Engels and the Luddites…
In 1768 the first automated sawmill in Limehouse was burnt to the ground by disgruntled sawyers. The British Parliament, which was quite decisive in those days, responded by passing an Act making the destruction of machines a felony punishable by death. This did not stop attacks on machines, like Cartwright’s power loom and his wool-combing machine, or papermaking and cotton-weaving machines.
Thought to have been named after Ned Ludd, an apprentice who allegedly destroyed two textile frames in 1779, Luddite riots became common. They led Frederick Engels to argue in his seminal book The Condition of the Working Class in England that the profits from automation went to owners; and that machines impoverished workers. Writing in 1845, he had a point.
Many of the early machines undercut the costs of the home-working artisanal producers. The new factories were not only much more productive, but the equipment could be operated by young – often very young – newly-trained workers. Though some artisans did find work in the new factories, wages were often lower.
In subsequent waves of the Industrial Revolution, the benefits of automation were more clearly spread. But the broader benefits did not, of course, stop threatened workers from objecting – generally through strikes, rather than by smashing machines. Some of the more colourful strikes included:
- 1900 – New York streetlamp lighters’ strike. Policemen tried climbing up and lighting the gas lamps, but children went along behind them, turning them off again.
- 1926 – General Strike in the UK was an attempt to prevent lower pay and worsening conditions for 1.2 million locked-out coal miners.
- 1945 – New York skyscraper elevator operators’ strike. This left workers either having to climb the stairs of the Empire State Building (it has 102 floors) or go home. The owners brought forward plans to install automated lifts.
- 1958 – US longshoremen’s campaign against containerisation. A resolution was agreed with unions, but the global introduction of shipping containers could not be stalled.
- 1968 – US telephone operators’ strike. 160,000 operators refused to place calls, but automated boards were already sufficiently widespread to enable management and their allies to handle most calls.
None of these strikes could stall the adoption of new technology. Indeed, they often had the opposite effect – encouraging owners to accelerate adoption.
The sector that bore the greatest change from the introduction of machinery was agriculture. From the 1930s to the 1960s the share of farms using tractors rose from 17% to 80%, with a parallel collapse in the use of horses. Without horses to feed and groom, farmers saw their productivity rise by 30%. Improved ploughing, sowing and automation of the food manufacturing chain brought further leaps in productivity. Between 1870 and 2015, the percentage of the US population working in agriculture fell from 46% to 1%.
You’ve never had it so good…
Periods of adjustment to new working practices certainly were distressing for those whose skills had been superseded. Many, especially the old, struggled to fit into the new economy.
Overall, however, society benefited. Technology brought us electric light, cheaper food and goods and the opportunities of travel. It took the drudgery out of house-cleaning and laundry. Displaced gas lamp lighters, farmhands, buggy drivers and house maids, generally, found new roles – often in more salubrious conditions. Many new products created their own demand and spawned new industries; cars and aeroplanes stimulated the development of tourism and road haulage. It is startling to recognise that nearly half of current American employment is in jobs that simply did not exist in 1870.
Technology eventually improved incomes. The productivity gains from the Second Industrial Revolution, from 1870 through to 1914, led to rapidly rising real wages across much of society. This trend continued for most of the 20th century. Strong growth in equity markets shows that capitalists continued to reap the financial benefits too.
Those who adapted better to the change tended to be the better educated and those able to learn new skills. This effect led to ‘class’, which used to be defined by job (manual labour, professional), now commonly being defined by educational level (degree or no degree).
Explosion of average wages, driven by industrial revolutions
The computer age…
In my own working lifetime, technology has continued to have an impact. When I first worked in the City, I was unusual in having a pocket calculator. Most offices had slide rules and log tables.
Younger generations struggle to imagine life without computers, but you do not have to have been working long to see how the rapidly increasing power and decreasing cost of computers is threatening to put skilled jobs at risk today.
Explosion of computing power since 1980
Understandably, in a time of rapid change, people are nervous. Trying to assess the impact of a technology seen to ‘replace’ workers is tricky and it is easy to over-simplify – as politicians do when they instinctively decry ‘lost jobs’. Once proven, new technologies will be deployed somewhere, even if local politicians try to slow their adoption to protect jobs. So will computers make us redundant? History suggests not. There will, however, be winners and losers.
How computer technology is developing…
Technological advances are taking robotics from heavy manufacturing towards a range of new sectors. Most industrial robots are currently used in automotive manufacturing, making repeat movements with heavy metal components. Improved sensors and smaller robot arms allow automation to grow in sectors such as food handling, textiles and even surgery in ways previously unknown. This automation seems mainly of the ‘productivity-enhancing’ type, rather than the labour-replacing type – though some low-skilled jobs, such as fruit sorting, will undoubtedly be lost.
But the new wave of automation also extends computing power into artificial intelligence. It does not mean that computers are starting to think more like people, but they are now faster at reaching the same conclusions (usually in highly specialised and technical areas). Their growing ability to learn from human-generated patterns is powering translation software and the development of things like autonomous driving. Here again they are often faster than humans, but only when faced with patterns they recognise.
The Moravec paradox reflects some of these limitations. It tells us that computers are great at doing things we find difficult; yet struggle with things even a baby can do. They can win a game of chess but cannot put the pieces back in the box. They follow rules, while we follow patterns. In the Henn na Hotel in Nagasaki they have recently disposed of their robots: they were waking up guests who were heavy snorers and asking them to repeat their ‘requests’!
Recently computer scientists have looked at various jobs and tried to analyse how much they rely on logical processing and how much on ‘value judgment’ – the aspect that computers find hard. The idea is not so much that computers should replace humans, but supplement them.
Impact of computers on jobs…
Since the 1980s, the growth in computerisation has driven new divisions in the labour market and the distribution of incomes. In the US income inequality has increased – particularly over the past 20 years. Blue-collar manufacturing workers have been especially affected and this trend will continue. The contentious list of ‘at risk’ trades for the next wave of automation includes farming, transportation, food preparation, retail and construction.
The number of jobs that require little further education has remained the same: catering, security guards etc. However, jobs that require value judgments – management, teachers and other professions – have become more productive due to the support of computing. Where the value judgment is the larger part of the job, this has led to higher pay, but where the computerisation allows more candidates to qualify for the job, pay has tended not to rise. However, job numbers in these areas have all risen as the improved product attracts greater demand. Trades where automation may create more jobs than it replaces include healthcare, financial and legal services.
As in the past, automation is creating friction. Perhaps for political convenience, attention is being deflected towards China, with accusations of unfair trading. It is hard to separate the two issues as offshore manufacturing savings increasingly come from automation in those locations, rather than low wages. Tariffs are unlikely to offset them. With US unemployment currently low and wage inflation rising, we may see this issue as a more modest vote winner in the 2020 elections than it was for Trump in 2016.
In the 1920s, John Maynard Keynes predicted that productivity would rise eightfold over the next century and that the working week would therefore shrink to 15 hours. In fact, productivity has risen ninefold, but the working week has not shrunk. Instead of opting for greater leisure as incomes have risen, we have found new things to buy.
The jobs that the new wave of automation challenges may well extend further into white-collar than earlier waves. Politicians threatening to ‘tax robots’ may have misunderstood the situation: computing power is hard to tax and in manufacturing, there is little point in slowing the use of robots unless the Chinese and others promise to do so too.
Many jobs will be challenged and have been challenged. The working environment today is one that could be barely guessed at a decade, let alone two decades, ago. Many of our children and grandchildren will need to retrain mid-career and hopefully such opportunities will continue to expand. Some have argued that as wealthier families invest more in their children’s education, they will cope better with the changing job market than others. This seems a strong argument for the state to improve education for all and throughout life, as social mobility may reduce as skill levels and adaptability challenges rise.
How we invest around these trends…
In the Artemis Global Select Fund, one of our largest investment themes for some time has been Automation. One of our larger investments, Daifuku, is a medium sized Osaka-based business, little known amongst investors, but well known in retail and distribution as the global leader in warehouse automation.
While we remain excited about technical developments in this area, it has been a poor performer over recent years as the US-China trade dispute has made multinational companies slow to implement their capital investment plans. This is merely delaying the inevitable. Companies with global supply chains are recognising the need to revisit and extend their plans for automation to help reduce offshore manufacturing costs and accommodate recently introduced tariffs. It is unlikely that President Trump intended to stoke a boom in the mainly Japanese automation sector, but we hope to thank him for it. It is striking how few of the world’s leading automation companies are based outside Japan.
Despite sales of advanced automation being sluggish, most of our investments across the rest of the portfolio continue to benefit from technological change. Most obviously, companies in our Online Services theme are driven by increased computing power, especially chip-design companies such as Synopsys. In our Screen Time theme, Vodafone and Comcast connect digital networks. From Scientific Equipment, makers such as Thermo Fisher to Nestlé are using big data to anticipate changing consumer trends. Most of the companies we invest in use modern technology to plan and drive their growth.
For some companies, the threats and challenges seem larger than the opportunities. We are avoiding the banking sector at present. Some banks have cleaned up their balance sheets and are finally seeing some loan growth. However, new financial operators such as PayPal, Revolut and Square are based on low-cost, technology-based systems that can take blocks of income from the traditional banking model. Banks are trying to keep up with modern technologies, but institutional sluggishness and the concern of cannibalising existing profit centres often hold them back.
We are investing during a period of very rapid technological change. Companies that make the most of the opportunities will, no doubt, thrive and provide strong investment returns for years to come. Those that lag are worth avoiding even when they look ‘cheap’. Technology does not depend on economic cycles, nor particularly on trade deals. Hopefully our politicians will embrace the advances and will anticipate the social issues that may result. In the long run, a more productive economy can be good for us all.
Simon Edelsten manages the Artemis Global Select Fund; visit the Artemis website for information about the fund’s performance and positioning.
THIS INFORMATION IS FOR INVESTMENT PROFESSIONALS ONLY. IT IS NOT FOR USE WITH OR BY PRIVATE INVESTORS. The fund is an authorised unit trust scheme. For further information, visit www.artemisfunds.com/unittrusts. Third parties (including FTSE and Morningstar) whose data may be included in this document do not accept any liability for errors or omissions. For information, visit www.artemisfunds.com/third-party-data. Any research and analysis in this communication has been obtained by Artemis for its own use. Although this communication is based on sources of information that Artemis believes to be reliable, no guarantee is given as to its accuracy or completeness. Any forward-looking statements are based on Artemis’ current expectations and projections and are subject to change without notice. Issued by Artemis Fund Managers Ltd which is authorised and regulated by the Financial Conduct Authority.