Minimum wages effects on low-skilled workers in less developed regions of China

Purpose - The purpose of this paper is to estimate the causal effect of minimum wages on the employment of low-skilled workers in less developed regions of China. Design/methodology/approach - Based on data from the China Health and Nutrition Survey, a double-difference (DD) methodology is used to compare the employment of low-skilled individuals before and after a minimum wage increase in their provinces with a comparison group of individuals in provinces that did not have a minimum wage increase. Also, a triple-difference methodology (DDD) is used that also includes an additional control group of highly educated workers as a within-province internal comparison group that should not be affected by a minimum wage increase. Findings - No evidence of an adverse employment effect is found in any of the 36 different estimates, consistent with recent US evidence that uses a similar DD methodology. Research limitations/implications - The data are not national representative; rather heavily weighted towards the less developed Central, Western and parts of the Eastern Regions of China. This may partially explain the absence of the theoretically expected adverse employment effect. Other related reasons are discussed, including: lack of enforcement in those less developed regions; a large presence of state-owned enterprises in the regions where employment security clause remains intact; the relatively less developed labour markets in the regions including where employers may behave in a monopsony fashion in their labour markets; shock effects; and cost offsets from reduced fringe benefits and increases in the pace of work. This paper was unable to disentangle the separate effect of these possible factors. Originality/value - This is one of the few studies on minimum wages in China to focus on low-skilled workers in less developed regions, to use individuals as the unit of observation rather than aggregates, and to provide causal estimates based on DD and DDD methodologies.