Analysing workplace safety policies in hong kong with a simulation method

Abstract

Despite its impressive economic performance and high per-capita income, Hong Kong's workplace safety record lags behind leading industrial countries. This paper develops a computable equilibrium model of workplace safety in Hong Kong that predicts the changes in safety levels after substantial policy modifications. We find that accident rates would fall if workers' compensation insurance was more experience-rated or if fines for violations of the safety codes were raised. We also find that a progressive injury tax aimed at accident-prone firms would sharply lower the accident rate. In contrast, large increases in workers' compensation benefits have only a minor effect on the safety level.

Keywords:

• Compensating differentials
• workplace safety
• computable equilibrium model

Acknowledgements

The authors thank the financial support from the Hong Kong Research Grants Council (Grant No. LC829/96H). The comments made by two co-investigators on the RGC project, Ho Lok-Sang and W. S. Siebert are gratefully acknowledged. The authors also thank Thomas Kniesner and John Leeth for their advice on the simulation and Ng To-Hong and Gary Wong for their excellent research assistance. Finally, the authors thank Mark Hitchen from the UK Health and Safety Executive and Kam Chi-kit from the Hong Kong Labour Department for providing us with data and information used in this project. The authors retain the responsibility for any errors.

Notes

¹Based on the figures published by the International Labour Office's Bureau of Statistics, the occupational fatality rates in 1995 are 9.8, 5 and 1.1 in 100,000 workers for Hong Kong, the US and the UK, respectively. US figures have to be read with caution as the Bureau of (Continued )Labor Statistics in the US treats traffic fatalities while working as workplace deaths while for both Hong Kong and the UK, treatment depends on the circumstances. For example, a truck driver killed by being struck by a car while he was walking from his truck to make a delivery would be a workplace death but if he died inside his truck it would not.

²For a recent review on the econometric studies on this subject and their related problems, see Viscusi & Aldy (2003).

³Wei (1999) tried to use separate labour market data in the UK to uncover the underlying demand function for safety. However, it is often difficult to test the exclusion condition for such kind of econometric studies.

⁴As there are no Census data available for 1995, we use the 1991 Census to estimate the hedonic wage equation for 1995. We assume the labour market has not changed much during the period 1991 to 1995. We converted 1991 wages to 1995 wages using the annual changes in the nominal wage index.

⁵According to statistics collected by the Labor Department, the most dangerous industry in HK has a fatality rate of 0.00098 per 100,000 worker-years in 1995.

⁶This is actually the average fatality rate from our 1991 Census sample. However, published figures from the Labor Department (Report of the Commissioner for Labour 1995) indicate that the average industry fatality rate in 1995 is about 0.0001, which is quite close to this figure.

⁷According to estimates by the Health and Safety Council in the UK, the total output costs of accidents are about 0.5% of total profits.

⁸According to industry claims workers' compensation premiums range from 0.1% of payroll costs for the safest firms to 10% of payroll costs for the riskiest firms (South China Morning Post, 11 December 1993).

⁹This means firms have to pay 55% extra in premiums to the insurance companies for the same amount of benefits to reach workers' estates in a no-load insurance system. This was estimated as the case by England (1989)

¹⁰The details of this model and the results are available from the authors.