In the mid-20th century, high-paying factory jobs were an engine of American economic growth and created a strong middle class. But during the late 20th and early 21st centuries, middle- and low-income workers’ wages stagnated.
New research shows that job design—the choices employers make about the actual tasks workers are assigned—could provide a route to higher-paying jobs for the future.
“Explanations for rising inequality and stagnating wages usually stop at the factory gates and at the superstore’s sliding doors,” the study’s authors write. “But wages are also affected by the choices that employers and managers make about the tasks workers do.”
The researchers found that when employers raise task complexity and autonomy in a given job, new hires’ pay gets a boost and grows faster in subsequent years. Earnings increased between 6 and 10 percent when jobs shifted from no complex or high-autonomy tasks to fully complex or high-autonomy tasks.
Complex tasks require either tacit knowledge that can be learned on the job or technical knowledge rooted in formal training, or both. For example, banks can give tellers responsibility not only for clerical work but also for loan assessments.
Autonomous tasks are subject to relatively little managerial oversight and give workers some control over how they do the work. For example, call centers can reduce the use of scripts and give staff greater discretion in conducting conversations with customers.
Increasing complexity or autonomy for frontline workers can increase productivity and decrease the cost of supervision, which in turn gives employers more latitude to increase workers’ pay. This study, by Dylan Nelson at the University of Illinois, Nathan Wilmers at MIT Sloan, and Letian Zhang at Harvard Business School, is one of the first to show how this works.
New data for measuring task complexity and autonomy
Case studies have shown that “high road” employers can maintain profitability by raising frontline workers’ pay and autonomy. But quantitative evidence about the effect of this job design on pay is limited, partly because the data needed to answer the question are so challenging to assemble.
The researchers linked wage records from the unemployment insurance system to hundreds of millions of digitized job posts and multiple surveys. This created a unique dataset that allowed them to measure changes in tasks for the same employer and occupation.
One concern is that increasing the autonomy or complexity of a job might simply result in workers with more education and experience being hired into the job. This is important because if job design only resulted in people with higher credentials filling previously accessible jobs, it could raise pay in a given job but increase inequality by reducing opportunities for lower-wage workers.
The researchers found this was not the case. In this study, relatively little of the earnings boost was due to a changing mix of workers. Instead, improved work organization increased workers’ earnings even without workers getting more schooling or experience.
“Our findings are relevant both for managers seeking to create higher-paying jobs for frontline workers and policymakers focused on incentivizing that creation,” the researchers wrote. “We show that organizational workflow changes, voluntarily implemented by employers, have the scope to increase workers’ pay.”
Nelson, Wilmers, and Zhang’s research was supported by the Upjohn Institute’s Early Career Research Awards program. The program began in 2007 and has supported more than 200 scholars, furthering the institute’s mission of informing employment policy and practice with timely research.