Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs Flashcards
(Article: Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs)
What is the content of paper (Introduction)?
- People fear that robots, algorithms, and AI may destroy jobs on an unprecedented scale. Robots are said to be taking over many mechanical processes and effectively expelling factory workers from manufacturing. In the past old jobs were replaced but also new ones were created –> candlemakers by electricians, saddle-makers by engine drivers and so on.
- Mainstream economists posit an automatic adjustment process: if new technology makes it cheaper to replace labor by machines, the prices of goods will tend to fall. Lower prices mean spending power is released to buy other goods and services. New businesses are set up to capitalize on these demand shifts, and jobs are created.
Can job creation continue to offset job destruction indefinitely? Today’s worrier’s insist that ‘this time it’s different’. - One reason for that is that the process of technological change appears to be faster than ever before
- Innovation no longer simply threatens the production-line jobs of unskilled workers, but also those of the middle classes. –> AI will soon supersede the skills of many lawyers, doctors, and architects
(Article: Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs)
What aboutt The industrial revolution and job destruction?
- Carl Frey is an economist, and his new book focuses on a comparison between the industrial revolution of the late eighteenth century and the computer revolution of recent decades.
- Frey’s view is that technological change in this period was primarily ‘worker-replacing’, keeping wages down, and thus two generations of workers gained little. It was only later that working-class living standard and western Europe employed technologies that were ‘worker-enabling’. By this he means technologies that “augment the capabilities of some workers and enables them to perform new functions, making them more productive in a way that also increases their wages”.
- Frey thinks that now, however, technology may have switched to being ‘worker-replacing’ once more and jobs and wages are threatened again. Frey points out the similarities between today and the early nineteenth century. Real wages are said to be stagnating, there is ‘hollowing out’ of middle-income jobs, and the rewards of new technology are going to super-rich entrepreneurs and executives.
(Article: Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs)
Which jobs will go – and how many?
- Frey and Osborne found out that on the basis of linking experts’ opinions and the objective manual tasks each job has involved, that 47% of American jobs are at high risk of being replaced by new technologies. This work has received an astonishing amount of attention from academics with Obama even using their estimates for his estimations.
- Of course, this has led to a lot of alternative approaches being tried. For example, some argue that one needs to look at the tasks which people actually perform, rather than those constituting a generalized description of the occupation of the kind. Therefore, individual-level data were drawn and led to a dramatic reduction in the proportion of jobs predicted to be at high risk of automation –> only 9% are at high risk.
- Another researcher (Roger Bootle) thinks robots are over-hyped, saying half of their industrial applications are still in automotive manufacturing, where they have been operating since the 1960s. Furthermore, many types of technological advances are likely to run into regulatory problems. –> there have been long-running rail strikes as a consequence of proposals to reduce or abolish the role of guards on trains.
(Article: Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs)
Why are Labor markets more complicated, yet more reassuring than you think?
- One problem with Frey’s approach is that there is very little evidence of widespread job loss as a result of technological innovation. The reasons for some job losses are not due to automation but because of competition from low-cost competitors in Chine etc.
- Frey’s emphasis on technological change tends to exaggerate its significance for employment. The pessimistic conjecture of Frey about the drastic loss of employment due to automation is indeed not as simple as he thought.
- The labor market is much more sophisticated than that: New technologies create new jobs and new sub-jobs (Content creator etc.). Furthermore, millions of people switch jobs in order to get a better salary etc. This would leave the old job (if not replaced) labeled as ‘lost’ but that is certainly not due to automation. Moreover, old-fashioned/not tech advanced jobs are on the rise: fitness trainers, hairdressers, tattooists, or chefs. These are all ‘new’ jobs, yet would not have been unfamiliar in the days of the Roman Empire, and it is difficult to see that demand is likely to fall off or that they are likely to be replaced by AI/automation any time soon.
(Article: Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs)
What us the policy?
- One widely discussed proposal is to tax robots. The journalist Bootle wrote that we should tax robots to offset the tax loss as people become unemployed or face lower wages.
- But a tax on botos would be far more difficult to impose than people seem to imagine. The first problem is how you define robots. If you think of robots as a C3PO you could just tax this discrete unit. But what if they are a smart production line or an aircraft’s autopilot system? Would the tax be imposed on hardware or the software that powers it?
- If governments attempted to impose a robot or AI tax there would be huge legal and tax avoidance and evasion as well as a substantial diversion of investment into ‘almost robots/AI’. Bootle sees this as undesirable, distorting choices and slowing productivity growth. He points out that any country imposing a robot tax in isolation would discourage inward capital investment. And if job protection is really the aim, he points out it might make more sense to reduce/abolish existing employment taxes.
(Article: Shackleton, J.R. (2020). Worrying about automation and jobs, Economic Affairs)
What us the conclusion?
These two contrasting books may reassure worriers that the collapse of work is not yet imminent. They do suggest, however, that the pace of change is accelerating and today’s labor market entrants have to be prepared to be flexible and enterprising.
There is no obvious way as yet in which government intervention can improve on the workings of the free market to maintain employment, and a case has not yet been made that technology is leading to such large irreversible shifts in income and wealth as to justify new taxes and major redistributive measures.