Abstract
Most of the literature on the rapid changes in institutional arrangements in China's construction industry have either limited themselves to a description of these changes or relied on opinion surveys to draw conclusions. Few researchers have conducted serious empirical examinations, especially regarding the ownership of developers or contractors. To fill this gap, we empirically examined the effectiveness of compulsory construction supervision arrangements on the quality of construction projects in mainland China, explained the seemingly inexplicable phenomenon that projects undertaken by state‐owned enterprises (SOEs) are of better quality than those undertaken by non‐SOEs, explored the law of tort in relation to construction supervision firms, tested the published official data from 1991 to 2001 on construction ‘good quality rates’ with regression against per capita values and powers of machinery as well as the capital–output ratios of both SOEs and non‐SOEs, identified the determining factors affecting construction quality, and demonstrated the limitations of using the principal–agent theory in analysing the construction industry. The results supported the corollary of the Coase Theorem that resource allocation in terms of output quality will be affected by changes in the institutional arrangements.