An article that appeared in the Business Day, September 4 unpacks the costs attached to implementing the ANC’s recently proposed NHI.
Servaas van der Berg and Heather McLeod
Government is presently considering a National Health Insurance (NHI) scheme to come to grips with a fundamental service delivery problem ‘ the continued failure of the public health system to offer quality care. Though the 16% of the population belonging to medical and bargaining council schemes receive good health services, public health care quality is so inadequate that 30% of people without medical scheme cover pay for private treatment out of pocket. Although only 4% of households claim that expense has prevented them from visiting a health facility, there is a serious need for access to quality care.
Our fear is that the proposed NHI will fail to meet expectations of the poor, will leave medical scheme members (including the working poor) worse off, will be massively expensive or even completely fiscally unaffordable, and will require far more doctors and nurses than are available. The danger is that it could well become a highly costly failure that will further increase frustration with service delivery.
The outlines of the proposed scheme remain sketchy (from a leaked 200 page ANC document and a 4 page elucidation in ANC Today):
· The NHI will be a publicly administered fund that receives funds, pools resources and purchases services on behalf of the entire population.
· Universal coverage is a starting point: ‘All South Africans will be equally covered to access comprehensive and quality health care.’
· Generous benefits (‘a comprehensive range of health benefits‘) are promised.
· Free access to quality health care at the point of service for all, ‘no upfront payment will be required by the doctor or hospital‘.
· Everyone above the tax threshold will contribute through a national payroll or income tax.
· The NHI will purchase services from accredited private and public hospitals and from doctors. Exactly how this will work is unclear, but the document emphasises public hospital services with private doctors as providers of primary care.
· All of this, it is said, will be done within the current funding envelope: ‘requiring no increase in total health care spending as a percentage of GDP.’
· At an individual level, ‘Contribution will be less than what members and their employers currently pay to medical schemes.’
There are a number of contentious assumptions, explicit or implicit, in the plan. Firstly, it implies that more money will successfully convert struggling public hospitals into viable service institutions, while somehow getting private hospitals and GPs to provide services at lower costs. Secondly, when politicians promise universal and free access to quality health care at the provider of one’s choice, their audience probably expects more than simply continued access to a better funded local clinic or Chris Hani Baragwanath Hospital. Thirdly, the document assumes that additional visits to doctors can simply be purchased by the NHI fund. Most poor people seeking treatment presently see a nurse (mainly in public clinics), while the wealthier see doctors. Demand projections indicate that satisfying public preferences would require double the current number of doctors’ visits, as the public presently interprets quality care as access to private facilities and doctors. Finally, a core and untested assumption is that medical schemes are delivering inefficient care that the state can supply at a much cheaper rate. What would full comprehensive care really cost in the public sector if quality is improved to the levels people expect?
Due to the lack of detail provided, it is difficult to determine what it will cost to implement the plan. The only hint in the document about what constitutes a ‘comprehensive range of services‘ is that cosmetic surgery would be excluded. This means important choices have been postponed on what procedures to exclude ‘ reflecting the lack of technical grounding of the proposal. The ANC document left cost projections for the future, saying ‘the contribution rates for formal sector workers could start with a nominal percentage in the initial years and increase to x% … and real increase from general tax revenue at x%’. A costing by Sule Calikoglu and Patrick Bond offers a range from R134 billion to R231 billion, with a preferred model of R205 billion for 2006. A preliminary costing by Di McIntyre, John Ataguba and Sue Cleary produced a cost of R77 billion in 2010, escalating to R169 billion at current values by 2020.
One way of estimating the envisaged expenditure is to work backwards from the excellent private sector cost data. Though the NHI promises a ‘comprehensive range of health benefits‘, we assume, to be on the conservative side, that the fund would provide only a Basic Benefit Package (BBP), entailing provision for basic primary care in addition to the Prescribed Minimum Benefit (PMB) package that all medical schemes have to include. The graph shows actuarial costs for different age and gender groups for this package of benefits in private medical schemes. After the first years of life, costs generally rise with age. Women have higher costs in the childbearing years, while men incur higher average costs at older ages because they do not survive as long as women. Applying these same costs to the full population (who are younger and therefore cheaper to serve than medical scheme members), R5 140 per person per year would be needed for this package in the NHI (excluding administration costs and costs associated with HIV/AIDS), i.e. R251 billion for the full South African population. This can be compared to a more restricted and a more comprehensive package of benefits:
- Prescribed Minimum Benefits or PMBs only R156 billion
- Basic Benefit Package (PMBs plus primary care) R251 billion
- Fully Comprehensive Benefit Package R334 billion
Even if one accepts the optimistic view that public sector provision could be up to 30% cheaper (though this ignores service quality) and applies that 30% reduction across the board, the funding required to fund the NHI at BBP benefit levels is a full R176 billion ‘ a massive amount compared to budgeted income tax revenue of R206 billion and public health expenditure of R84 billion!
A payroll tax of over 17% would be needed to fund such a conservatively estimated revenue need. (To exempt poorer workers would require higher rates or subsidies from general tax revenue). Alternatively, to fund the NHI through an income tax would require tax rates to increase by 85% across the board ‘ the top marginal tax rate of 40% would have to rise to 74%! Our estimate of the tax costs is similar to that of a 2005 Ministerial Task Team, who declared an NHI ‘unaffordable at the existing level of economic development‘, even with fiscal substitution (i.e. reducing other public health spending). Laying claim to present medical scheme contributions for the NHI (around R76 billion) is not a solution for getting financial resources for the public system, as scheme members will want to continue receiving the benefits currently covered.
The NHI proposal can only be taken seriously once a proper analysis of its costs, fiscal consequences and affordability has been undertaken. The current proposal is beyond what the country can afford. Moreover improving the quality of health care requires more real resources ‘ doctors, nurses, hospitals, equipment ‘ that do not automatically follow funds, but depend on management and incentives in the public health system. History teaches us that improving such management is difficult to achieve in a moderate time frame. Indeed ‘(t)he most urgent need is to ensure that the public sector is well-capacitated to deliver the envisaged services at all levels‘ ‘ and this is not mainly a question of money.
Professor Servaas van der Berg
Professor of Economics and National Research Chair in the Economics of Social Policy, University of Stellenbosch
Professor Heather McLeod
Extraordinary Professor in Statistics and Actuarial Science at the University of Stellenbosch and Visiting Associate Professor in Public Health and Family Medicine at the University of Cape Town
To access the above graph, click on the linked pdf file.