Unlocking the PBFF. An evaluation of New Zealand's Population-Based Funding Formula

Abstract

New Zealand’s District Health Boards (DHBs), which are responsible for delivering health services to their populations, receive their share of public funding according to the Population-Based Funding Formula (PBFF). The PBFF is intended to fairly distribute funding between DHBs according to the relative health needs of their populations and the costs of providing services to meet those needs. In 2014/15 the PBFF distributed $12 billion amongst DHBs, representing over two-thirds of the total health budget. Consequently, the PBFF exerts considerable influence on the health sector.

Since its introduction in 2003, the PBFF has been subject to vigorous criticism. Significant differences in per capita funding, in conjunction with disparate financial circumstances, across the 20 DHBs has resulted in ongoing tensions within the health sector. Compounding uncertainty over the fairness of the allocation process, there is limited publicly available information regarding the methods underpinning the PBFF and little independent scrutiny of them.

The aim of this thesis is to independently evaluate the PBFF, elucidating the current allocation methods and evaluating the capacity of the core determinant of PBFF allocations, known as the cost weights, to explain variations in need, use and expenditure.

Four main studies – corresponding to chapters in the thesis – are undertaken to evaluate the PBFF. First, the methods underpinning the PBFF allocations are investigated and a methodological précis developed. Second, the PBFF is viewed from a comparative perspective and the main features of, as well as nuances behind, funding formulae across seven high-income countries, including New Zealand, are considered. Third, the relationships between health need, use and the PBFF cost weights are explored using data from the New Zealand Health Survey. Fourth, the predictive accuracy of the PBFF is evaluated at the individual and, crucially, the DHB-levels using routinely collected inpatient discharge data, as well as Census data, and compared against alternative models informed by richer demographic and clinical data.

Overall, the results of these studies suggest that, despite its stated intentions, the PBFF does not support the fair allocation of health funding between DHBs. The vast majority of funding is allocated according to DHBs’ population demographics. However, the agglomeration of the population into relatively few demographic clusters, in particular with respect to ethnicity, appears to obscure important differences in needs, use and expenditure between DHBs. Comparisons with international funding formulae underscore the relatively simplistic approach of relying exclusively on demographic data as a proxy for more fundamental drivers of demand for health care. Consistent with international evidence and practice, incorporating clinical data reflecting casemix or clinical complexity into the PBFF could considerably improve the predictive accuracy and equity of allocations between DHBs.

These findings provide a cogent argument for revisiting the rationale and construction of the PBFF. Given the centrality of the PBFF to the New Zealand health system – which emphasises population-level health as well as the reduction of inequalities – there is a pressing need for transparent and inclusive discourse on the impacts of the PBFF, as well as alternative funding formulae, on different populations in New Zealand.