For the past 30 years, we’ve taken it for granted that major companies have globalized their production and fragmented their supply networks to far-flung corners of the world in search of the best value.
“Offshoring” has long been a subject of debate in conversations about economic growth and job protection, but there has been very little empirical research into why the phenomenon has occurred. Within academic circles, it’s often assumed that in addition to seeking out lower costs, companies have been driven to fragmentation by improvements in technology that have allowed them to communicate more effectively with suppliers overseas.
However, there has been little evidence to support that assumption, says Teresa Fort, an assistant professor of business administration at Tuck.
“There’s been a lot of theoretical work, a lot of models that people have written down, but it’s really hard to look at the plant level or firm level and see exactly who is doing it and what factors are driving them,” says Fort, who used to work as a research assistant for the U.S. Census Bureau. Fort’s Census Bureau connections gained her access to an exclusive dataset showing the exact location of all U.S. manufacturing plants, along with their sales and employment. By tying those data to statistics on the level of imports by firms, it’s possible to assess how and why plants are sending out their production.
Fort is presenting her findings in a new working paper that paints a rather surprising picture of offshoring. For starters she finds that the number of plants engaged in sourcing customized materials from overseas suppliers is quite small.
Nationwide, only 2 percent of plants are offshoring these inputs. However, that percentage changes dramatically when you consider the size of the company. Among firms, 27 percent of sales and 19 percent of employment occurs at companies with one or more plants that source the majority of their fragmented production from foreign locations. “Not a lot of plants are offshoring customized inputs, but those that do tend to belong to really big firms,” concludes Fort.
Interestingly, it’s more common for plants to outsource their production to suppliers in other U.S. states. Compared to the 2 percent of plants sending work overseas, almost 30 percent of them are sourcing from other domestic plants.
Here, too, there is variation. Plants in states with high wages, such as those in the Northeast and West Coast, are much more likely to send their work to lower-wage states such as those in the South. In contrast, plants located in low-wage states that fragment production are more likely to send their work overseas, since they must look elsewhere to access lower labor costs. It’s a much more nuanced picture of outsourcing, and doesn’t fit so easily into political categories.
“You can get both Obama and Romney to say we don’t want to ship jobs overseas. But the political economy ramifications are a lot more complicated when you talk about Boeing shipping jobs from Washington to South Carolina,” says Fort. “There will be a debate within the country about whether that’s a good thing or a bad thing.”
Fort’s study also undermines the idea that advances in communication are driving jobs overseas. When she analyzed the data of plants engaged in high-tech forms of manufacturing, she found that while advanced technology can facilitate movement of production overseas, such movement is limited by the sophistication of the receiving country. “Companies are offshoring to countries with cheaper labor, but those countries also tend to have lower levels of technology—so higher tech firms need to source from higher tech locations.”
As a result, higher tech firms that offshore are more likely to source from more expensive countries such as Japan, but less likely to source from countries such as Bangladesh, which have less technology.
Since wages in higher technology countries are often comparable to those in the U.S., increases in communication technology might actually aid in returning jobs to the U.S. At the same time, the greater worry—at least for some states—may be losing jobs to other lower wage states, where easier communications can decrease costs associated from coordinating production across locations, but highly skilled workers can also still take advantage of those technologies.
“The differences between states in terms of worker skills is much smaller than the differences between countries,” says Fort. “South Carolina is going to be a lot more homogenous in terms of skills of workers compared to Bangladesh.”
Given that distinction, offshoring may not be nearly the problem in the future that “onshoring” could be.
Expect to see that issue coming soon to a political debate near you.
T. Fort, “Firms’ Organization of Global Production: Theory and Evidence,” working paper.
Teresa Fort is an assistant professor of business administration at the Tuck School of Business.