The French newspaper Libération devoted four pages of a recent issue to an investigation of the idea of a minimum basic income for every adult. Among the surprises for an Anglo-Saxon reader, for whose politicians such a notion is so far out of mind as to be science fiction, is the discovery that in other countries it is very much on the map. The left-wing Spanish party Poderemos has it in its manifesto. The Finns are considering it, and the Swiss will vote on it in a referendum later this year. The practical Dutch are testing it in a pilot in Utrecht.
If Martin Ford is right, such initiatives, far from being a rush of Corbynitis to the head, are necessary and urgent. Rather than signalling the demise of capitalism, he suggests in his impressively researched and soberly argued book, The Rise of the Robots, a basic minimum wage may turn out to be the only way to save it.
Ford is a successful Silicon Valley software entrepreneur, which at first glance makes him an unlikely technological doomsayer. But unlike gung-ho contemporaries such as Ray Kurzweil who enthuse about the coming ‘singularity’, the moment when machine intelligence surpasses the human kind, and the achievement of immortality (sic), Ford concentrates on what is happening in the here and now to make a powerful case that this time it really is different; meaning that the threat in his subtitle – ‘technology and the threat of a jobless future’ – is very likely to come true.
The techno-optimists’ cheerful view of the economic consequences of technological advance is based on faith and precedent. They point out that since the Industrial Revolution first disrupted craft-working in the 18th century, each succeeding wave of progress has produced new markets and industries that generate more and better-paid jobs than they destroy. This is Schumpeter’s ‘creative destruction’ in action, a virtuous circle in which technological innovation drives higher wages and increasing demand that floats all boats. Better education will keep human capability ahead of the machines; economic growth will provide a steady flow of new jobs in partly-mechanised sectors and others that spring up to serve them.
But although many economists (and almost all politicians) are still parroting the old mantras about a return to growth, Ford notes that the positive relationships had started breaking down long before the financial crash of 2008, and substantially pre-dating the current round of techno-acceleration. (Ford writes mainly about the US, but in economics as in other spheres the UK can be relied on to imitate the transatlantic experience with a short time lag, if in slightly less extreme form.)
Thus Ford lists ‘seven deadly trends’ that have been ticking away behind the economists’ comfortable assumptions like a colony of deathwatch beetle: average wages stagnant since the 1980s; a shrinking labour share of GDP; declining labour-force participation (particularly among less qualified men); jobless recoveries (it took until mid-2014 for US employment to regain pre-crash totals, by which time the working population had increased by 15 million – in fact the US economy has put on no net new jobs this century); soaring inequality; a diminishing premium and growing underemployment for graduates; and the polarisation of the jobs market between well-paid full-time employment for the very few and part-time and freelance for the many – ‘uberisation’, let’s say.
Now take this malign dynamic and supercharge it with the most powerful general purpose technology (ie whose effects will leave no industry untouched) ever devised – one, moreover, whose advance is accelerating with the undiminished momentum of Moore’s Law. Machine intelligence is improving by leaps and bounds. IBM’s Watson supercomputer defeats human champions not only at chess, a bounded problem, but also at the game show Jeopardy!, a cryptic, unbounded one. Watson is now being deployed commercially. While Artificial Intelligence (AI) is still the ‘narrow’ variety, the ‘strong’ version, or Artificial General Intelligence (AGI), is now being vigorously pursued not just in research labs, as in the past, but competitively by ambitious, well-resourced giants such as Google, Facebook and Apple, pushed by powerful commercial incentives to make it work. Another AI, Artificial Intuition, is in the works.
A recent Forbes contribution (entitled none too subtly ‘Deep Learning And Machine Intelligence Will Eat The World’) could not have put it more clearly: ‘The effects of this technology will change the economics of virtually every industry. And although the market value of machine learning and data science talent is climbing rapidly, the value of most human labor will precipitously fall.’ Publications (reputedly including Forbes) already employ software to write news stories and reports; in a few years time 90 per cent will be machine generated, in one estimate. In many other industries, automation, says Ford, is simply ‘the logical next step’. Beware, he warns: if you’re working with computer software, you’re probably training it to replace you.
In a much quoted 2013 report, the Oxford Martin School suggested that 47 per cent of US jobs would be susceptible to computerisation in the next two decades. Later estimates raise that to 80 per cent. Given that the essence of computerisation is enabling more to be done with less, is it conceivable that new industries based on it will be labour intensive? Looking at early evidence from those avatars of the new economy YouTube (which had a value of $1.65bn when it was acquired with its workforce of 65) Instagram ($1bn and 13) and WhatsApp (a staggering $19bn and 55), the answer seems pretty clear. Uber and Airbnb just underline the point: for the first time new technology is not only creating fewer jobs than it consumes: it is also creating worse ones. The circumstantial evidence keeps flooding in. Thirty per cent of US science and technology graduates are currently labouring in jobs that don’t need degrees; most UK graduates are in non-graduate jobs. ‘The assumption that we will transition to a more productive … economy just by increasing the conveyor belt of graduates [the method used in the past] is proven to be flawed,’ says the Chartered Institute of Personnel and Development (CIPD).
So far, so not very good at all. But if the scenario follows Ford’s trajectory, the effects will be massively self-reinforcing through the resulting demand deflation. Ford quotes the following exchange between Henry Ford II and union boss Walter Reuther on a factory visit: ‘Walter, how are you going to get those robots to pay your union dues?’ ‘Henry, how are you going to get them to buy your cars?’
At the extreme, radical inequality is unsustainable in the most basic sense: by hollowing out the middle classes, the winner-takes-all economy becomes a contradiction in terms. There won’t be anything to take, because the very rich simply don’t consume enough to keep the wheels turning. Being partly the result of what happens when average consumers do keep consuming, but using debt rather than cash to do so, the 2008 crash should be a warning here. As Ford relates, companies are already abandoning the middle market to chase the 1 per cent of super-spenders; despite increasingly frantic advertising, the average US car is now 12 years old, a forerunner of the tangible consequences of rising inequality. In this context, the gathering cloud of graduate debt overhanging the US and UK looks increasingly ominous not only for individuals with increasingly uncertain earning prospects but also for the economy as a whole, while the policies that created it are revealed as the monstrous false economy they seemed to many at the time.
This is the background against which Ford sets out his proposal for a universal minimum basic income. Ironically, the original progenitor was the English radical Tom Paine, who advocated it as a blow for social justice against the cruelty of emerging capitalism. By contrast, for Ford it is primarily a prop to keep that system going. As he concedes, the idea is controversial, with weighty considerations on both sides. Yet the need for solutions may be even more pressing than he thinks it is. Ford didn't foresee another emerging result of global inequalities, the rising tide of immigration. And like almost every other commentator, whether optimist or pessimist, Ford has internalised, and thus leaves out of his reckoning, the most deadly trend of all: the pernicious incentives at the heart of today’s shareholder capitalism.
This is the first great wave of technological evolution whose justification is not that it benefits humanity but that it benefits shareholders. The lesson of the last 30 years is that Investments driven by self-interest and shareholder value – where the benefits are supposed to accrue to one group in society – do not produce a generalised increase in wellbeing, because they are designed not to. As Jeff Pfeffer succinctly puts it, ‘Economic performance and costs trump employee [and societal] wellbeing’. Under today’s incentives, investments in accelerating technology will just destroy more of it. To be clear about this, consider a quote from the founder of a start-up planning to automate the production of customised gourmet hamburgers: ‘Our device isn’t meant to make employees more efficient. It’s meant to completely obviate them.’ Or this from another start-up entrepreneur warning that executive jobs too are in the firing line: ‘It will not be possible to hide in the C-Suite for much longer. The same cost/benefit analyses performed by shareholders against line workers and office managers will soon be applied to executives and their generous salaries’. Oh yes, the name of this start-up: iCEO.
At this point, as Ford notes logically if bleakly, at the end of his book, falling demand may run out even for further automation. It might. But it would surely be unwise to wait to get that far to find out.
****I'm away at the moment: next piece beginning of September****