Uni of Newcastle Boardroom Lunch, Sydney Business Chamber, Thursday, August 23, 2018
I’m pleased to be invited back with fellow alumni of the university to talk about the disrupted economy. But though, as someone working for a newspaper, I’m only too well aware of just how disruptive digital disruption can be to whole industries and the people working in them - with, no doubt, much more disruption to come – if you’ve come today hoping for confident, detailed, spine-tingling predictions about which industries are next and how many thousands of jobs will be lost, I’m sorry to disappoint you.
Rather, I’m going to do roughly the opposite and give you the antidote to all the frighting predictions about how our world will be turned upside down and how disastrous it’s going to be. I’ll give you seven things to beware of.
First is, beware of futurists – supposed experts who speak with great confidence and specificity about what the future holds. Their forecasting record is even worse than economists’. Humans have an insatiable desire to know what the future holds, which gives us an unquenchable appetite for the output of all manner of soothsayers and witchdoctors. We keep forgetting the simple truth: no one, but no one, knows what the future holds. It follows that people who claim to know should be viewed with scepticism. If they’re genuine experts – scientific experts - their predictions will be hedged about with qualifications. Practitioners of futurology, however, are part of the entertainment industry. There’s great demand for their product and they know that the more confident they sound about their predictions, the more they’ll please their audience.
Second, beware of believing that “this time is different”. It’s always possible that this time really will be different, but just remember that people say that every time – during every boom in the economy or the sharemarket or the property market – and it almost never is. This is relevant to the proposition that digital disruption will be bigger and faster and more devastating to employment than we’ve ever seen, because the installation of labour-saving technology has been occurring continuously since the start of the Industrial Revolution more than 200 years ago, and in all that time we are yet experienced technology-caused mass unemployment. This time it may be different, but a betting man or woman would say the greatest likelihood is it won’t be.
Third, beware of believing urban myths. Futurology teems with confidently asserted, seemingly scientific predictions that are so widely repeated they take on the status of accepted facts. In the 1980s, when I first joined the debate about the employment effects of computerisation and the future of work, one of the assumed facts was that each word-processor that firms installed would replace the jobs of, I think it was, 12 typists. But where did this factoid come from? A Harvard study? A presidential report? Few of the people who repeated it ever knew – or bothered to inquire. But someone traced it back to an advertising claim made by a firm selling word-processors.
One current widely quoted factoid is that 40 per cent of Australian jobs are at high risk of automation in the next 10 to 15 years. Hear that one before? The origins of this one have been traced. CEDA commissioned some engineers to take the figures in a study of the American job market and simply modify them for Australia. The American study was made by some entrepreneurial academics at Oxford. But when economists examined the Oxford study – as some Australian and OECD economists did – they found the methodology rough and unconvincing. The OECD study said a more believable estimate for Oz was not 40 per cent but 13 per cent.
Another widely quoted factoid is that “today’s 15 year-olds will likely navigate 17 changes in employer across five different careers”. Heard that one before? I’ve seen it quoted in formal reports and footnoted, without being able to trace its supposedly academic origins. People who tell us how dramatically the world of work is going to change and how much education and training need to change in anticipation of that change say our youngsters need to be taught, among other things, the skill of “critical thinking”. I agree. So let’s apply a little critical thinking to that factoid. Did you note how remarkably numerically specific it is? That’s suspicious. Next, ask yourself how easy it is to believe those remarkable numbers? 17 different employers; five different careers. But the key question we should always ask of futurology is: how would you know? Do you have a time machine? I find it impossible to imagine how you could take all the children born 15 years ago, peer into each of their working lives over the next 55 years and decide how many jobs and careers they’ll have. And do it with such accuracy it could be summarised in two numbers, 17 and five. It’s literally in-credible. The one thing to say for it is that cramming that many employers and careers into just 45 years wouldn’t leave much time for unemployment.
Fourth, beware of assuming that whatever machines can do for us, we’ll want them to do for us. Futurists make themselves expert on what’s coming next from such things as artificial intelligence. What they don’t know much about is human nature. But leave out the human factor and your predictions about how our world will change stand a high chance of being off-beam. Take the confident predictions a decade or two back that, by now, we’d all be working from home. It’s certainly technologically possible, even cheap and easy to organise. But it’s happened to only a limited extent. Why? Because it doesn’t suit many bosses, and it doesn’t suit many workers.
Or, take the long-established move to online financial markets for shares, bonds and currencies. In principle, this means financial traders could be spread around all the cities and towns of Australia with decent internet links. In practice, almost all the traders remain crammed into two or three streets of the Sydney CBD. One of the unexpected features of the knowledge economy is that more and more of the nation’s GDP is being produced in the CBDs of Sydney, Melbourne, Brisbane and maybe Perth. Why? Because knowledge transfers better when people are face-to-face.
To take a more prosaic example, it will soon be technologically possible for cafes and restaurants to be manned by robots rather than people. But I don’t see much of it happening. Why not? Because few people would enjoy eating in such impersonal circumstances. Let me make my own futurist prediction. I think it won’t be long before many firms bring their overseas call centres back on shore. Why? Because so many of their customers hate it.
Fifth, beware of assuming that what you see is all you get. What everyone sees with computerisation, automation, digitisation and the rise of AI (artificial insemination) is all the jobs being destroyed. Non-economists assume that’s the end of the story. Economists earn their living by knowing that it’s not. You have first to remember that some new jobs – not all that many - are created by the workers needed to sell, distribute, maintain, replace and modify all the robots. Much more significantly – but much harder to see – is the realisation that the instillation of labour-saving equipment – including robots – doesn’t destroy value, it increases it. That’s why it happens. It doesn’t reduce productivity, it increases it. And increased productivity equals increased income. Ideally, that increased income is manifest in higher wages for those who remain in the industry and lower prices for the industry’s customers. Failing that, the extra income stays in the hands of the owners of the firms in the industry. But wherever the gains from higher productivity end up, when they are spent somewhere in the economy they create new jobs. This is why economists say that new technology doesn’t destroy jobs, it “displaces” them. It reduces employment in the industry applying the new labour-saving technology, but increases employment in many other, often utterly unrelated industries – usually in the services sector and, if our experience over the past 30 years is any guide, particularly jobs suitable for women. This is why 200 years of labour-saving investment have yet to create significant levels of technological unemployment and why overall employment has continued to grow almost every year.
Sixth, beware of forgetting what economists, in their pompous way, call the “policy reaction function”. This is just a fancy way of saying there’s what the market does, and then there’s what the government does in response to what the market’s done. If it doesn’t like the outcomes the market has produced – or, more likely, if too many voters don’t like those outcomes - it will intervene to correct those outcomes.
Finally, beware of assuming from all I’ve said to put future technological disruption into a more realistic context that I think nothing much will change, or that there’s nothing to worry about. I think a lot will change and there’s a fair bit to worry about. But these are things you can ask me about now.
