In 2016, LinkedIn CEO Jeff Weiner shot videos like this to present his company’s ambitious vision for a global “Economic Graph” mapping every worker, company, job, skill, and school…on the planet. Yes, all “three billion people in the global workforce,” Weiner promised. I would stare back and say, “Really?”
Around this time, I decided to test the relevance of professional networking platforms like LinkedIn to work-seekers in emerging economies—specifically, unemployed youth in South Africa. Safe to say, these work-seekers — most of whom lacked a high school degree and two-thirds of whom lacked job experience— are not LinkedIn’s core customers. Nor is South Africa part of LinkedIn’s core market. But to produce a global economic graph, LinkedIn would need to reach them. When encouraged and taught, would South African youth use LinkedIn? Would using the platform affect their job outcomes? It turns out that the answers are yes, and yes — with implications not just for LinkedIn, but also for those of us working in international development.
Beginning in 2016 and ending in January 2019, RTI International funded and conducted a randomized control trial (RCT) of LinkedIn usage in South Africa in partnership with Harambee Youth Employment Accelerator, economists from Duke University’s DevLab, and LinkedIn. We found that work-seekers trained and supported to use LinkedIn were 10 percent more likely than their control-group peers to gain employment, an effect that persisted twelve months post-graduation. These youth found and retained jobs in South Africa’s banking and insurance industries, often in entry level call center positions. See the full paper with co-authors Laurel Wheeler, Rob Garlick, Patrick Shaw, and Marissa Gargano here.
Our analysis indicates that LinkedIn training increases account creation and platform use, which then helps work-seekers convert job applications into job offers and retain those jobs. LinkedIn use likely alleviates either or both of two information failures in the job market. First, LinkedIn provides demand-side information that helps firms screen work-seekers. Second, it provides supply-side information that helps work-seekers target their job search and perform well in interviews. Related to both, LinkedIn also provides a referential benefit, where firms and work-seekers may tap networks to gain third party references. We also find that it is not terribly difficult or costly to teach LinkedIn. The training RTI developed with Harambee cost roughly $20 dollars/person ($48/person in purchasing power parity) to implement, took four hours to complete (spread across an eight week employment readiness course), and produced a 66 percent increase in LinkedIn account creation. You can find the curriculum and approach we used here.
It is probably worth noting that I do not work for LinkedIn, and LinkedIn did not pay for or influence the results of this study. They did kindly, and bravely, provide access to data for the 1638 participants in our study, giving us critical information on account creation, account use, and network development. We combined these data with participant background and training performance data from Harambee, and survey data we collected at the beginning of training, end of training, and 6 and 12 months after training. We used these data to track LinkedIn uptake, differential usage, and other non-platform factors like changes in attitudes, aspirations, and education engagement. Despite our findings that the platform may provide an information premium, we did not find evidence that it reduced job search costs, changed participants’ education experience, or influenced participants’ future orientation (i.e. hopes and aspirations).
South Africa has 7.1 million active users of LinkedIn—about 40 percent of its workforce—and 240,000 jobs posted. Some of these users include our study participants, and their employers. These are the groups Jeff Weiner needs actively using the platform to realize his Economic Graph. Our data suggest it is worth their while. Yes, really.