Distinct skill sets contained in low and higher-income occupations could be one reason for the disappearance of middle-wage jobs in the U.S., according to a new study. | Iyad Rahwan
The U.S. workplace is divided into two main categories — high-wage occupations associated with social-cognitive skills such as complex problem solving, and low-wage occupations aligned with sensory-physical skills such as equipment maintenance, according to a new study published in the July 18 issue of Science Advances.
The study also concludes that transitioning from low-income to high-income occupations is difficult because the skills associated with each type of job are so different. This polarization of skills has notable consequences for wages and provides important insights into the controversial "hollowing" of the American middle class, brought about by the relative decrease in middle-wage jobs in the American workplace, authors of the analysis said.
Iyad Rahwan, an author of the paper from Massachusetts Institute of Technology's Media Lab, was surprised to find that workplace skills are highly polarized. "There are two dense clusters of skills … [and] the transition between these two clusters is not smooth. Rather, there is only a narrow bridge of generic skills connecting the two clusters," he said.
The transition from low to high wage jobs could be difficult for some, because it would require low-wage workers to take time off to go to school to learn new analytical and management skills. Therefore, low-income workers are "stuck" relying on the low-wage skill set, the authors said.
To date, there have been at least two major culprits traditionally associated for the hollowing of the middle class. Some experts and policymakers suggest technical changes such as the automation of routine work are responsible for shrinking the pool of U.S. middle-wage jobs, while others fault globalization and the removal of U.S. jobs to other countries.
To study occupational polarization within the American labor, economists have tended to group jobs into two main categories: white collar and blue-collar. White-collar workers are well-paid and are inclined to rely on higher educational qualifications, whereas blue-collar workers are largely paid less than white collar workers and perform more physically laborious jobs.
These classifications, however, do not capture the complex connections between skills within each job, Rahwan said, and questions remain about the detailed dynamics of skill-related trends in the U.S. workplace and how workplace skills drive economic inequality more broadly.
Learn more about U.S skill polarization across occupations and cities, with the Skillscape site created by Iyad Rahwan and colleagues at the MIT Media Lab.
Rahwan and his colleagues used a network-based approach to create a new methodology that analyzes job-skill interrelatedness in the labor market to identify distinct clusters of skills. "Our approach, in contrast [to traditional methods], is completely bottom-up," said Rahwan. "We look at the skills themselves, and ask, which skills occur frequently together in the same job?" This allowed the authors to see what kind of structure emerged directly from the data.
The study relied on a database of skills called O*NET, which the Department of Labor collects. The researchers' dataset consisted of 161 workplace skills, including skills ranging from persuasion and to finger dexterity, related to 672 U.S. occupations to map skill relatedness. The authors found that the job-skill network was composed of two distinct clusters — social-cognitive skills and sensory-physical skills — with high-income workers tending to use more social-cognitive skills and low-income workers tending to use more sensory-physical skills.
The researchers suggested that the connections between skills may shape the trajectory of how workers change from one occupation to another. Individual workers, for example, take advantage of the similarity between skills that they already have and skills required by another potential job to facilitate their career mobility. Therefore, workers are more likely to transition to occupations requiring skills in the same skill cluster, the authors said.
Polarization in the job-skill network can thus hinder the career mobility of individuals, with low-income workers stuck relying on the low-wage skill set, they said. Since occupational transitions usually require changes in workers' skills, the polarized network of skills revealed in this study should constrain mobility between low-wage sensory-physical occupations and high-wage sociocognitive occupations, Rahwan and colleagues explained.
"This should lead to disproportionately high employment below a certain cognitive threshold, rather than a smooth distribution of employment across the range of cognitive values," the researchers write.
Future studies are still necessary to track changes, such as technological advancement, in the workplace industry over time. For example, accountants today can use computers for detailed calculations whereas accountants 20 years ago were required to do calculations by hand. "The nature of work is constantly changing," Rahwan said. "Tracking these kinds of changes over time is crucial to understand how technological and economic forces shape the future of work."