Abstract

     This study utilizes a threshold model to examine the nonlinear relationship between working hours and job satisfaction, using the open-access data of the 2018 China Family Panel Studies. We address the endogeneity of working hours utilizing an instrumental variable-based two-stage residual inclusion approach. The threshold model shows that the effects are indeed different. Working more than 9 h reduces workers’ job satisfaction, and these reductions are even greater among those working more than 12 h. Heterogeneous analysis reveals that working long hours reduces the job satisfaction of female employees more than that of their male counterparts; the job satisfaction of unmarried individuals is unaffected by how long they work, whereas that of married workers declines when they work longer hours. Also, although the job satisfaction of wage-employed workers falls with an increase in the number of hours worked regardless of how long they work, that of self-employed workers falls only when they work more than 12 h. Poor physical health mediates the adverse effects of long working hours on job satisfaction. Finally, working long hours reduces individuals’ short-run hedonic well-being but does not affect their perceptions and feelings towards various facets of life in the long run.

Introduction

     According to the World Health Organization (WHO) and International Labor Organization (ILO), about 36.1% of the global workforce spends more than 48 h/week working (Messenger, 2018). In Europe, this number was 16% in 2015 (Eurofound, 2017). Overwork is endemic in many Asian countries (Yamashita, Bardo, & Liu, 2016). For example, Chinese Labor Law states that the statutory working time is 8 h/day, and the average working time is no more than 44 h/week. Some organizations, especially in the technology sector, expect employees to work from 9 am to 9 pm, six days per week (Dong, Wu, Ni, & Lu, 2021; ILO, 2018); there is a growing backlash against this draconian “۹۹۶ working system”. In Japan, the term karoshi, referring to overwork-related death, reflects Japan’s culture that exalts overwork. Working long hours is a universal concern with far-reaching implications for society.

Conclusions and policy implications

     Working long hours is exalted in developed countries, and developing countries are embracing overwork too. The draconian “۹۹۶” work regime in China exemplifies this phenomenon. However, there is a growing discontent towards working long hours, and it is rebuked by policymakers, media, and workers themselves, who are becoming dissatisfied with jobs requiring them to work long hours. This paper analyzes the nonlinear effects of working hours on job satisfaction using a threshold model. Considering that job satisfaction is endogenous, we employ the IV-based 2SRI model to address the endogeneity. This is the first paper to examine nonlinearities in the link between the number of hours worked and job satisfaction while endogenously determining the thresholds delineating the work durations associated with changes in job satisfaction.

     We find that how long people work indeed affects their job satisfaction. The nonlinear estimates derived from the threshold model confirm four specific thresholds for the number of hours worked/day signifying changes in job satisfaction: 8.571 for the singlethreshold model; 7.143 and 9.000 for the double-threshold model; and 12.000—in addition to the two threshold values for the double-threshold model—for the triple-threshold model. The findings of the threshold model confirm larger reductions in job satisfaction at higher thresholds of the number of hours worked. The effects of working hours on job satisfaction for males and females vary, as do the effects for married and unmarried individuals. Our results show lower job satisfaction from working long hours for males and married individuals but not for females and those who are unmarried. The type of employment people are engaged in also influences their job satisfaction differently—working long hours may lower the job satisfaction of wage-employed people more significantly than of those who are self-employed. Working long hours in both the agricultural and non-agricultural sectors significantly reduces job satisfaction once working hours reach the above thresholds. The mechanism analysis reveals that poor physical health mediates the adverse effects of working long hours on job satisfaction. Furthermore, working hours are negatively associated with people’s selfreported happiness.