What if you had the information to identify the fastest growing jobs in your country, or even in your hometown, and the skills you need to learn to get hired in them? And what if this data could also help you decide which jobs you could most easily transfer to with just the skills you have today should you want a career change or if your current job might be soon automated?
Obtaining the information to answer these questions is a primary concern for policy makers, educators, students and workers everywhere at a time where the exponential growth of digital technologies, combined with the rapid development and deployment of robotics, arti cial intelligence, the Internet of Things, and new platform technologies like LinkedIn, is accelerating the pace of technological change and creating important shifts in the workforce.
Obtaining better and timelier insights into the changing demand for skills and occupations has become a pressing challenge for all countries. However, despite a global debate about the future of work and the changing demand for skills, few studies have been able to capture skills-level data in a way that is dynamic, cost-effective, and reflective of the different markets, industries, and occupations that characterize different geographic localities.
In this paper, we provide new evidence to characterize changes in the demand for skills associated with shifts in occupations for a sample of 10 of the 20 G20 countries, using information available from LinkedIn profiles as a new and unique source of dynamic labor market data on occupations and skills. A unique feature of LinkedIn’s data is the availability of granular measures of skill importance by country and occupation. This data allows us to examine how similar occupations may differ in their skills composition across dferent countries, and to measure the corresponding shifts in skill demand associated with changes in occupations for each locality. While the results are only representative of the subset of workers with LinkedIn profiles, they provide an important perspective on an increasing, and arguably very relevant, portion of the labor market.
A second aspect we tackle in this paper is the degree of transferability of workers across declining and emerging occupations. Understanding the transferability of skills is one of the most promising areas of analysis available through LinkedIn data. Throughout history, the development and expansion of new tasks and occupations have helped to outweigh job losses caused by waves of automation. However, this expansion can be slowed—and transitions be more costly–if recent graduates and/or workers who lost their jobs due to automation do not have the required skills to perform these new tasks or occupations. Possessing and promoting more transferable skills—i.e., those skills that are important to and shared across di erent occupations—may help individuals to better withstand labor market disruptions in a dynamic digital economy.
In this paper, we assess the transferability of workers employed in declining occupations to expanding portions of the economy as a first step in identifying the set of policies that may be needed to accelerate reallocation and economic adjustment, and to create more resilient learning and labor pathways for individuals. We leverage LinkedIn’s granular skills data to create a distance measure that estimates how close two occupations are based on the skills that workers in those two occupations share. Potential uses of this analysis are identifying those occupations into which workers in declining occupations can transfer, identifying alternative career paths for workers wishing to switch occupations, and identifying transitions with high potential for employment growth.