While the furor around robots taking our jobs has largely died down in recent years (not least due to the lack of any real evidence that it’s happening), it remains inevitable that the introduction of new technologies will cause disruption in the labor market.

Research from the Kellogg School explores historical periods of technology-driven disruption to see if there are any patterns around the kinds of workers that get disrupted, and indeed how that disruption affected their current and future income.

Exposure to risk

The researchers developed an approach to gauge the exposure of workers to new technology. They do this by examining the tasks associated with occupations and determining whether any similarities exist between those tasks and the descriptions provided in new patents.

The analysis found that from 1850 to 1970, manual laborers were most exposed to new technologies. From the 1970s onwards, jobs where routine cognitive tasks were regularly performed also saw a significant increase in exposure as technologies began to do more thinking as well as heavy lifting.

Interestingly, this process meant that those workers with the highest salaries saw the biggest slowdown in their income, despite having the most advanced skills.

Unequal distribution

As we might expect, there are winners and losers from any new technology, and the distribution in terms of productivity and standards of living was far from evenly spread, even though society in general improved with each new technology.

Across four categories of jobs, there were some noticeable differences. For instance, for those in manual occupations, there was particular exposure to technological change. While that is somewhat expected, the study also found that those in more cognitive occupations were not immune either, especially if they required repetitive tasks.  Indeed, by the 2000s, the risk for those in this group was on a par with those without a college degree.

This was found to have a noticeable impact on employment prospects, as the researchers identified a clear link between exposure to technology and employment rates. What’s more, despite wage data only being available from the 1980s, there was also a link between exposure and lower income.

This risk was found to be particularly pronounced among older workers, who suffered slower wage growth than their younger peers given the same level of technology exposure. The researchers believe this might be due to the lower investment made by younger workers in skills that had become obsolete, while also having more time to invest in new skills.

Most to lose

Interestingly, seniority offered no protection. Indeed, those with the highest incomes suffered the most from technological exposure. The wages of these workers slowed at twice the pace of the average worker in the same occupation at the same level of exposure.

This was particularly so in roles where a lot of experience was required and a deep investment in particular skills was made. When those skills were automated, the impact was severe. This strongly suggests that it’s not as simple as technology automating jobs, and it’s more likely that technology changes the skills that are required instead as the nature of the job changes.

This results in people who might have spent considerable time, money, and energy learning and mastering a craft that is no longer relevant. In the worst case, they could be laid off, or in the best case, their wages could stagnate as a result of their skills no longer being in such high demand.

Low exposure

It has often been argued that the rise of automation will free up humans to do what humans do best, which tends to revolve around softer skills. To an extent, this was evident throughout past technological disruptions, where jobs in the interpersonal category had relatively low exposure to disruption. In essence, technology has always struggled to replicate human-to-human interaction.

This was especially so among those with high interpersonal skills, for whom incomes remained strong even when they had high exposure to technology. This seems to be the case today, with the infamous analysis from Frey and Osborne also suggesting that professions like nurses would be little impacted by the wave of automation that was set to wash over us.

The researchers also found that technology wasn’t negative in a uniform manner. When they assessed patents across various industries, they found no real overlap with occupational tasks. In other words, they believe that the impact on workers’ incomes was probably because the new technologies had helped them become more productive.

This helped some people get ahead, while those whose skills were no longer valued fell behind, at least while they adapted to the new situation.

The lessons seem to be twofold from the research. Firstly, being willing and able to learn and adapt almost continuously will be a crucial skill to have. The researchers point to MOOCs and other forms of online learning that are inexpensive ways of doing that, but they also urge policymakers to consider subsidizing lifelong learning more than they currently do.

The second is that soft skills remain a key differentiator between man and machine, so it’s important that these skills are given the same emphasis as more technical skills, which often seem to be the focus of industry and policy attention today. It’s often said that if we don’t learn from history then we’re doomed to repeat it, and so it’s important that we learn from previous technological transformations to avoid the disruption they caused.