When it comes to automation, the topic can often be framed in a way that accentuates its seeming inevitability. The inevitable march of technology can make things appear as though all automation is positive automation. Research from Dartmouth and MIT suggests that this isn’t always the case, and that excessive speed of automation may not always be a good thing.
The researchers explain that while investments in automation usually result in higher productivity in firms, and therefore often more employment, it can be harmful to those who are displaced, especially if they have few alternative options. As a result, the researchers believe that the speed of automation should be halved.
Slowing down
“Firms do not necessarily take into account the consequences that automation has for their workers. Instead, they tend to focus on the value that automation will bring to the firm and its shareholders,” the researchers explain. “Automation can benefit society as a whole. But it also comes at a cost in the short run. It displaces workers who can be financially vulnerable.”
As a result, the authors believe that automation should be taxed in order to slow it down sufficiently that workers can successfully retrain and then transition into new jobs. Indeed, in an ideal world, the revenue raised by the taxation would be specifically used to finance retraining programs or otherwise be used to compensate displaced workers.
They are at pains to point out that this shouldn’t be viewed as a permanent measure, although it would be sufficiently permanent to last a few decades, after which the researchers hope that any displaced workers will have transitioned into new forms of work.
“Automation should only be taxed while existing generations reallocate,” they explain. “There is no reason to tax automation in the long-run, if new cohorts can get the appropriate training and choose occupations that are less at risk of automation.”
Optimal policy
While the likes of Bill Gates have argued for taxing the implementation of robots and similar technologies, the researchers believe that there remain significant gaps in the literature when deciding on the best policy to help combat any negative externalities associated with automation, such as the displacement of workers.
They aimed to plug that gap and better understand how, or even if, governments should intervene in the automation process to better support those who are negatively affected, while, of course, continuing to support the kind of innovation underpinned by automation.
They identify two main sources of friction that workers tend to face when they’re confronted by automation that disrupts and displaces them and their work. The first of these involves the transition they have to make into new roles, which might involve long periods of either retraining or unemployment.
The second friction then revolves around the difficulties workers have in terms of borrowing against any future income as that future income is inevitably more uncertain as a result of the impact automation has had on their livelihood. Therefore, from a policy perspective, the researchers argue that automation can often be both inefficient and excessive.
Pace of change
They attempted to understand what is a more desirable pace of change to allow innovation to occur but also for workers to adapt to any disruption to their livelihoods. They developed a quantitative model that aims to replicate the fundamental dynamics involved in automation and the displacement of labor seen since the dawn of industrial robotics in the 1970s.
This was a period that was marked by a prolonged period of decline in employment in occupations typified by routine tasks. Such occupations included file clerks, machine operators, and welders.
“Our research shows theoretically and quantitatively that there is a rationale for slowing the speed of automation,” the researchers explain. “The government can deliver broad benefits to the economy by taxing automation and reducing its pace by half. As a result, workers are better off, and this improvement in welfare is comparable to a permanent increase in consumption of 4%. ”
There has obviously been a huge amount written on the future of work and both how many people might be disrupted by technology and how this risk could be mitigated. While the history of technology does suggest that adaptation does occur in time, it also suggests that in the short-term there is often considerable disruption, with this likely to lead to periods of political instability as those individuals and communities that are disrupted resort to populist politicians promising easy answers.
With nearly 13 million people employed in manufacturing in the U.S. alone, it’s a matter of considerable importance, but automation is now something that promises to affect sectors far beyond manufacturing, such as customer service and transportation.
Suffice to say, this is not the only suggestion on the table, and other researchers have advocated slightly less “nuclear” options and promoted things like mobility vouchers to help people move to areas where employment prospects are greater. Such suggestions remain on the table, however, and it very much remains to be seen whether policymakers are able to act sufficiently swiftly to provide the support displaced workers need. As a result, slowing down automation might actually be the most sensible choice of all.
