IU researchers partner with LinkedIn to publish labor flow study

A small team of researchers at Indiana University, in collaboration with LinkedIn, has created the first-ever global map of labor flow—how people move between jobs and industries in geographic regions.

The research, which was published in an academic journal Aug. 1, began in the fall of 2014 with the launch of LinkedIn’s Economic Graph Challenge, a global research competition that sought out researchers and academics to pitch ideas about ways to better understand global economic issues by utilizing LinkedIn’s array of user data.

The world’s largest professional social networking website has a global user base of more than 500 million people who frequent the site to create individual, resume-type profiles of professional experience and accomplishments, as well as connect with colleagues, classmates and potential employers.

“They’re sitting on this huge trove of information about people: where they work, career paths, what jobs they transfer into, [and] how people move within different industries,” said Kevin Fryling, senior news and media specialist for science at IU. “It’s just so much information that you need experts to try to comb through it all.”

IU’s team was one of 11 finalists chosen out of 200 applicants in May 2015. Each finalist was awarded $25,000 by LinkedIn. In 2017, IU’s team was selected as one of just two teams to continue research past that year—the other being Massachusetts Institute of Technology.

“It was good that we had more time to make our research more robust,” said Jaehyuk Park, one of the students on IU’s research team. “We could do a lot more analysis to make the research we had make sense in other areas and contexts.”

The participating researchers from IU—four Ph.D. students who worked alongside Yong-Yeol Ahn, an associate professor in IU’s School of Informatics, Computing and Engineering—were granted special access to LinkedIn’s job history database from its users between 1990 and 2015, totaling about 130 million job transitions between more than 4 million companies.

Park said he and Ahn both have backgrounds in economics and studying the labor market, so they together formed the idea to map global labor flow for the LinkedIn challenge.

“The work the team from Indiana University has done stands out in terms of showing how you can sketch out the structure of the global economy—at a large or small scale—using talent flows,” Guy Berger, principle economist at LinkedIn, said in a written statement to IBJ. “We look forward to seeing what the eventual impact of their work, as well as that of the other teams, will be for the 3 billion members of the global workforce.”

IU’s research findings argue that geo-industrial clusters—specific geographical regions defined by labor flow—provide useful insights into the growth and movement of the economy. Park said the research is also intended to help people and policymakers prepare for the future of the labor market.

“By just keeping track of people’s movement, we can detect people or industries or regions that are related to each other,” Park said.

Despite four years of rigorous data analysis, the research still has some bias because it is mostly U.S.-based, Park said.

“If local governments or even federal governments could collect data about skill sets of people, then that means they can use this in a more representative way,” Park told IBJ.

The data doesn’t dive into individual communities, but Fryling said this leaves opportunity for further research to be done.

“They’ve shown how we can do this, but individual analyses to answer specific questions will need to come later,” Fryling said. “Whether they do that in specific studies, I’m not sure. This is like the super higher level, it’s the global economy.”