One of the most important underlying issues of the digital revolution that’s now underway is: what will happen to the jobs?
Artificial intelligence, robotics and automated decision-making, it is clear, will displace a lot of jobs – professional as well as manual. It’s cheaper and more efficient for very many tasks than using humans. Robots and algorithms don’t fall sick, seek pensions or threaten strike action. They’re getting better all the time. Five questions arise:
How many jobs will there be in future (and where)?
What kind of jobs will they be (if they are ‘jobs’ at all as we now understand them)?
What will job changes do to inequality?
What difference will there be between rich and poor countries (with their very different costs of capital and labour)?
What should be done about it? – in terms of policy; in terms of skills.
Not enough space here to answer all of these today. Some thoughts on the bigger picture, inspired by discussions at the recent IGF, this week; on the gig economy and the changing nature of work and skills next time.
How many jobs will be out there?
It’s hard to say. There’ve been some bleak estimates if you’re starting out on your career right now. One much-read study from 2013 suggested that 47% of jobs in the United States were at “high risk” of losing out to automation over the next ten or so years.
That caused a stir but it’s been widely criticised. Methodologies are unclear, American data are not representative, foresight analysis is difficult.
It’s clear, though, that lots of jobs will go as automation grows. Autonomous vehicles, if they take off, will displace drivers. Computers are replacing service workers now as robots continue to replace those who work in factories. Professional jobs – like aspects of the law and medicine, even journalism – no longer are immune.
Will there be new jobs?
The first question is whether there’ll be new ones to replace them. Optimists here argue two things.
First, they suggest, ‘history teaches us’ that innovation will create new jobs while it destroys old ones. Look what happened, they say, during the Industrial Revolution.
Second, they say, there’ll be scope for new jobs in areas that machines aren’t good at, like – and it’s usually these two they cite – ‘the caring and creative sectors’. Where empathy rather than efficiency’s what matters.
To someone trained as an historian, like me, that phrase ‘history teaches us’ rings big alarm bells.
Yes, industrial jobs were created during Europe’s Industrial Revolution while agricultural jobs declined. Yes, technological innovations do enable new things to be done that, in the past, could not be done. But that doesn’t mean the same will happen now. Why?
Will new jobs be created?
First, the pace of change today is far faster than it was during the Industrial Revolution. Old jobs are going to go much more quickly, leaving less time to adapt. Will new jobs be created at the same speed?
Second, will those new tasks that do emerge be jobs (for people) or work (for more machines)? The same logic that means machines are cheaper and more efficient at the old tasks are likely to be true of new ones too.
Third, the rate of population growth today’s much higher. Developing countries already have high levels of unemployment and underemployment. They need rapid job growth to address those current problems and the needs of the many millions leaving school each year. One-to-one replacement will not be enough.
Fourth, none of this was plain sailing back in that (European) Industrial Revolution. Many historians see it as a time of wage stagnation. Working and living conditions in the urban slums that grew up as workers left the land for factories were very poor. There were severe social tensions. The long-term gains of the Industrial Revolution were accompanied by short- and medium- term pains.
So… will new jobs be created?
Some, perhaps many, new (types of) jobs will certainly arise, but how many and how fast we simply cannot tell at present. It seems more likely, as things stand, that job creation won’t be enough to match displacement or meet the needs of growing populations. At any rate, policymakers need to plan for that scenario rather than hoping for the best.
And they need to do so in their national contexts. Some economies are more vulnerable to job displacement than others because they depend on different sectors (services, manufacturing, agriculture, oil and mining) and because of variable labour costs.
Governments and businesses will be quicker to replace workers with machines where labour costs are high (in richer countries) than where they’re low (in poorer ones), and where digital infrastructure’s more reliable than where it’s less.
What about equality and inequality?
Important issues arise here concerning (in)equality. Employment’s the main source of income, and therefore of security and welfare, for people everywhere. Incomes from jobs are unequal, thanks not just to skills but also to the balance between labour supply and demand, to the status that’s attached to different jobs, and sometimes to the type of employee (women are still often paid less than men, younger than older workers, for the same jobs).
Will job transition for the digital age change patterns of income inequality? In some ways, clearly, yes because supply (of appropriately skilled workers) and demand (for types of work machines do better) will be changing.
As clerical jobs are displaced by computers, the pay that goes with them is going to fall relative to other workers – but under-met demand for other groups (say, for computer coders in the short term, until their role’s also taken by the computers they once coded) may lead to higher wages.
On balance I’d suggest, it’s likelier that income inequality will rise than not. Optimists think that employers, including governments, will take the opportunity of efficiency savings from replacing labour to improve the quality of services and create new job opportunities that do so. It’s just as likely, though, that employers, especially governments, will bank efficiency savings through cost reductions, shedding labour rather than reskilling or upgrading.
What of the caring and creative sectors?
This is the other argument that’s put by optimists, but there are problems here as well.
First, the scope for growth’s not as big as they suggest. If traditional job numbers fall, more of those who lose their jobs will care for their dependents without pay, rather than relying on professional carers.
And creativity’s a limited recourse. Not everyone’s got the talent. More importantly, demand for creative content’s limited by human lives and lifespans. I already have access to vast reserves of music, films and books: far more than I’ll ever have the chance to hear, watch or read. The marginal value of more content, from unknown, untried artistes, is small and falling.
Second, those in caring and creative professions are among the worst-paid workers. More care workers won’t mean better-paid care workers but more people on low wages. With the exception of the biggest names, actors, musicians and writers are at the most precarious end of the employment spectrum.
And third, I wonder if this message has got through to those responsible for policy. Lots of governments are designing their skill strategies for the digital age, but they’re strategies concerned with digital skills. They want to teach kids coding rather than caring. I’ve not seen one digital skills strategy that places emphasis on these non-digital skills for future carers and creatives.
For the moment, then, I’d say we don’t know for sure what’s going to happen to jobs in the digital age, but we’d be wise to prepare for downsides as well as hope for upsides. Those downsides could include fewer jobs, lower real incomes and greater inequality. Responsible governments should be planning how to maximise employment gains and mitigate employment losses.
Next week, I’ll look at gig work and the changing nature of employment.
Image: “Trapped by Artificial Intelligence” by Lew (tomswift) Holzman used under Creative Commons license.