1. Modelling the resource requirements for universal coverage: A case study of South Africa Di McIntyre Health Economics Unit University of Cape Town SHIELD project: Strategies for Health Insurance for Equity in Less Developed Countries

2. Background Proposal to introduce universal coverage in South Africa: – Heated debate with particular concerns about ‘affordability’ Modelling useful to explore likely implications of different financing reforms: – Resource requirements (and impact of key design features) – Revenue potential

3. Broad approach used Identified the key scenarios: – What kind of financing reforms are being considered, proposed, or might be feasible? Resource requirements: – Population x utilisation x unit costs (Excel spreadsheet model) – Absolute amount and relative to GDP Revenue: – Health sector share of government budget; additional tax or mandatory contributions

4. Reform scenarios Status quo: – Limited growth in private insurance (<20% of population); limited improvement in public sector Extended schemes: – Mandatory extension of private insurance to all formal sector workers and dependents (<40%) Universal system: – Core of substantially improved public services; access to all; mainly tax funded including additional payment by formal sector

5. Core methods and assumptions Benchmarked baseline15 year projections PopulationSources: Census, schemes, SHIELD household surveys Disaggregation: Age & sex, insurance status Reliable projections (AIDS impact not adequately accounted for in official stats) UtilisationSource: SHIELD household surveys; benchmarked against HMIS & scheme data Disaggregation: Age & sex, insurance status, type of service Normative targets and cross-checked with international experience Unit costsData from Treasury & HMIS Scheme claims Real increases: trends & improved resources Expenditure benchmark Audited public & scheme expenditure

6. Excerpt of model

7. Coverage and utilisation

8. Total expenditure as % GDP

9. Public expenditure as % GDP

10. Public funding requirements Status quo: – Will decline to 10.5% of government budget Extended schemes: – Will decline to 9.4% of government budget (but increased costs of covering civil servants) Universal system: – Increase share of government budget to 15% plus – Proportional income tax of 4% (2% employer & 2% employee) – Or, progressive income tax of 1.2% to 6%

11. Conclusions UC is lowest cost option in SA, but places greatest burden on public funds and requires: Increased allocations for health in line with Abuja target, and Additional taxes Modelling illustrates likely resource requirements for reform scenarios, but also provides insights into reform design (sensitivity analyses)

12. Pre-conditions for success Cost containment: – Minimise unit costs, careful purchasing from non- government providers – Minimise administration costs: avoid fragmentation – Contain utilisation: strengthen primary & community services; improved gate keeping – Need for ongoing M&E, refine model with updated data

13. SHIELD partners For more informaton visit: http://web.uct.ac.za/depts/heu/SHIELD/about/about.htm Funding: European Commission and IDRC