards a Capitation Formula for Competing Health Insurers. An Empirical Analysis, 1992

Van Rensburg H.C., Harrison D., History of Health Policy, Chapter 4, South African Health Review 1995, Health Systems Trust and Hemry J. Kaiser Family Foundation, 1995,m pp.53-71

White Paper for the Transformation of the Health System in South Africa, published as Notice 667 of 1997 in the Government Gazette no. 17910.

Wilson VM et al, Case Study: The Washington State Health Care Authority, Inquiry 35, Summer 1998, pp.178-189.

World Health Organisation, The World Health Report 2000, Health Systems: Improving Performance, Geneva.