1. The Future –
    Utopia or Dystopia?

    Employment and Technology
    in the 21st Century

    Lawrence Burns, Investment Manager. Third Quarter 2017
  2. Progress in artificial intelligence (AI) has been phenomenal. Deep learning, in particular, has allowed machines to learn through data and experience rather than on the back of pre-programed models. It is enabling breakthroughs in natural language recognition and computer vision in a wide range of applications. Given these and other advances, futurists, journalists and academics are increasingly asking themselves what roles, if any, will be left for humans.
  3. According to leading technology innovator Sebastian Thrun, “we are just seeing the tip of the iceberg”. While no doubt a great technical achievement, the recent progress in autonomous vehicles, for example, does beg the question of what will become of taxi drivers, truck drivers and even pilots when it is cheaper, safer and more efficient for computers to undertake these roles. A widely-quoted study by Oxford academics Frey and Osbourne has even claimed that as many as 47% of today’s occupations are at risk of being automated in the next 20 years. Will there be an employment crisis? An explosion of inequality? And, will the upshot of this be a resistance to technological progression and, therefore, the blue sky outcomes we are relying on?

     

    Image: Remains of airliners at Mojave storage facility.
    © In Pictures Ltd./Corbis Historical/Getty Images.

  4. Before panicking too much, it’s worth remembering that the fear that machines would make human labour obsolete is as old as machines themselves.

    As far back as the industrial revolution, Luddites smashed weaving looms, fearful they would take their jobs. And it has been no less emotive throughout the 20th century.

    In the 1930s, British economist John Maynard Keynes coined the term ‘technological unemployment’, predicting that the working week would shrink to 15 hours as the need for human labour diminished. President John F. Kennedy went as far as declaring in 1962: “I regard it as the major domestic challenge, really, of the ‘60s, to maintain full employment at a time when, automation, of course, is replacing men”.

    At the dawn of the computer revolution in 1982, and in the face of mounting historical precedence, the view that the rise in technology would lead to underemployment stubbornly persisted. Nobel Prize-winning economist Wassilly Leontief was adamant: “Past experience cannot serve as a reliable guide for the future of technological change … it is a simple fact that fewer people will be needed”.

    Technology and employment present us with a fascinating paradox. While the past few hundred years have seen countless innovations from the spinning jenny to the computer, all of which are demonstrably labour saving, it simply is not the case that huge technological leaps and increases in automation have either rendered human labour obsolete or led to an increase in long-run unemployment. Quite the opposite, in fact. The proportion of the population employed has increased materially and done so in almost every decade for the past 125 years. In the US, for example, 52% of the population was employed in 1890, today the figure is 65%.

    A woman weaves cotton on an industrial loom.
    © Hulton-Deutsch Collection/Corbis Historical/Getty Images.
  5. Understanding the Historical Paradox of Automation

    The interplay between automation and employment is complex, influenced by factors such as how automation affects end demand and whether or not it substitutes (rare) or is complementary to (more common) human labour. On first glance, it might seem a given that automating a task should reduce employment on the grounds that there is less work left for humans to do, but the reality is that because automation often lowers the costs of producing goods or services, it can stimulate demand. And more demand can actually mean greater employment. 

    Returning to the earlier example, in the industrial revolution, the emergence of mechanical weaving looms reduced the labour required to produce a yard of coarse cloth by a staggering 98% and lowered the end costs to consumers substantially. Demand grew exponentially. Far from a shrinking profession, between 1830 and 1900, the employment of textile workers in America quadrupled to meet the increased demand.

    In the 1970s, ATMs were introduced into American bank branches, seemingly threatening the jobs of the 250,000 bank tellers. Despite the mass job losses predicted, the number of bank tellers has doubled to 550,000 today, with 100,000 added in the first ten years of the 21st century. Instead of eliminating bank tellers, the ATMs lowered the cost of operating branches, and the banks opened far more branches. Better still, it freed up bank tellers from the lower-value task of dispensing cash to customers and gave them more time to forge relationships with their customers and sell them financial products. ATMs, therefore, created new opportunities for bank tellers to undertake higher value tasks, leveraging their problem solving and interpersonal skills. Indeed, Boston academic James Bessen found that between 1982 and 2012, employment grew significantly faster in computer-based occupations such as graphic design.

    This is a partial answer to the automation paradox. How many textile workers are there in America today? And what about the bank branches that are closing now, replaced by apps, you ask? In the examples above, the machines complemented human labour, which became more valuable as a consequence. However, in some cases, the role being undertaken can be fully automated, significantly decreasing the need for human workers. Also, at some point, demand increases don’t keep up with additional productivity gains and the labour required decreases. What history shows us is not that technology doesn’t eliminate jobs, but, rather, that what is true for a single job or industry is never true for the economy as a whole. 

    This is best understood through agriculture. In the long run, spectacular improvements in productivity surpassed increases in demand. The proportion of the US workforce employed in agriculture declined from 40% in 1900 to just 2% today. Not only did technology eliminate jobs, but in the last century, it annihilated most of them. Nevertheless, as we have already seen, long-run employment in the economy as a whole did not decrease.

    So why didn’t it? Although automation improved productivity in agriculture, thereby lowering production costs and the price of goods, it could not continue increasing demand. There is a limit to how much food a household can consume, so instead of increasing demand, the net result was that food started to take up an ever smaller proportion of US household income, declining from 45% in 1900 to under 10% today. This gave households more money to spend elsewhere, it increased their purchasing power and the wealth of society as a whole.

    As a society, we could have responded by working less. Theoretically, as US economist David Autor points out, a US worker wishing to attain the average living standard in 1915 could do so by working just 17 weeks a year, one third of the time. However, people don’t view this as an attractive trade-off between work hours and income, given 1915 standards would be wholly inadequate to people today. 

    This increased productivity, and the wealth it creates gets spent on products and services that were previously too difficult or too expensive for us to afford. Demand for restaurant meals, hairdressing, manicures, personal fitness and tourism has grown rapidly over the past century, creating additional jobs. More importantly, humanity’s ingenuity uses the wealth created and time saved to invent new products and services, which then drive consumption. In the words of Thorstein Veblen, “invention is the mother of necessity”. Many of the industries we work in today, such as healthcare, electronics and computing, barely existed or had not been invented a century ago to the extent that one-third of the jobs created in the US in the past 25 years, were in occupations or fields that did not previously exist.

    This is why labour-saving inventions have not reduced long-run employment. Human insatiability means we always want more. Human ingenuity always provides something new or better for us to produce and consume, reinvesting the time our inventions save and the wealth they create into bigger and better things. This is not just the story of why we have jobs, but, the story of why we advance as a species.

    It is easy to understand why countless great minds were firm believers in technological unemployment. If you went back to 1900, when 40% of the workforce was employed in agriculture, and told them that only 2% of the population would be employed in agriculture by 2010, they would surely wonder what work people would be doing. It’s almost certain they wouldn’t predict the electronics or computer industry, yoga instructors or app developers. Throughout history, forward thinkers have been able to realise that automation could eventually lead to fewer jobs in certain industries, but, what they weren’t able to do was imagine a sufficient number of new jobs that could replace them.

  6. Is This Time Different?

    The historical precedence would indicate technological progress doesn’t cause long-run unemployment. At least part of the current anxiety is likely to reflect the age old problem of failing to imagine the jobs of the future. The scale and speed of the technological shift in the next few decades may be profound, but, it has been rapid and profound before. In 1900, 40% were employed in agriculture. By 1950, following the green revolution, the figure had plummeted to 11%.

    There is no guarantee the future will follow the past. It is credible that advances in AI could mark a break with previous patterns of automation, just as the automobile eventually replaced the widespread use of horses: up until that point, technology in the form of the saddle and carriage had complemented the horse, not supplanted it. 

    A long-held reason against automation totally substituting for human labour rather than predominately complementing it has been that of Polanyi’s paradox. Polanyi’s work in the 1960s highlighted that many human skills are only understood tacitly and we therefore “know more than we can tell”. This has put a theoretical constraint on automation because if we cannot understand and codify how we perform a task then it becomes impossible to program a computer to do it. Deep learning arguably changes this, because it doesn’t require programing; only data and experience to learn, such as observing the undertaking of a task. The other argument for why this time is different is that the next few decades will, in the view of some, see us approach singularity. This would see human ingenuity surpassed by AI. The underlying reasons for the continuation of long-run employment have been human insatiability (part of the human condition and unlikely to ever be quenched) and human ingenuity. The latter of those could, therefore, be threatened by superior AI ingenuity.

    Though arguments for this time being different may seem compelling, there are two things to keep in mind. The first is that while advances in computing power may increase the number of calculations per second, this only provides raw computing power. It doesn’t give us creativity or empathy. Deep learning could eventually turn that computing power into such abilities, but, building creative and empathic computers is still a magnitude of difficulty greater than getting them to recognise images or speech. We don’t yet understand the limits of this approach or of the data sets required to fuel it.

    The second factor to keep in mind is long-term human improvement. Theoretically, we could eventually develop machines that exceed natural human biologically-enabled ingenuity, mastering even creativity and empathy. At this point, the longstanding human competitive advantage of shifting our labour to more complex tasks will no longer be viable. The machines will simply be better at everything, taking over both work and invention. Nevertheless, it is feasible, however, that rather than obsolescence, we improve ourselves through a combination of genetic engineering and by merging with the very technology that threatens our usefulness. Such a change isn’t just confined to the realms of science fiction.

    We have, to some extent, already become inefficiently fused and augmented with technology.

    We use computers on our person to supplement our cognitive capabilities, allowing us to recall and access information instantaneously, navigate and communicate across vast distances. The process of fusing humans with technology at a biological level is being targeted by Elon Musk’s new start-up called Neuralink, with the objective of developing ‘neural lace’ technology, which would implant tiny electrodes into the brain, with a view to improving memory and cognitive abilities. As technology moves speedily forward, there is no reason to believe that we will have to be constrained by the glacial progress of evolutionary biology and this is how we may escape the fate of horses in the 19th century.

  7. Painful Adjustments

    Automation has created enormous wealth and hasn’t yet dented long-run employment, but it is not without high cost for those people and communities that are displaced by it. Hard-learned skills become obsolete and workers are often unable to adapt and retrain. Automation may continue to bring society greater wealth and a rise in living standards, but, there are adjustments costs for groups, which we should not gloss over.

    In the industrial revolution, mechanical weaving looms displaced the highly-skilled artisans who used handlooms at home to produce cloth. The new machinery created huge wealth and employed far more people, but, it was not the artisans who benefitted from it, but a new, low-skilled section of the workforce who found employment operating and maintaining the machines. Those new roles fell mostly to children and women. It turns out the Luddites, who were the artisan textile workers, weren’t wrong to worry for their future. The same could happen again. An accelerated period of unmitigated automation could lead to bad fortune for many and therefore even social unrest if not properly managed.

    There is no fundamental rule that the wealth created by automation will be distributed evenly even in the long run. The evidence today points to job polarisation and a hollowing out of the middle class, driven by a combination of technologically-driven automation and technologically-driven globalisation. In America, the number of high-skilled, high-wage jobs requiring a higher level of education continues to grow robustly. Higher educational requirements make it difficult for American workers to be supplanted by cheaper labour pools in the developing world. Furthermore, the roles are sufficiently complex or non-routine which make them difficult to be automated by software or machinery.

    Low-skilled, low-wage and low-education jobs are also growing robustly. Jobs within food services, cleaning and elderly care are not disrupted by cheaper overseas labour pools, as they have to be close to where their services are consumed. The roles often rely on non-routine tasks and interpersonal skills, meaning they also cannot be easily automated. However, these jobs do not command high wages because there is no shortage of workers to fill them.

    Those in the middle are arguably even less lucky. Middle-skill, middle-wage and middle-education jobs are declining in America and Europe. They are at threat from both cheaper overseas labour pools and from automation, both blue collar factory jobs by machines and white collar clerical jobs by software. Numerous middle-skill jobs follow clear processes that are arguably more susceptible to codification. Middle-skill occupations accounted for 61% of all jobs in America in 1979. Today, they account for just 43%. 

    The shockwaves from this hollowing out are keenly felt. Inequality is increasing and so is disillusionment with the economic system that has served the Western world so well in the past. It is also partly responsible for the polarisation of our politics. As the middle is eroded, it is removing rungs from the economic ladder harming socio-economic mobility. Both globalisation and automation do assist the poorest in society by lowering the price of goods, but, this is often less perceptible and of little comfort in a society that does not act to mitigate the impacts of technology.

  8. Middle-skill, middle-wage and middle-education jobs are declining in America and Europe.
  9. A Societal Response

    Inequality and suffering is not determined by technology, but our response to it.

    The progression of technology, just like the tools it creates, can be used for good or ill. How we distribute the enormous wealth and how we mitigate the impacts are entirely in our own hands. Whether it leads to a utopia or a dystopia is not determined by the development of technology but the response of our society and its institutions.

    In the past, societies have taken action to deal with the impacts of automation. In the late 19th century, it was realised that agricultural employment was declining, industry was rising, and that children would need additional education to earn a living. In 1900, the typical American had an education equivalent to middle school (ages 11-13) and 11% were illiterate. In response, over the first four decades of the 20th century, the United States became the first nation in the world to deliver universal high school education up to the age of 16. Tellingly, it was the farming states that led the high school movement.

    The high school movement helped reskill the American workforce for a more industrial-focused economy. The increased supply of high school graduates (at the time viewed as highly educated), reduced their earnings premium and consequently reduced inequality. The same process of heavy state and public investment was used after the Second World War to increase college and university education. 

    Education is humanity’s response to automation. It is our software upgrade to ensure our workforce remains relevant. It is our way of upgrading the human workforce within the constraints of the hardware that is human biology, while we wait for Musk’s neural lace. It is also our key remedy to inequality. In the words of Horace Mann, “education is the great equaliser”.

  10. The work of Goldin and Katz views this as a race between education and technology. Their work shows that in the first half of the 20th century, educational attainment raced alongside, if not ahead, of the technological advances that tend to increase relative demand for skills. In 1890, the average years of schooling was 8; by 1940 it was over 12 years. The increase in educated labour meant that supply kept pace with demand, keeping both educational premiums and inequality in check. The last few decades have seen increases in educational attainment slow as technology now races ahead. Contrary to popular belief, the present discounted value of college relative to high school graduates net of tuition fees has risen materially in the US. In 1965, the premium was worth $213k for men and $129k for women. Today, the premium, adjusting for inflation1 is $590k for men and $370k for women.

    Those at the vanguard of AI are clear that, in the words of Sebastian Thrun, education is the “antidote to the ongoing AI revolution”. Andrew Ng, Google and Baidu’s former AI chief, like Sebastian Thrun, has also founded his own innovative education company because he even believes AI scientists “have an ethical responsibility to step up and address the problems we cause”. The response must surely be for education to not just increase in depth and reach, but, to focus more on that which computers cannot do. To focus primarily on harnessing human creativity rather than repetition and standardised testing. AI could also be used to create better learning systems that are tailored to individuals and are adaptive, while online access could improve access to lifelong learning.

    Beyond education, just as we did after the previous waves of automation, we will also have to use the great wealth created by it to strengthen our welfare state to help those left behind and to limit the tragedy automation can cause. If automation is as pervasive as some fear, we would certainly have the resources to do so, with a basic universal income now being widely suggested.

     

    1. Both sets of figures are in 2009 US dollars.

     

  11. What can we take from this? Technological progress does not necessarily equate to mass unemployment, but does require us to rise to the challenge of redistributing the wealth it creates. There is no reason to materially slow technology, but, there is great reason to reinvest its bounty in education and a strengthening of social safety nets. The importance of an education system, particularly one that harnesses creativity, will have increasing value in the future. I think education should become increasingly core to our views on different economies. These issues also reinforce my concern that cheap labour, the comparative advantage of emerging economies, could start to erode. Is there any better way than education to assess which emerging markets will prosper in the decades ahead? Overall, the present represents the pinnacle of human technology, wealth and living standards and I believe the future will continue to edge ever closer to utopia, rather than heading to dystopia.
  12. Important Information and Risk Factors

    The views expressed in this article are those of Lawrence Burns and should not be considered as advice or a recommendation to buy, sell or hold a particular investment. They reflect personal opinion and should not be taken as statements of fact nor should any reliance be placed on them when making investment decisions.

    This communication was produced and approved on the stated date and has not been updated subsequently. It represents views held at the time of writing and may not reflect current thinking.

    Baillie Gifford & Co and Baillie Gifford & Co Limited are authorised and regulated by the Financial Conduct Authority (FCA).

    Baillie Gifford & Co Limited is an Authorised Corporate Director of OEICs.

    Baillie Gifford Overseas Limited provides investment management and advisory services to non-UK Professional/Institutional clients only. Baillie Gifford Overseas Limited is wholly owned by Baillie Gifford & Co. Baillie Gifford & Co and Baillie Gifford Overseas Limited are authorised and regulated by the FCA in the UK.

    Baillie Gifford Investment Management (Europe) Limited provides investment management and advisory services to European (excluding UK) clients. It was incorporated in Ireland in May 2018 and is authorised by the Central Bank of Ireland. Through its MiFID passport, it has established Baillie Gifford Investment Management (Europe) Limited (Frankfurt Branch) to market its investment management and advisory services and distribute Baillie Gifford Worldwide Funds plc in Germany. Baillie Gifford Investment Management (Europe) Limited is a wholly owned subsidiary of Baillie Gifford Overseas Limited, which is wholly owned by Baillie Gifford & Co.

    Persons resident or domiciled outwith the UK should consult with their professional advisers as to whether they require any governmental or other consents in order to enable them to invest, and with their tax advisers for advice relevant to their own particular circumstances.

    Important Information Hong Kong

    Baillie Gifford Asia (Hong Kong) Limited 百利亞洲(香港)有限公司 is wholly owned by Baillie Gifford Overseas Limited and holds a Type 1 licence from the Securities & Futures Commission of Hong Kong to market and distribute Baillie Gifford’s range of UCITS funds to professional investors in Hong Kong. Baillie Gifford Asia (Hong Kong) Limited 百利亞洲(香港)有限公司 can be contacted at 30/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong. Telephone +852 3756 5700.

    Important Information South Korea

    Baillie Gifford Overseas Limited is licensed with the Financial Services Commission in South Korea as a cross border Discretionary Investment Manager and Non-discretionary Investment Adviser.

    Important Information Japan

    Mitsubishi UFJ Baillie Gifford Asset Management Limited (‘MUBGAM’) is a joint venture company between Mitsubishi UFJ Trust & Banking Corporation and Baillie Gifford Overseas Limited. MUBGAM is authorised and regulated by the Financial Conduct Authority.

    Important Information Australia

    This material is provided on the basis that you are a wholesale client as defined within s761G of the Corporations Act 2001 (Cth). Baillie Gifford Overseas Limited (ARBN 118 567 178) is registered as a foreign company under the Corporations Act 2001 (Cth). It is exempt from the requirement to hold an Australian Financial Services License under the Corporations Act 2001 (Cth) in respect of these financial services provided to Australian wholesale clients. Baillie Gifford Overseas Limited is authorised and regulated by the Financial Conduct Authority under UK laws which differ from those applicable in Australia.

    Important Information South Africa

    Baillie Gifford Overseas Limited is registered as a Foreign Financial Services Provider with the Financial Sector Conduct Authority in South Africa.

    Important Information North America

    Baillie Gifford International LLC is wholly owned by Baillie Gifford Overseas Limited; it was formed in Delaware in 2005. It is the legal entity through which Baillie Gifford Overseas Limited provides client service and marketing functions in America as well as some marketing functions in Canada. Baillie Gifford Overseas Limited is registered as an Investment Adviser with the Securities & Exchange Commission in the United States of America.

    Potential for Profit and Loss

    All investment strategies have the potential for profit and loss, your or your clients’ capital may be at risk. Past performance is not a guide to future returns.

    Stock Examples

    Any stock examples and images used in this article are not intended to represent recommendations to buy or sell, neither is it implied that they will prove profitable in the future. It is not known whether they will feature in any future portfolio produced by us. Any individual examples will represent only a small part of the overall portfolio and are inserted purely to help illustrate our investment style.

    This article contains information on investments which does not constitute independent research. Accordingly, it is not subject to the protections afforded to independent research and Baillie Gifford and its staff may have dealt in the investments concerned.

    All information is sourced from Baillie Gifford & Co and is current unless otherwise stated.

    The images used in this article are for illustrative purposes only.

     

    Ref: 39798 ALL WE 0080

  13. Lawrence Burns

    Investment Manager
    Lawrence graduated BA in Geography from the University of Cambridge in 2009. He joined Baillie Gifford the same year and has spent time working in both the Emerging Markets and UK Equity departments. Lawrence is a co-manager of the International Concentrated Growth strategy as well as a member of the EAFE Alpha Portfolio Construction Group. He travels extensively researching his particular interest in how pervasive technology and China are changing our world.