Institutional arrangements and construction safety in China: an empirical examination

Abstract

Previous studies merely attempted to identify factors affecting safety performance from the micro perspective of project management, while they never questioned why those factors, such as lack of the top‐management support and training, existed. A macro perspective, informed by the corollary of the Coase Theorem, was adopted here to show that suitable institutional arrangements are essential for better safety performance in a construction industry. Province‐level construction safety records in China from 1994 to 2000 were used to test the effects of institutional arrangements on construction safety in China. Major factors affecting construction safety were found to be the implementation of construction safety laws, the rate of subcontracting, while neither the extent of using temporary workers, nor the availability of resources, nor the level of per capita GDP have any effects.

Keywords:

• Safety
• Coase Theorem
• institutional arrangements
• farmer worker
• subcontracting

Notes

1. These were enacted on 30 September 1989 by Order No. 3 of the Ministry of Construction of China, effective from 1 December 1989.

2. Clause 45 of the Construction Law.

3. Relevant provisions can be found in the Construction Project Safe Production Administration Ordinance, which was enacted on 24 November 2003 by Order No. 393 of the State Council and became effective on 1 February 2004.

4. Clause 18 of the Safe Production Law.

5. Clause 2 of the Safe Production Licensing Ordinance.

6. This is stated in the Provisions on the Procedures of Reporting and Investigating Serious Construction Accidents.

7. This is stated in the Provisions on Supervision and Administration of Safe Production in Construction, which was enacted on 9 July 1991 by Order No. 13 of the Ministry of Construction of China.

8. Clause 23 of the Construction Project Safe Production Administration Ordinance.

9. Source: China Building Industry Year Book 2001.

10. The data also contained a few mistakes. For instance, the total number of accidents was not equal to the sum of the accidents for individual classes. They were amended to make the results more acceptable.