Why, for me, the long-run impact of automation on unemployment has been recently overstated

Really sorry not to have written for so long – exams and general schoolwork really got in my way. For the summer, though, I think I’ll be back to regular writing! 🙂

We’ve seen in recent years a sharp rise in attention to the potential automation of a wide array of different jobs and its long-term effects on the global economy. Such concern has in part led to the touting of different hypothetical policies that could help assuage the problem, such as the famed Universal Basic Income (UBI).  Some have welcomed the idea – after all, you can see how not having to do any work may seem attractive. Others despair, fearing of widespread unemployment and subsequent social unrest compounded by economic uncertainty. However, for me, whether you’re concerned or jubilant at the idea (or somewhere in the middle), there may be reason yet to tone down your forecasts of a robot-led revolution of the labour market. This is for two principal reasons:

  1. We must never underestimate the capacity of humankind to continue innovating, while respecting the notion of Knightian uncertainty. It’s important to note that machines have been slowly gaining the ability to do more and more of the jobs we currently do since the first Industrial revolution in Britain, in the late 18th century. Even then, less than 1 in 20 of the UK’s labour force is unemployed as of May 2017. That’s why, here, I’m making a distinction: it is very possible that robots may take many of the current jobs we do, but I definitely don’t think it’s as possible that robots may take as many of the jobs we will do in the future. We can see from technology’s current progress that it may not be so long before the 9-5 office jobs of our time begin to stop becoming available to humans, however we will never be able to see exactly how many and which jobs may be created in future. We live in a world dominated by Knightian uncertainty, essentially the idea that there are infinitely many outcomes that can arise from any given situation, all of which we cannot predict. Furthermore, so many new jobs have come into existence in the last 10 years that people are making top 10’s of them. Taking the uncertain future and the unpredictable past hand-in-hand, we can see that the possibility of human ingenuity aided by automation driving job creation is a very real one, and this does provide some room for optimism. Machines such as the smartphone have driven the growth of companies like Facebook and Uber, so who’s to say we won’t see another groundbreaking development that could create scores of jobs across the globe? Of course, basing a prediction off the past can go catastrophically due to the very uncertainty outlined above. However, I can say with a reasonable degree of confidence that with the innovation seen even today, it would be unlikely that automation would take jobs from us without giving at least some of them back.
  2. Polanyi’s Paradox. This essentially means that “we know more than we can tell.” Some things we do, like writing a poem, we can understand but not explain to others the intricacies of how we actually do it. Many products intuitively require a degree of creativity to be produced, something which to this day computers or robots have not been able to fully replicate. Take the iPhone – a combination of machines could very well be programmed to assemble the iPhone from its constituent parts, but it’d be a big stretch to suggest that Steve Jobs’ initial vision of a multi-purpose device operated by touch could have come even close to being replicated by the robots of today. So while the more manual parts of production may continue to be automated, companies operating in sectors that rely on creative spark and vision would still need to employ creative thinkers, those with ideas out of the box that can revolutionise and mold entire industries. More formally, of course, what I’m talking about is known as research and design (R&D for short). As companies see a cut in costs of production arising from robots automating manual production processes, it’s indeed a possibility that they may invest more into R&D, generating significant employment. Of course, some corporations operating in monopolistic or oligopolistic markets may not, in fact do this. It’s also clear that in the short term there may be significant unemployment as people become educated and re-educated in areas where they have a realistic chance of employment. The threat of machine learning also calls this argument into question, however despite the evidence of self-driving cars, we haven’t seen much in the way of evidence suggesting robots could, in fact, disprove Polanyi’s paradox, and so for now, it’s fair to say that the Polanyi’s paradox argument does hold some weight.

The arguments now set out point to a dramatic shift in composition of the labour market, but not to a dramatic shift in unemployment numbers themselves. From history, we’ve seen that in the long-run, automation actually generates more jobs than it displaces, but in my opinion, the major concern for both developed and developing nations should be how to highly educate vast numbers of people in line with the needs of a dynamic labour market. It’s easy to see people needing to be more educated to be able to carry out jobs that machines cannot yet do, and requiring different skillsets to what the labour market generally requires now.

For this reason, it’s so, so important that more people are encouraged to seek the high-skilled professions that many feel are still only for the elite. To me, it’s not a stretch to say that the future of the global economy depends on it.


By Shrey Srivastava

A finance and economics enthusiast, and someone who wants to share his views with the world.


  1. A well-written and informed article Shrey. Keep it up! I agree with your fundamental argument and the conclusion you have reached. You raise a key point – ensuring that occupational mobility is achieved in the labour market will help if technology/machinery threatens to displace workers in current sectors. I would encourage you to look into David Ricardo’s view on technology and machinery – I would like to understand where your opinion lies in regards to his views.


  2. Good Read. Disruptive changes at slow or fast speed have been part of development of human kind! We will manage to replace parts of mundane tasks through automation- small scale and large scale. But who will conceptualise the automatiion – how and why? Startrek had automation in a number of areas,and that is the way, I think, we will go, in my opinion. So I agree with you and this scare that automation will cover everything is only a fantasy. I also agree that skill set needed would be substantially high and unit a chip to increase human learning is invented, we have to continue to get educated..Good read overall!! Keep it up!!


    1. Thank you very much! I very much think that the automation will be conceptualised by a variety of different companies competing with one another in a variety of different markets. Cutting costs, ceteris paribus, increases profit margins and so to gain a competitive edge over other companies in the marketplace, I feel that companies may automate (I think we’re seeing this a bit in financial services already). Thank you very much for the comments and the praise! 🙂


  3. Wonderful article, Shrey! Do you entertain the possibility that computers and AI could be programmed to think like humans? In this case, perhaps there should be concern towards automation and the threat of AI, even in creative industries? Keep up the insightful writing!


    1. Thank you very much 🙂 Yes, I do – and I think that is the fundamental difference between the current wave of automation and those which preceded it. However, looking at the current state of technology in the area (for example, neural networks), it appears such technology is relatively far from becoming a reality. That said, I don’t have a crystal ball and so this article would definitely need to be heavily revised if such technology become commonplace.


Leave a comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: