Technological Change and Work: What can history tell us about the future of work?
This is part four of a summary of my PhD thesis. You can find part one here, part two here, part three here, and more information about the thesis here.
In the last three posts, I explained the headline findings of my PhD thesis, which used historical case studies to analyze how innovation changes jobs and access to employment. In this post I will outline what this research suggests about the opportunities and risks of future technology.
To recap, there are competitions and tradeoffs during periods of technological transition. First, the number of jobs available is the result of a race between increasing productivity and demand for goods and services. Instances of technological change that rapidly raise output per worker can produce unemployment because demand increases more slowly, but technology that contributes to a virtuous cycle of growth and rising demand can increase employment opportunities.
Second, major technological changes or “macroinventions” can also expand the range of tasks required to make goods or provide services, and companies respond by separating these tasks into many new occupations. This stratifies work as the new jobs vary widely in pay, safety, and employment strain. Technology can allow for the creation of new jobs but also inequality on multiple dimensions.
While industrial technologies may seem far removed from today’s world of automation, the underlying process is the same: replacing human capacities with inanimate ones. In the Industrial Revolution, this meant replacing human strength and dexterity with metal and wooden implements driven by water or steam power. In the 21st century, it involves replacing human pattern recognition, calculation, and cognition with artificial intelligence. We should expect that the task range will increase again, and that work will become more specialized — indeed, the wide variety of new digital occupations shows that this process is already occurring. Distinctive task bundles and skill requirements can produce opportunity as well as inequality.
Third, after most companies have adopted similar macroinventions, their continuing search for higher productivity leads to a race between productivity and job quality. Many smaller “microinventions” that can reduce work strain or increase occupational safety do not fulfill that potential because managers demand higher effort levels or increase operating speeds. These operating practices partially or completely offset the job-improving opportunities of new technologies in order to increase output per worker.
What do these three big picture findings mean for the future of work? The first, most straightforward implication is that we need to be conscious of tradeoffs and of the disparate impacts of technology.
Textile mechanization rapidly reduced prices for consumers, but it also caused large-scale unemployment. As Jane Humphries and I have written, this job destruction had important effects on women’s employment and family earning patterns into the 20th century. Similar technological unemployment caused by automation could impact highly gender-segregated jobs in the future, such as trucking or logistics, with major implications for household incomes and wellbeing. We can celebrate technological breakthroughs that render difficult, tedious, and demanding jobs unnecessary, but we should not discard the workers who have been replaced by innovation.
Macroinventions may produce a wide variety of new jobs, but also increase inequality at work on multiple dimensions. Incomes, work intensity, and occupational risk can all vary substantially between jobs. Without strong labor bargaining power, managers will maintain high levels of job-related inequality and do little to improve the worst occupations. In the Industrial Revolution, only well-enforced regulations and powerful unions were able to limit working hours and substantially improve safety after decades of poor conditions and long hours. The flip side of job stratification is that without technological change, low-income countries may miss out on the opportunity to create new middle-class jobs. An equitable approach should encourage technological adoption while ensuring a floor of labor standards to improve pay and safety, and reduce the intensity of the worst jobs.
Labor intensification offers another potential route to productivity growth after companies have exhausted the potential of major inventions. Contemporary economies have many examples of this, especially in the gig economy with firms that use rising effort demands to increase output per worker. Labor standards and worker bargaining power are again important: the most unionized workers in the case studies I analyzed were those who were best able to resist labor intensification. Historical and contemporary workers in highly-competitive sectors (like textiles, which are traded internationally) are particularly susceptible to work intensification.
Innovation has huge potential to improve human wellbeing by easing our labor, improving quality and access to health care and other personal services, and producing new and better goods with fewer hands and less energy. However, decisions about the technology that companies choose to adopt and how they use it have tradeoffs. In the past, managers who focused on profits and productivity disregarded workers and their wellbeing. A future of good work that embraces the opportunities of technology needs to strike a balance between the accepting the benefits of innovation for workers, consumers, and companies, giving bargaining power to employees, and establishing high minimum standards for working conditions.
