Are we ready for the impact of automation?
By Malminderjit Singh
While automation may change the way we work in the future, it is not going to eradicate jobs, says Malminderjit Singh, Editor of THink: The HEAD Foundation Digest.
“Robots will take over our jobs!” some of the more sensational headlines these days scream. Other reports argue that automation will make most of the jobs we know and do today, redundant. Let us examine the probability of this prophecy being realised and to what extent.
The key drivers of a potential rapid displacement of jobs are automation, disruptive innovation, the skills gap and globalisation. Much of the intensity of job displacement in the future hinges on the variation of these factors and how they interact with one another. But there are other underlying factors within these domains too that may have a louder say, than some may imagine, in the fate of work of the future.
Will robots take over our jobs?
Automation has certainly been one of the main reasons why many doubt the longevity of current jobs. The lure of increasing factor input productivity by automating a job function fully, or part thereof, is one that employers may find attractive against the backdrop of rising wages. It may be easier to think of more labour-intensive job functions to fall prey to the jaws of automation, but recent trends have shown otherwise. Even white-collar professions such as lawyers and doctors now have part of their work automated and performed by artificial intelligence and deep learning software systems.
This shows that machines these days do not just do manual tasks but ones built on cognitive skills too. Therefore, it is not a matter of automation replacing white or blue-collar jobs or manual ones, but how routine a task may be. The greater the replication of the task, the easier for the machine to internalise, master and be productive in it and, eventually, carry it out more precisely.
So, if you are a radiologist, chances are that in the future you may have a machine taking over your job. But if you are a radiologist who also takes on other tasks in the hospital, it is rather unlikely that a machine will be able to displace you since your work is not routine.
According to a study done by Oxford University-based scholars Carl Benedikt Frey and Michael Osborne, future computering could put as many as 47 percent of US jobs at risk. They reported in ‘The Future of Employment: How susceptible are jobs to computerisation?’ that the industries most vulnerable to automation are transport and logistics as well as office support, while even some service-level jobs may be at risk.
What does this mean for the future of the job market going forward? For one, there may be a decline in the demand for specialised mid-level jobs. The more specialised skills that are routine, like radiology for instance, can easily be replicated by machines and, hence, replaced. Learning a specialised skill or knowledge in the future may be less appealing and necessary.
There may be a decline in the demand for specialised mid-level jobs.
Certainly, with the spread of artificial intelligence and the information economy, the need to develop specialised knowledge may be less necessary than to learn its application and management. The impact of this on the job market, and even the higher education sector, is a possible swing towards more general qualifications.
The exception to this rule is high-level specialised jobs, such as architecture, where a reasonable degree of discretion is required therefore making the job less routine. But this is where automation could have a further impact; not just on the workforce, but also on society. As it seems that the less routine a task the safer it is from being displaced, there are two distinct groups that should benefit from this conclusion: high-level skilled jobs and low-level unskilled labour.
This makes the mid-skills jobs the most vulnerable to be replaced by machines. This could lead to greater wage inequality in the future and make the income divisions in society more pronounced, which in itself leads to several accompanying social challenges. Younger mid-level workers may have the option to upskill and move up the value chain if opportunities exist. Older workers though, may be less able and willing to do so and hence, may either be out of work or force to take low-skilled jobs. Either way, underemployment and unemployment may be more pronounced due to this phenomenon.
But the real question is if these predictions will hold true. The advent of automation and its perceived threat to jobs is not new. As early as the Industrial Revolution, the threat of automation became the scorn of textile workers who believed their livelihoods were at stake.
Throughout the course of history, experts warned of the impending threat of technology and automation on jobs – eminent economist John Maynard Keynes coined the term “technological unemployment” in the 1930s; in the 1960’s US President John F Kennedy in a speech at the United Chemical Workers Convention cautioned that the biggest domestic challenge was to maintain full employment and not allow automation to cause unemployment; a group of Nobel prize winners in 1964 sent US President Lyndon Johnson a note to warn him of the impending danger of an automation revolution, which they argued would alter the production process with less human capital input, dividing society into a skilled elite and unskilled underclass.
The reality is that, despite all of these bleak projections, technology has led to more job creation than unemployment. An explanation for this, as put forth by Massachusetts Institute of Technology economist David Autor, is that as automation is introduced to make certain tasks more productive, there is an increased demand to employ workers to carry out tasks around it that have not been automated.
The reality is that, despite all of these bleak projections, technology has led to more job creation than unemployment.According to a study published in McKinsey Quarterly in 2015, computerisation in most cases does not lead to the replacement of workers but automates portions of the tasks they perform. A special report by The Economist magazine on automation in June 2016 illustrated this point through the example of automated teller machines (ATMs). The introduction of ATMs in the US saw the average number of human tellers employed by banks fall from 20 percent per branch in 1988 to 13 percent in 2004. But this reduced the banks’ operational cost per branch allowing them to open more branches in response to customer demand.
As a result, the number of urban bank branches increased by 43 percent over the same period, thus increasing the number of jobs. So human tellers lost part of their previously routine jobs to machines but then took on new roles like customer service, which computers could not automate. Similarly, while driverless technology may put drivers out of jobs in the future, they will then have the opportunity to train and operate as safety drivers.
There is thus enough evidence to show that while automation may change the way we work in the future, it is not going to eradicate jobs. What it will do is replace the more routine tasks that we do, allowing us to take on more complex functions that cannot be automated. This will require us to learn new skills so that we can take on these new roles.
Are we skilled enough?
As we have seen, the ability to traverse the future economy will depend on how skilled we are to do so. Whether automation manifests itself as job creator or job robber will be determined by our capabilities and if we can take advantage of the opportunities it presents us.
One of the greatest challenges in equipping ourselves with such skills and capabilities is even identifying them in the first instance. Since many of the jobs of the future do not even exist as yet, it is rather difficult to be exact about what skill sets these may require without being presumptive. The term ‘21st century skills’ is a buzzword these days, but again, besides a few common core skills that everyone talks about like coding and data analysis, any indication of which skills will be prerequisite in the future are just speculative.
Another argument is that core skills remain constant regardless of whether it was the 21st century or decades before. Realistically, basic skills such as numeracy and literacy are essential tomorrow as they have been in the past and today. It is difficult to imagine someone taking on programming before attaining a considerable level of competency in mathematics.
But herein lies the problem. Travel to many parts of the developing world and you will find that the education systems there are woeful in even providing numeracy and literacy abilities that are competent globally. If they cannot even meet the needs of today’s basic skills, how will they be able to keep up with future more complex developments?
Such scenarios create skills gap in the developing world. As a result, we may see a reversing of the trend of previous years where jobs moved from the developed world to cheaper destinations on cost competitiveness. As jobs become more specialised in the future and so demand commensurate skills, we may see them returning to more advanced countries where education systems have been able to prepare a skilled and competent workforce.
For the workforce in the developing countries, a lack of skills may cost it dearly in hanging on to jobs. Since automation may replace many of the routine tasks, these economies are likely to struggle to find niche industries to be competitive in unless they can make substantial leaps in their skills agenda.
For the workforce in the developing countries, a lack of skills may cost it dearly in hanging on to jobs.But even in developed countries, the ability of governments and systems to recognise future trends and anticipate their implications varies. Singapore is one example of a country that has taken a proactive approach to bridging the skills gap through initiatives such as SkillsFuture. But even in countries where many of the trends of future jobs and skills originate from and are determined, gaps still exist. For instance, in the United States, the number of skilled workers needed is growing and there will be a projected shortage of 3.5 million workers by 2025.
The challenge of having a skills gap globally is that firms may not be able to find enough workers to fill high-level jobs that they create. Thus, while automation may replace routine jobs, businesses may not be able to use the cost savings to create more jobs higher up the value chain because they simply do not have an adequate pool of talent to rely on.
This could lead to two possible outcomes – one, companies have less incentive to automate as they do not think they will successfully be able to move up the value chain without the necessary talent pool or second, companies automate and successfully replace routine jobs but then, due to a shortage of talent, are not able to scale up and create new and skilled employment. Which way the situation swings depends on a number of factors including the level of business costs, type of industries, factor productivity rates, the visibility of skills gaps and the mobility of labour across industries and geographies.
Uberisation of jobs
The disruptive innovation of business models and industries is a factor that is rapidly influencing the nature of future work. Today, the term Uberisation, named after ride sharing company Uber, has actually become an adjective to describe a workforce trend that sees the conversion of existing jobs and services into discrete tasks that can be requested on-demand.
The development of this ‘gig economy’ has quickly transcended the workforce as startups across industries and countries transform spare capacity into productive resources. Besides the transport industry, Airbnb has made anyone with a vacant home into a landlord. In this region, online grocery shopping company honestbee has turned those not in the workforce into productive shoppers. Same with local logistics company Ninja Van, which taps on the lunch hours to allow workers of any companies to double up as delivery personnel.
The gig economy does have its pros and cons. On one hand, it creates a useful alternative for those of out of job and not actively looking for work. The disruptive economy could thus be a salvation for those who lose jobs to automation in the future and need an urgent stream to sustain their livelihood. However, the attraction of flexible work arrangements could also lure many from the mainstream workforce to the gig economy.
For one, this could hollow the workforce and worsen the skills gap. A low-skilled worker, for instance, who was displaced by automation may find it easier to do gig jobs delivering parcels, driving for Uber or doing concierge services rather than upskilling and moving up the value chain. This is a net loss to society as it stagnates skills development and worsens the ability and incentives of firms to create new higher skilled jobs. In addition, the gig economy also does not provide social insurance or a social safety net for employees through medical coverage, pension payments and retirement savings plans, increasing the fiscal burden on the state for this segment of society in the future.
Barriers to jobs
Disruptive companies thrive on scale though as they strive to be the large multinationals of the future. And that is where globalisation comes into play as a factor. This essay has already discussed how the mobility of labour could help determine the impact of the skills gap on the willingness of companies to automate. But there is more that globalisation, or the lack of it, could impact the future of jobs.
The events of the last year have shown that the trend of global liberalisation of the last four to five decades is now reversing. The increase in protectionism or the inclination to look inwards is adversely impacting the global trade and economic regime.
Dipping into economic theory, David Ricardo’s theory of comparative advantage and specialisation, which underpins the modern convention of free trade that we had been used to, is based on a core principle – that free trade will allow countries to specialise in the production of goods and services that they have a comparative advantage in producing. Should the movement of goods and services become less free, will countries still find it as attractive to specialise? In such a scenario, will automation and technology adoption continue at a high pace?
As such, the decrease in globalisation could slow down the loss of jobs to automation.
Humans need apply
The jobs of the future could not be under as much threat as some may fear. It seems likely that automation will replace routine jobs, particularly at the lower and middle-skilled levels. But there is also an opportunity for automation to lead to more job creation higher up the value chain, which requires society to upskill.
But if current trends of disruptive economies and anti-globalisation pick up steam, automation may be less attractive to companies and jobs in the future may not be displaced by machines as rapidly. Humans still have a substantial role to play in the future workforce and any concerns otherwise may be exaggerated and misplaced.
Malminderjit Singh is Editor of THink: The HEAD Foundation Digest. He is an Editorial, Communications and Government Affairs consultant to a number of private sector and non-profit organisations. He is also a speech writer and adjunct academic.
This article was previously published in THink: The HEAD Foundation Digest and reproduced with permission. The HEAD Foundation is a Singapore-based think tank that is focused on the research, policy influence, and effective implementation of education for development in Asia.