Data for: Transient and persistent inefficiency traps in Chinese provinces

Technical efficiency indices are estimated by using the parametric stochastic frontier analysis. The regression is based on a fixed-effect panel translog stochastic production frontier model, which covers GDP, capital stock approximated by the perpetual inventory method, education augmented labour input, and a time variable. Persistent inefficiency trap is indexed by the difference between an underachieving province’ persistent (or average) efficiency level and national average efficiency level. Transient inefficiency trap is indexed by the difference between the underachieving province’s overall efficiency gap and its persistent efficiency gap. Determinants of transient inefficiency trap are identified by a panel data model. Determinants of persistent inefficiency trap are identified through a cross sectional data model.