The Efficiency of New Zealand District Health Boards in Administrating Public Funds: An Application of Bootstrap DEA and Beta Regression

ABSTRACT

This study uses quarterly data from 2011 to 2018 to evaluate the technical efficiency of New Zealand District Health Boards (DHBs) in providing hospital services. It examines how efficiency is affected by various patient structures and contextual factors. An intertemporal data envelopment analysis and bootstrap approach are used to compute the bias-corrected technical efficiency scores, followed by highly flexible beta regression to assess the relationship between technical efficiency and related factors. The results indicate that the technical efficiency levels of New Zealand DHBs have not improved since 2011, and on average DHBs could increase their provision of hospital services by approximately 12%. Furthermore, most of the poor performing DHBs operate in the area of high socio-economic deprivation. The results from beta regression show that DHBs providing hospital services in highly deprived areas are associated with a decreasing level of technical efficiency as the proportion of surgical, acute, Māori and Pacific inpatient increases. However, an increase in capital to labour ratio improves the technical efficiency of these DHBs. Therefore, policymakers need to formulate comprehensive strategies involving a longer time horizon that facilitates capital investments in critical technology and capacity development to improve the long-run efficiency performance of DHBs operating in the area of high deprivation.

KEYWORDS:

• Data Envelopment
• efficiency frontier estimations
• beta regression
• public healthcare management
• New Zealand district health boards
• bootstrap

Notes

1. Reported in New Zealand dollars.

2. The term decision making unit was introduced by Charnes, Cooper, and Rhodes (1978) in their seminal paper that introduced DEA into the efficiency literature.

3. Pure technical efficiency reflects the managerial ability of using resources efficiently by following the best practice frontier in the industry.

Additional information

Funding

This work was supported by the Auckland University of Technology, New Zealand [16927380].