Do Hospital Reform and Ownership Matter to Shenzhen Hospitals in China? A Productivity Analysis

Kok Fong See*, Ying Chu NG

*Corresponding author for this work

    Research output: Contribution to journalJournal articlepeer-review

    2 Citations (Scopus)


    The present study enriches the literature on hospital performance by analysing the total factor productivity growth of Shenzhen hospitals for the period 2005–2013 during which the 2009 Reform was implemented in the hospital sector of China. Unlike most previous studies, we adopt the global metafrontier Malmquist productivity index approach which allows the relaxation of the constraint of technological heterogeneity for hospitals by ownership. The analysis showed that there was a slightly rise in productivity of 0.68% on average. The technological regression experienced by the sampled hospitals out weighted the improvement in efficiency change. By hospital ownership type, public hospitals exhibit an average productivity decline by 0.67%, while non-public hospitals enjoyed an average productivity growth of 3.28% in general. Technological change is mainly the source of productivity change. While public hospitals are found to be the technology “leaders”, non-public hospitals have been catching up with the frontier over time. High bed utilisation was found to enhance technological progression, boost productivity growth, and allow hospital to catch up with the frontier. However, 2009 Reform exerts negative impact on the productivity growth and technological regression of the sampled Chinese hospitals.
    Original languageEnglish
    Pages (from-to)145-155
    Number of pages11
    JournalEconomic Analysis and Policy
    Publication statusPublished - Dec 2021

    User-Defined Keywords

    • Chinese hospitals
    • Hospital reform
    • Metafrontier framework
    • Ownership
    • Total factor productivity


    Dive into the research topics of 'Do Hospital Reform and Ownership Matter to Shenzhen Hospitals in China? A Productivity Analysis'. Together they form a unique fingerprint.

    Cite this