Productivity changes in Taiwanese hospitals and the national health insurance

Abstract

In 1995 Taiwan launched the National Health Insurance (NHI) programme, which changed the method payment for hospital finances. This article examines the impact of this financial reform on hospital productivity during this period. Using the Malmquist productivity index approach, a hospital's change in productivity is decomposed into quality, efficiency, and technological change components. Factors affecting efficiency and productivity are also assessed. Results indicate that most hospitals experienced a significant productivity slowdown due to declines in technology and quality of service, but efficiency did significantly improve. The inception of the NHI programme does greatly improve a hospital's productivity and quality of service, but decreases efficiency. Public hospitals' efficiency improvements are significant.

Acknowledgements

The author thanks two anonymous referees for their valuable comments and suggestions. Special thanks are also offered to Professor Chee-Ruey Hsieh, Institute of Economics, Academia Sinica, Taiwan and Professor Wei-Der Tsai, National Central University, Chungli, Taiwan for their comments on an earlier draft of the paper. The empirical data provided by the Department of Health in Taiwan is greatly appreciated.

Notes

1. The three prior major programmes are Labour Insurance, Government Employee Insurance, and Farmer Health Insurance. Members of the public enjoy medical insurance coverage through enrolment in the plan corresponding to their occupation [NHI Annual Statistical Report, Taiwan 1995].

2. For derivations of these bootstrap approaches, see Efron and Tibshirani [1993], chapters 12–14.

3. The p values for the Z statistics of the Wilcoxon signed-rank test are less than 0.10 for all comparisons with/without the incorporation of the attribute variable.

4. The number of each bootstrap replication is 2000, which is larger than the square of the sample size 1600 [Atkinson and Wilson, 1995].

5. Average length of stay is calculated by inpatient days divided by inpatient frequencies.

6. Severity of illness is computed by the ratio of surgery frequencies divided by inpatient frequencies.

7. Bed utilisation rate is calculated by inpatient days divided by total beds.

8. Market concentration ratio is computed using total hospital beds as the ratio of total aggregate hospital beds in the medical region.

9. The White test shows that heteroscedasticity does not exist in the error term of the OLS regressions.

10. The VIF (Variance Inflation Factors) test shows that the problem of multicollinearity among the regressors is not serious.