The Amazon Jobs Effect: Kenosha County, Wisconsin

By / 2.1.2018

Amazon is the fastest US company–and perhaps the fastest company anywhere–to 300,000 workers. Its rapid expansion is creating tech-enabled work in virtually every corner of the country, with our estimates showing that fulfillment center jobs pay 31% more, on average, than brick-and-mortar retail jobs in the same area.

Now, there are all sorts of interesting questions about what happens next. Some people have worried that the fulfillment center jobs will fade away as the operations get increasingly roboticized. By contrast, our view is that fulfillment centers will become critical hubs for the new “Internet of Goods“: By lowering the cost of shipping and creating a pool of tech-enabled workers, areas with ecommerce fulfillment centers will  have a head start in attracting the next wave of manufacturing startups.

The answers to these questions, of course, bear on the important debates about the value of tax and other public incentives for Amazon fulfillment centers and the company’s HQ2. I haven’t gotten involved in these discussions directly, because they really are about the shape of the future economy. If you think that robots are going to eat all of our jobs, then tax incentives never make sense. If you think that we are just at the beginning of the transformation of  physical industries and the creation of a new wave of tech-enabled jobs in physical industries such as distribution, manufacturing, and agriculture, then offering tax incentives to get a piece of the future is far-sighted thinking.

However, in the midst of all of these very interesting discussions, I really must address a new study from the Economic Policy Institute which purports to show no employment gains from the opening of an Amazon fulfillment center. More precisely,  “[t]wo years after an Amazon fulfillment center opens in a county, overall private-sector employment in the county has not increased.”

Really?  This result does not pass the smell test. You can raise all sorts of long-term questions. But in the short-term, if you build a giant new fulfillment center, first you get construction jobs in the years before the center opens. Then you get the workers themselves. There’s no plausible mechanism by which those jobs can crowd out other jobs in the samecounty in the short-term.

And I’m not sure about their sample of counties. I look on their list of ‘Amazon’ counties (Appendix Table 3), and it doesn’t include Kenosha, Wisconsin, where the construction of an Amazon fulfillment center in 2013 and 2014, and its opening in June 2015, added thousands of jobs into the local economy.

The chart below plots private sector jobs in Kenosha County against private sector jobs in all of Wisconsin.


You can see that right around the time that Amazon arrives in Kenosha, county employment turns up, driven in large part by the increase in warehouse jobs.

Indeed, Kenosha County is effectively becoming a tech-enabled distribution-manufacturing hub. After Amazon opened its doors, the county attracted companies like Haribo, the German candy giant (and originator of gummy bears), which is building its first North American factory in Kenosha.