Using confidential US Census micro data we find three results. First, there is no evidence that Chinese import competition generated net job losses. In low-human capital areas (for example, much of the South and mid-West) manufacturing saw large job losses, driven by plant shrinkage and closure. But in high-human capital areas (for example, much of the West Coast or New England) manufacturing job losses were limited, with much larger gains in service employment, particularly in research, management and wholesale. As such, Chinese competition reallocated employment from manufacturing to services, and from the US heartland to the coasts. Second, looking at the firm-level data we find almost all of the manufacturing job losses are in large, mul- tinational firms that are simultaneously expanding in services. Hence, these large firms appear to have offshored manufacturing employment while creating US service sector jobs. Indeed, we show large publicly traded US firms do not seem to have been negatively impacted by the rise in Chinese imports. Finally, the impact of Chinese imports disappears after 2007 – we find strong employment impacts from 2000 to 2007, but nothing since from 2008 to 2015
https://nbloom.people.stanford.edu/sites/g/files/sbiybj4746/f/bhkl_posted_draft.pdf