Can intermediate input trade liberalization affect worker health in a developing country like China, and if so, how? Do the impacts differ between skilled and unskilled workers? What are the welfare implications of input tariff reductions once health factors are considered? Professors Haichao Fan of Fudan University, Faqin Lin of China Agricultural University, and Shu Lin of the Chinese University of Hong Kong develop...
Recently, a growing strand of literature has started to examine the health effects of trade shocks (e.g., Hummels et al., 2016; McManus and Schaur, 2016; Pierce and Schott 2016; Bombardini and Li, 2020). While most existing work focuses on import competition or export expansion, less is known about the role of input tariffs. Using input tariffs and the China Health and Nutrition Survey data (CNHS), this study (Fan, Lin, and Lin, 2020) finds that input tariff reductions adversely affect worker health through increased working hours. Moreover, input tariff reductions widen both the income and the health gaps between skilled and unskilled workers. Further welfare analysis indicates that ignoring health outcomes would substantially underestimate the welfare disparity between skilled and unskilled workers.
Four Stylized Facts
We first document four stylized facts regarding input tariffs, workers’ working hours, wages, and health before and after China’s WTO accession:Fact 1: There were large tariff cuts following China’s WTO accession, and the reductions applied mainly to imports of intermediate goods.
Fact 4: After trade liberalization, unskilled workers experienced larger increases in illness rates and working time but a smaller gain in income.
A Theoretical Model
To rationalize these facts and to motivate our empirical exercises, we build an illustrative model linking input tariff reductions to firm behavior, workers’ supply of labor, and their health status. In our model, firms choose the number of workers and each worker’s optimal working time. The optimal working time and the number of employed workers are substitutes and are both decreasing functions of the wage rate. To hire workers, firms need to pay a search cost (e.g., Helpman and Itskhoki 2010; Helpman et al., 2010). The impact of a reduction in input tariffs on firms’ demand of working time can be illustrated intuitively using a simple diagram (Figure 1). A reduction in input tariffs lowers a firm’s marginal costs and increases its profit. The demand curve for working time is accordingly shifted to the right.Workers in our model choose optimal working time to maximize their utility. The supply curve of working time is an increasing function of the wage rate. Based on existing findings in the labor economics literature (e.g., Juhn et al., 1991, 2002), we assume that the elasticity of labor supply to the wage rate is lower for skilled workers. As a result, in Figure 1, the supply curve of working time for skilled workers is steeper than for unskilled workers. Consequently, a rightward shift of the demand curve following trade liberalization has differential effects on skilled and unskilled workers. While equilibrium working hours and wages increase for both types of workers, the increase in working hours (wages) for unskilled workers is larger (smaller). Existing studies in the medical literature have provided overwhelming evidence on a positive association between working time and the likelihood of experiencing illness or injury (e.g., Sparks and Cooper, 1997; van der Hulst, 2003; Dembe et al., 2005). Rod et al. (2017), Hamermesh et al. (2017), and Berniell and Bietenbeck (2018) provide further causal evidence. Longer working hours, therefore, lead to a higher probability of becoming injured or ill, both the income gap and health gap widen following trade liberalization.
Figure 1. Optimal working hours and wage payment
We then put the predictions of our theoretical model to a test. Following the literature (e.g., Amiti and Konings 2007; Topalova, 2010), we construct a prefecture-level input tariff shock measure using the Chinese provincial-level input-output table and the prefecture-level initial industry composition of labor. Employing the CHNS database, we estimate the following specification.
, where i and t represent the individual and year, respectively. measures the health condition of individual i in year t. Our main measure of workers’ health status is a binary variable that takes the value of 1 if an individual has experienced illness or injury in the past four weeks and 0 otherwise. We also consider two alternative measures for robustness. First, since the CHNS survey provides information about whether a worker has felt uncomfortable or exhausted in the past four weeks, we will use this binary variable as an alternative. Second, we also utilize workers’ subjective evaluations of the severity of illness or injury as another alternative. The self-evaluated score ranges from 1 (the least severe) to 3 (the most severe). In addition, we will also examine the effects on different types of illness using a set of binary indicators.
is a comprehensive set of individual-level controls, including gender, years of education, age, age squared, disease history, smoking, possession of health insurance, occupation type dummies, and employer ownership dummies. We also control for a set of prefecture-level covariates, , in our benchmark specification. Following Facchini et al. (2019), we include in a set of prefecture-level measures of other policy changes, including the elimination of trade uncertainty (Handley and Limao, 2017; Erten and Leight, 2019; Facchini et al., 2019) and export licenses (Bai et al., 2017), changes in quotas due to the expiration of global MFA, and changes in tariffs on China’s exports abroad. In addition, we also include in a prefecture-level measure of air quality to capture the effect of air pollution on health. is the year fixed effects, which captures yearly shocks common to all individuals. We also include prefecture fixed effects, , to control for all time-invariant differences across prefectures in our regressions. is the error term.
What Do We Find?
We find that manufacturing workers in Chinese prefectures that had a larger exposure to input tariff reduction shocks experienced a significantly higher likelihood of suffering from illness or injury. This finding is not affected by controlling for other trade shocks and pollution variables. It is also robust to using only the two waves of the survey data closest to China’s WTO accession. In a comprehensive set of tests, we also rule out the possibility that our results are driven by pre-existing trends. Additional evidence from the placebo tests lends further support to our hypotheses.
Table 1. Benchmark results
Table 2. Welfare disparity between skilled and unskilled workers (for a one standard deviation reduction in input tariff shock)
Amiti, Mary, and Jozef Konings. 2007. “Trade liberalization, intermediate inputs, and productivity: Evidence from Indonesia.” American Economic Review, 97: 1611–1638.Bai, Xue, Kala Krishna, and Hong Ma. “How you export matters: Export mode, learning and productivity in China.” Journal of International Economics, 2017, 104: 122–137.
Erten, Bilge and Jessica Leight. 2019. “Exporting out of agriculture: The impact of WTO accession on structural transformation in China.” Review of Economics and Statistics, forthcoming.
Facchini, Giovanni, Maggie Liu, Anna Maria Mayda, and Minghai Zhou. 2019. “China’s “Great Migration”: The impact of the reduction in trade policy uncertainty.” Journal of International Economics, 120: 126–144.
Fan, Haichao., Faqin Lin, and Shu Lin. 2020. “The hidden cost of trade liberalization: Input tariff shocks and worker health in China.” Journal of International Economics, forthcoming.Hamermesh, Daniel S., Kawaguchi, Daiji, Lee, Jungmin, 2017. “Does labor legislation benefit workers? Well-being after an hours reduction.” Journal of the Japanese and International Economies, 44, 1–12.
van der Hulst, Monique, 2003. “Long workhours and health.” Scandinavian Journal of Work, Environment & Health, 29(3), 171–188.