At a time when workers around the world are concerned about how technology will affect their jobs, some recent research underscores the importance of on-the-job learning as a way to better understand the economics of new technologies.
Prasanna (Sonny) Tambe (pictured above) and John Horton conducted a retrospective survey of Adobe Flash workers to answer the question: What happens when a technology becomes obsolete?
The results, described in The Death of a Technical Skill: Evidence from the Demise of Adobe Flash, show that because technology workers are forward-looking and rapidly switch to new skills, market supply adjusts relatively rapidly to shifts in demand for a new technology. Tambe is Associate Professor of Operations, Information and Decisions at the Wharton School at the University of Pennsylvania, and a Digital Fellow at MIT IDE; Horton is an Assistant Professor at MIT Sloan. As the paper notes: “The waxing and waning of the demand for speciﬁc human capital is commonplace in a dynamic economy, and the waxing and waning of IT or STEM skills is particularly important.”
At a recent IDE seminar, Tambe described the research and the implications for the spread of new technologies that require complementary workforce skills. Prior work (e.g. Wachter et al 2009, Couch and Placzek 2010) studied the impact of events such as mass layoffs and plant closings on displaced workers and found earnings losses on the order of 15% of higher. In contrast, Horton and Tambe examine a particular, high-skilled technical context where on-the-job learning is both common and rapid. They found that wage losses for workers who continued to use the technology were relatively muted because supply adjusted so quickly.
Support Gone in a ‘Flash’
The specific context was that of Adobe Flash. In 2010, Steve Jobs announced that Apple would no longer support Adobe Flash—a popular set of tools for creating Internet applications, Tambe explained. In the years following, the use of Flash declined precipitously. Although Adobe abandoned the product—leaving those skills with no perceived future—most specialists acquired new skills, primarily through on-the-job training.
“Using data from an online labor market, we show there was no detectable reduction in Flash hourly wages, or even in the number of applicants per Flash job,” he said. In fact, “the reason wages stayed flat was that the negative demand shock for Flash quickly became a supply shock: Flash specialists transitioned away from Flash, and new market entrants were less likely to specialize in Flash.” Over time, they acquired new skills for the evolving market.
The implications are that while new technologies will always displace old ones, markets adjust quickly enough to keep wages steady or at least on a slow decline, according to Tambe. This means that the wages employers pay workers to use technology systems that are becoming obsolete will not necessarily fall much below the wages they pay workers to use more mainstream technologies. In fact, workers using these older technologies may even command relatively high wages for years after the technology begins to fall out of favor, as has been the case with programming languages such as COBOL.
As the paper concludes: “Our results are consistent with the labor market not being a signiﬁcant impediment to technology adoption: the ‘old’ skill did not become relatively cheaper, and the price of ‘new’ skills (at least to Flash workers) would, if anything fall, as supply increased and/or workers were willing to work at a discount to increase their human capital.”
Read on Medium, here.