To contribute to the debate about costing Universal Health Coverage (UHC), the WHO Country Office in collaboration with WHO HQ developed the attached brief to help think through the implications.  “How much will the reform cost?” is a common and sensible-sounding question in planning any major health reform. Indeed, it is a question that national finance authorities have to ask. Unfortunately, the answer is complicated – although the expectation is that it is simple and straightforward. The accounting view proposes an average cost of service production that can be calculated and these unit costs can be used to project future scenarios based on different levels of service use. The implicit assumption is that cost and production functions for health services are essentially fixed. This paper outlines why it is misleading and potentially dangerous to use current unit costs as the basis for projecting the future.

To contribute to the debate about costing Universal Health Coverage (UHC), the WHO Country Office in collaboration with WHO HQ developed the attached brief to help think through the implications.  “How much will the reform cost?” is a common and sensible-sounding question in planning any major health reform. Indeed, it is a question that national finance authorities have to ask. Unfortunately, the answer is complicated – although the expectation is that it is simple and straightforward. 

The accounting view proposes an average cost of service production that can be calculated and these unit costs can be used to project future scenarios based on different levels of service use. The implicit assumption is that cost and production functions for health services are essentially fixed. However, it is misleading and potentially dangerous to use current unit costs as the basis for projecting the future for three reasons.
 
First, we don’t know where we are on the cost function. Cost is a function that describes the relationships between inputs and outputs. Any unit cost that we observe in a cost accounting study is just one point on a curve – but we do not observe that curve. Second, we don’t want to lock today’s inefficiencies into future estimates. The observed costs –or more accurately, expenditures –also embody inefficiencies that exist in the health system.  Third, changing the health system’s cost structure by re-configuring service delivery is a legitimate objective of many health reforms. Therefore, costing studies can be very useful – but only if they reveal information on the underlying cost structure of service delivery and enable the modelling of different scenarios using various assumptions about prices, the impact of incentives, changes in service delivery configuration, and levels of service use, e.g. primary care driven system.
 
For example, there are high cost scenarios (e.g. costing based on a hospital-centric configuration, using prices similar to those that currently exist in the private sector, and paying providers on a fee-for-service basis); as well as low cost scenarios (e.g. unified national pool that incorporates risks of the whole population, focused on primary care service delivery with total expenditures driven by current public sector prices, and payment mechanisms that operate within an overall budget cap).
 
Another critically important policy aim is to increase administrative efficiency. For example, would the new administration of National Health Insurance (NHI) be additional to the current bureaucracy (e.g. for quality assurance mechanisms)? Will the information systems required to run the pooling fund replace those currently used? What will be the implications for productivity in response to changes in provider payment methods? These fundamental choices need to be considered as planning proceeds –and a good cost analysis can help to raise these issues that encompass not only the national health insurance fund but the implications for the system as a whole.
 
Ultimately, "what will the UHC cost" depends critically on how it is designed and implemented. In that sense, looking at costing scenarios and assumptions may be valuable for raising some core policy issues. The process thus brings to the surface key choices and implementation issues that have to be made to enable the government to sustain the reforms.
 
What will not be useful is a focus on getting to “the number” – some point estimate that offers a false certainty about what the resource requirements will be. By using cost-accounting, projected resource requirements frequently are considerably more than currently available or reasonably projected revenues. This can lead to an endless cycle of revisions and efforts to dream up new revenue sources—thus focusing on issues that have more to do with tax policy than health policy. The experience of Tanzania is instructive –where the focus on “filling the gap” diverted policy attention away from where it needed to be: how to enable their new National Health Insurance fund to pool diverse funding sources and develop into a strong purchaser to drive efficiency and equity gains.
 
In summary, “what will NHI cost?” is the wrong question – or at best – an incomplete question. It would be better to frame the costing question around the implications of different scenarios for implementing reforms towards achieving UHC.

Click here to read the complete brief.