The future of work when there is no work left to do
A story showed up in the press recently, sourced from a user comment on the Reddit discussion site. It was by a guy who was sacked from a San Francisco tech company that had employed him for six years. His crime? About eight months into his six-year stint, he had written a program (he was a programmer) that did his job for him. So for the balance of his tenure at the firm, he had been coming to work every day, playing video games and going to the gym. That is to say, he was still collecting his pay, but the program he had written was doing his job for him. On Reddit, this guy called himself FiletOfFish1066, which is great, but I’m going to shorten that to FoF.
One thing I found baffling about FoF’s story—and let’s assume for now that he is telling the truth—is why the company didn’t promote him rather than sack him. Why not get him to see if he could automate other jobs the company was paying people to do? Now, on one level, the firm doesn’t want to be seen to be rewarding this sort of behaviour, but from another, surely it would be to their advantage to use this guy’s skills rather than lose them?
What is really interesting about the story, though, is that it illustrates in microcosm a process that takes place every day, one that in many ways defines the modern economy: human workers are replaced by technologies of one sort or another. If you like, robots are taking our jobs.
This is hardly a new process, but there is a lot of evidence suggesting that with advances in computing power and the growing sophistication of things such as machine learning, 3D printing, driverless vehicles and other forms of advanced robotics, the process of technological unemployment is accelerating. Many suggest we are heading for a world in which large swathes of jobs will simply disappear because they can now—or will at some point in the next 20 years—be performed by machines. There are those who argue strongly that our concerns about this sort of technological unemployment are exaggerated, that although work will change and jobs will be lost, technology will—as it has in the past—create new jobs that people will be able to do. This is entirely reasonable, but I think it misses a bigger point about the future of work.
My basic position—the position I argue in my recent book, Why the Future Is Workless—is that ‘will a robot take my job?’ is the wrong question. The real issue has to do with the ways in which wealth will be created as we move from a manufacturing-based economy to one that depends on the manipulation of knowledge and information. So while I have no doubt that a large number of new jobs will be created in the near future, what concerns me is that there won’t be enough them, that the sort of information economy we are moving into simply doesn’t need workers in the way that a manufacturing-based economy does. Under such circumstances, the compelling question is not whether a robot will take your job but whether that will be a good thing or a bad thing. It is therefore less a technological problem than a social and political one.
To help us into this topic, let’s perform a thought experiment based on what happened with FiletOfFish1066. Imagine the company he worked for had kept him on instead of sacking him. And imagine he was able to automate, say, 20 per cent of the other jobs being done by paid employees.
Now imagine that instead of sacking the workers FoF identified, the company decided to keep paying them. Or maybe not so much ‘decided’ as was obliged to. Clearly the firm in our original example had no problems continuing to pay FoF for six years of work he wasn’t doing, so there is no particular reason why they couldn’t continue to pay the other 20 per cent of workers he identified. By continuing to pay their wages, the firm would in effect be paying a kind of hypothecated tax, and the money that would normally be saved by sacking people would continue going to the people sacked rather than into company profits. It seems like an outlandish idea, not just because we tend to think that it is the firm that should be rewarded for increasing productivity and efficiency, but because most of us would think there is something unethical about forcing a firm to pay wages to people who aren’t doing anything. Why should those workers get money for nothing? All fair points, but maybe not as self-evident as it seems at first blush.
So let’s try another variation on this. Instead of continuing to pay the wages to the 20 per cent of workers made redundant, what if the firm involved—and firms throughout the economy—paid the money they would’ve paid out in wages to the government, which then distributes it via various welfare programs? That is, what if we changed the usual way firms pay tax—on profits—and instead deemed them to have a certain number of employees whose wages they pay in perpetuity?
As governments throughout the world have found, taxing company profits is an incredibly inefficient way of collecting revenue. It is simply too easy for firms to minimise their levels of profitability or to shift profits from the country in which they are generated to a jurisdiction in which the tax rates are lower or non-existent. A recent report by the Australian Tax Office showed that nearly one-third of our biggest businesses paid no tax at all in the 2013–14 financial year,1 zero, and Australia is hardly unique in this respect. What’s more, over the last 40 years or thereabouts, the proportion of national wealth going to profits rather than wages has increased significantly. Combine this with persistently high levels of unemployment (especially among certain age groups and demographics); the increasing pressure of technology on jobs; and the increasing tendency of paid work to be offered on the basis of short-term contracts rather than as ongoing work built around a career across a lifetime, and you go a long way to explaining the massive increases in inequality that plague us. In other words, our system of employment and taxation is pretty broken. Our political and economic systems are straining under the pressure of being seen to reward the few at the expense of the many, and the many are starting to get angry about it—Brexit being just one example.
And here’s the thing: if predictions about large-scale technological unemployment prove correct—and I think they will for reasons I will explain—then all this is only going to get worse. If you think societies are under strain now with official unemployment rates around 6 per cent (and underemployment an increasing component of the employment figures), what’s it going to be like if that figure doubles or triples as more and better robots come online?
Which brings us back to our thought experiment. What if instead of taxing company profits as we now do, we came up with a formula that deemed firms to have a prescribed number of workers based on some combination of their level of market capitalisation and turnover, metrics they tend not to fudge because it affects investment and their overall professional standing. (Somebody could even come up with an algorithm to work out how many employees a given firm should be deemed to have.) The point is, this is a much harder tax to avoid. Firms would still get the advantage of any technological improvement; they just wouldn’t get to feed the savings automatically into profits.
Sounds crazy, right? But remember, we are just extrapolating here from the story of FoF who got paid for six years, for doing a job that he fully automated, by a company that didn’t even notice the difference. We are really just pretending that instead of workers’ employment being disrupted by technology, it is the firms that bear the brunt. Let’s imagine this system works and government income in the form of taxation is now more secure because the new tax regime is harder to avoid. Wouldn’t we also want to insist on improvements in the way in which governments spend the revenue they were collecting?
As I noted above, human workers are replaced by technology all the time: ATMs replace bank tellers, robots replace factory workers, algorithms replace lawyers.2 It is a story as old as paid employment itself, from the seed drill, to the cotton loom, to—coming soon—the driverless car. When it happens, redundant employees look for a job elsewhere doing similar work; retrain for another career; fill in the time earning a few bucks by maybe working in a café or driving for Uber; or maybe they go on the dole. More likely, they do a combination of all these things. The firms creating the unemployment may initially compensate the workers with redundancy packages of various levels of generosity, but ultimately much of the expense associated with these changes in the labour market falls on governments: unemployment benefits, retraining, education and the rest.
And—even in the increasingly threadbare welfare states of the developed world—this is how it is meant to work. In the long run, the firm saves money by paying out less in wages and thus increases its profitability, a process we measure in terms of efficiency and productivity, increases in which are almost axiomatically seen as a good thing. The theory is that this makes the firm stronger, enables it to increase the remaining workers’ wages, or maybe even expand the business in some way and employ more people. The theory also says that an increased profit means the firm will be paying more tax and so society in general benefits from the redistribution of wealth this causes.
The trouble is, the theory doesn’t hold up in practice, especially in a world in which the work is increasingly done by machines. When workers are replaced by technology, unemployment increases, money that formerly went to labour goes to profits, wealth concentrates in fewer hands, and because the wealthy save more than they spend, economies stagnate. As former US labour secretary Robert Reich puts it:3
Conservatives believe the economy functions better if the rich have more money and everyone else has less. But they’re wrong. It’s just the opposite.
The real job creators are not CEOs or corporations or wealthy investors. The job creators are members of America’s vast middle class and the poor, whose purchases cause businesses to expand and invest …
[But consumer] spending has grown more slowly in this recovery than in any previous one because consumers don’t have enough money to buy.4
All the economic gains have been going to the top.
Which brings us back to FiletOfFish1066. In hiding the technological unemployment he had created—his own—FoF overcame one of the key problems associated with losing your job: he managed to keep earning a wage. Not only that, he exposed the fact that the firm easily managed to pay that wage even though he was no longer doing his job: it took them six full years to even notice. Unfortunately, ‘hiding your technological unemployment from your employer’ is not an idea that scales.
Still, it might be fair to say that hiding technological unemployment from society as a whole scales very well. One of the enduring fantasies of the world post global financial crisis is that the jobs will come back, that there will eventually be enough work for everyone. At the last federal election, for instance, the Turnbull government ran on a promise of ‘jobs and growth’ without explaining where either was likely to come from. The theme of Donald Trump’s Republican Convention was ‘Make America Work Again’. The Obama Administration boasts of having ‘created’ millions of jobs and, indeed, the US unemployment rate has dropped dramatically over Obama’s second term. I cast some scepticism because the figures also show that nearly half the new jobs created have been at poverty
wage levels.
The idea that we are forever on the verge of a moment when the jobs will return is a fantasy that is shared across the political spectrum, from the most left-leaning trade union to the most right-leaning think tank, and every political party in between. Of course, each of them has different ideas about how we should ‘create’ these new jobs, but they all perpetuate, in one form or another, the idea that if we can just get the settings right—pull the right levers, as Paul Keating used to say—everyone who wants a job will be able to get one.
For some, the levers we need to pull are the ones connected to education: we need more STEM skills.5 We need more STEAM skills.6 We need more tradesmen. We need more vocational training that meets employer needs. For others the key is to cut wages: stop paying a minimum wage and penalty rates and you will create more jobs for everyone. Some think the secret lies in cutting company tax and regulation: make firms more profitable by reducing the ‘tax burden’ and they will employ more people. Cut red tape. Wealth will trickle down. Others insist the key to creating more jobs lies in restricting our exposure to the rest of the world: stop immigration, stop free trade. Stop signing us up to international treaties that allow international firms to ignore local laws and outsource labour to the developing world.There are probably other theories too, but I suspect all of them are bullshit. Not that some jobs couldn’t be created in these ways, just that it is bullshit to assert that they are likely to create the sort of full-time employment that we associate with the success of developed countries in the aftermath of the Second World War, that is, the sort of jobs that can sustain a person in a middle-class way across a lifetime.
One reason I say it is bullshit is because of the changing nature of the creation of wealth. We are not just going through an era of change, but a change of era. We are moving from a Fordist to a post-Fordist economy, from one that relies on the creation of goods in factories via production lines to one that relies on the manipulation of knowledge and data through networks. We are seeing the financialisation of the economy where not only is great wealth generated by algorithms buying and selling various financial instruments on world stock exchanges, but where the profitability of, say, a car dealership comes from finance rather than sales: selling cars is not profitable; arranging loans for cars is. Google, Facebook and Uber make their billions by controlling the platforms, not by creating the content. Yes, we still need to make stuff—whether it is cars or iPhones or growing food—but these are precisely the sorts of jobs that are most easily automated, and many already have been.
To some extent economies and societies have been able to absorb these job losses because new jobs have arisen in the service sector, as well as in the culture and knowledge industries, but even these white-collar jobs are now under threat of automation as machines get smarter. If you have, for example, the Xero accounting app on your phone, you have probably already put an accountant out of work. Law firms are increasingly using programs to search case law and for other sorts of legal research, thus potentially displacing entry-level lawyers. Even media companies are using software to write stories, particularly those involving sport and finance. Increasingly it is the professions that are being automated: technology completely reshapes the way in which knowledge is generated and distributed throughout society. As Richard Susskind and Daniel Susskind note in their book The Future of the Professions:
[W]e are on the brink of a period of fundamental and irreversible change in the way that the expertise of these specialists is made available in society. Technology will be the main driver of this change. And, in the long run, we will neither need nor want professionals to work in the way that they did in the twentieth century and before.7
Exactly how many jobs are at risk? Well, no-one really knows because, you know, the future. This leaves enough doubt in the discussion to allow some to argue that fear of technological unemployment is an unfounded anxiety. Nonetheless, there are a number of reputable studies that suggest we can’t afford to be blasé about what is happening.
The Oxford Martin School, for instance, released a report in 2015 that examined 702 job categories and assessed them against various criteria to see how vulnerable they were to automation. They concluded that within 20 years, 47 per cent of those 702 jobs were likely, or highly likely, to be done by machines. A more recent report by the McKinsey organisation was even more startling. They examined 2000 ‘individual work activities’ and came to this conclusion:
The bottom line is that 45 per cent of work activities could be automated using already demonstrated technology. The magnitude of automation potential reflects the speed with which advances in artificial intelligence and its variants, such as machine learning, are challenging our assumptions about what is automatable.
Did you get that? Forty-five per cent of those 2000 activities were vulnerable to already existing technology. We don’t even need to wait for it to get smarter, better or faster. Even if this post-Fordist world does create new jobs—and it undoubtedly will—it won’t create enough of them: the knowledge economy simply needs fewer workers than the manufacturing economy. What’s more, left to the vagaries of market forces, the jobs left over, especially for the unskilled, will be low-paying and extremely insecure. This trend is already apparent.
Anyway, hold all that in your head while we get back to our little thought experiment. Under the conditions of our game, those rendered unemployed by FoF (or by other examples of technological advance), wouldn’t have to hide their unemployment in order to continue to be paid. The government would be collecting tax based on the assumption that firms were still employing people and would be redistributing that income in the form of welfare payments. So it is the type of welfare payment that now becomes important. To see how this might work we need to introduce a distinction between what I will call ‘labour’ and ‘work’. For our purposes, labour is the form of work we do to keep ourselves alive. In other ages that would have included hunting or growing crops or whatever, but since about the time of the Industrial Revolution it has meant having a paying job. So labour in this sense is paid employment.
Work, on the other hand, is those activities that involve doing things that don’t normally attract a formal wage in the way that a job does. We could include everything from housework to caring for children and the aged, to education and community volunteer work. No-one disputes that these things involve effort—and therefore constitute work—but we don’t think of them as work in the same way we think of a paid job as work. Washing up dishes in a restaurant, for most of us, is a different thing from washing the dishes in our own home. What separates them is not the nature of the activity, but the social conditions under which it happens. It is no coincidence, therefore, that most of the tasks I have described as ‘work’—cleaning, caring, volunteering—fall on women because the social arrangements that we tend to associate with post–Second World War prosperity and full employment were also associated with the idea of the man as the breadwinner and the woman as the homemaker. Women were not remunerated by a regular wage but in the form of material support from the breadwinner. They were relegated to the private sphere with all its attendant invisibilities, dependencies and discriminations. To some extent Western societies no longer accept this division, although housework and the rest of it still falls disproportionately upon women, even as they hold down regular jobs outside
the home.
The point is, paid employment done in a workplace is still valued more highly than the work done in homes despite the fact that the formal economy would collapse without the ‘free’ labour provided by the informal economy of the household.
So what if we stopped making this arbitrary distinction between work and labour? What if we dispensed with the social habit of thinking that only those with a job deserve to be paid a living wage? What if we recognised the sort of work done outside the formal economy as valuable too and thus deserving of formal remuneration? What if our societies finally acknowledged how integral the informal economy is to the success of that society and remunerated all those involved in maintaining it? This would mean that the welfare payments the government distributed would not be offered in the form of unemployment benefits but in the form of a wage that sustained people regardless of their formal employment status. The wages extracted from firms for employees they don’t have would be redistributed to everyone who was without formal employment, except that it wouldn’t be money for nothing, but money for doing the formerly unremunerated work conducted in households, in volunteer organisations and the like.
The concept I am describing is generally called a basic income. This is a scheme where governments pay citizens a monthly, unconditional wage sufficient to cover the basics of life. The theory is that it is available to everyone and that it is not means-tested in any way. There is resistance to such a scheme because people freak out about the idea of anyone getting money for nothing (or at least, about other people getting money for nothing). The fact that every experiment with basic income—and there have been a number—shows that it makes people more productive doesn’t alter the opinion of the doubters, in the same way that the existence of scientific evidence for climate change doesn’t change the mind of the deniers.
Still, with growing concerns about technological unemployment, as well as ongoing problems with long-term and youth unemployment, and the rise of increasingly insecure jobs, a universal basic income (UBI) presents itself as an efficient way of redistributing national income outside the usual systems of the labour market. Indeed, many on the right champion a version of it precisely because they see it as a way of simplifying welfare payments, thus saving money and reducing the need for government bureaucracy. Around the world a number of trial schemes are in place, and Switzerland even put the idea to a national referendum (where it was rejected).
My inclination is that a basic income should meet three criteria: that it is unconditional, universal and sufficient. But I’m willing to admit that this form may not meet with general approval, so let’s continue our thought experiment and see what alternatives there might be. Let’s imagine a basic income scheme is in place, but people still insist that it must have conditions upon it.
Is there some way of imposing those conditions while negating the need for the expensive administration that means a huge proportion of the funds allocated for, say, the dole, are spent on policing compliance?
What if everyone who received a basic income had to download an app. And let’s imagine that the app worked similarly to the way the current game craze Pokemon Go works. In that game, the app is linked to your phone’s GPS, camera and mapping systems, and players walk around in the real world collecting Pokemons. As you walk along the street or by the river or wherever, little cartoon critters appear on your phone but they look like they are on the road in front you. You throw your little pokeball at them—again, an animation on your screen—and you capture them. The more you capture, the more points you get. Good clean wholesome fun for the whole family.
Okay, so what if instead of (or as well as) a Pokemon app we all had a Workemon app on our phones? Instead of Pokemons appearing on our GPS and map-enabled screens, what would show up would be jobs. These could be any sort of ‘work’ that someone needed done. They could even include the daily housework that needs to be done around your own home: the washing up, the laundry, the shopping, watching the kids, mowing the lawn or whatever. It could include studying or training or creating a work of art. It could include stuffing envelopes for a charity or a political party or caring for your infirm parent, your sick child, or babysitting someone else’s kid. Each task would have some sort of token value attached to it, and in order to qualify for your basic income payment, you would have to collect sufficient work tokens—Workemons—to ensure you met the conditions of your basic income payment. If you did happen to get paid work as well, well great, you would keep that money too, subject to the usual regimes of taxation that applied.
In a such a world, we would start to transform what we meant by ‘work’ and we would begin to overcome the prejudices that suggest a basic income is ‘money for nothing’. Workemon would be a much fairer way of distributing a living wage because it would better recognise all the work that we do in our lifetimes rather than just the stuff that is now considered worthy of a wage. Very few of us do nothing, but as it stands we only get paid for doing work that is recognised as a ‘job’ and that is a very blunt instrument for recognising and remunerating the stuff that keeps a society going.
A basic income has other advantages too. In a world in which employment is increasingly piecemeal and insecure, it provides continuity of income, a condition for wellbeing that cannot be overrated. So even if you were ‘between jobs’ you would have an income floor underneath you, thus eliminating the existential fear of not being able to support yourself or your family. Related to this is the fact that a basic income would also empower workers. Under our market-driven system, paid employment, especially unskilled work, is often offered on a take-it-or-leave-it basis because employers know people have no choice. The existence of a basic income would give workers that choice. As economist Steve Randy Waldman says, sure, people would be pickier about what jobs to take, but ‘that’s just another way of saying that workers would have greater bargaining power in negotiating employment, as their next best alternative would not be destitution’.
We often talk about the importance of work in terms of it giving our lives meaning and structure, and there is undoubtedly some truth in this. However, that meaning and structure are not necessarily intrinsic to paid employment: they can be a side effect of a system that just happens to reward some forms of work and not others. Meaning and structure arise from the remuneration and social recognition as much as they do from the work itself. Having a job is meaningful and satisfying because it allows us to keep ourselves alive because we are paid to do it. Under a basic income, where essentially all work—inside and outside formal employment—is recognised and rewarded, I have no doubt that our conception of ‘meaningful work’ would shift and broaden. We fear that without work our lives would lose meaning, but if the means of survival were taken out of the equation by a basic income, I have no doubt that this fear would be diminished and we would find meaning elsewhere.
Let’s remember why we are even talking about this and try to sum up. We began with the anecdote of FoF rendering himself redundant via technology, but that is just one example of a much bigger and clearly discernible trend. If we really are heading towards a future where ongoing employment that can sustain us at a reasonable standard of living across a lifetime is likely to become increasingly rare, then it is such a monumental change that we will barely be able to take it in. We need to start conducting these sorts of thought experiments if we are going to have any chance of creating a politics capable of responding to the changes that are coming.
So let me be blunt. The point of capitalism is to destroy jobs, not create them. Machines are not only cheaper to run than humans, they argue less and don’t make demands. Unemployment has always been a tool used to discipline workers and ensure they don’t get too many ideas about sharing the spoils. Automation is that process writ large and under such circumstances the promise that the new technologies will create as many jobs as they destroy is exposed as the politically naive, ahistorical wishful thinking that it is. Once you realise that, you realise it is only a matter of time until we are all workless and that we are going to have to come up with some other way than a job in order to support ourselves and our families.
It’s like climate change, folks. We could go on for the next 20 years having fruitless arguments about whether or not technological unemployment is a thing, arguing the toss forever about whether or not the robots are going to take our jobs, and then wake up one day and realise we have left our run too late. Alternatively, we could get out ahead of the predictable surprise and demand that the coming changes be enacted on behalf of the many and not just the few.
Welcome to the world of postwork, and what’s that video running in the background? Is it the Jetsons or is it the Terminator?
- See <http://www.abc.net.au/news/2016-03-22/ato-30pc-of-large-private-companies-pay-no-tax/
7266454>. - See <https://www.washingtonpost.com/news/innovations/wp/2016/05/16/meet-ross-the-newly-hired-
legal-robot/>. - See <https://robertreich.org/post98668011635>.
- See <http://t.umblr.com/redirect z=http%3A%2F%2Fwww.cbo.gov%2Fsites%2Fdefault%2Ffiles%2Fcbofiles%2Fattachments%2F43707-SlowRecovery.pdf+%2C&t=ZjlmOWZjOTAzNDAxOGExNTE2NzBjMDlkNjU3NWQ4MmM2ODJjNmFmZCxrUjNXdGM1Ug%3D%3D>.
- See <http//www.livescience.com/43296-what-is-stem-education.html>.
- See <huffingtonpost.com/john-tarnoff/stem-to-steam-recognizing_b_756519.html>.
- Richard Susskind and Daniel Susskind, The Future of the Professions: How technology will transform the work of human experts, Oxford University Press, Oxford and New York, 2015, p. 1.
