27 October 2016

Ignore the doomsayers – automation gives us more free time, not less


It was Giles Wilkes, now of the Financial Times, who said that the New Economics Foundation should stand for “Not Economics, Frankly”. And reading today’s Guardian, I could see what he meant.

The paper carries an op-ed by Anna Coote, a senior fellow at the think-tank. She’s making the remarkable claim that as productivity rises, we will all have to work harder and harder to keep our jobs and incomes. This is as a prelude to our being told that actually, we should all be forced – yes, forced – to work fewer hours instead:

All the main UK political parties insist that the only successful economy is one that grows, preferably faster than other economies. Growth calls for greater productivity: getting more output per unit of input. The system is greedy for more resources, but workers and machines have to do more for less. More efficient processes (including more robots) reduce the amount of human input required. So those who have jobs must work harder – and longer hours – to hang on to what they’ve got and to keep the economy growing.

The reason this claim is worth highlighting is that it’s an (extreme) example of an idea that’s becoming increasingly common on the Left – the idea that the employment market is a rat race, that it will become ever more so as global competition and automation continue to bite, and that something drastic needs to be done.

The thing is, it’s just not true.

In Coote’s version of the argument, it’s the middle sentence there where things start to go wrong. Productivity is usually defined as (but doesn’t have to be) the output per unit of labour input. So rising productivity (of labour) means that for any given level of output we need less labour input – or, of course, we get more output for the same input.

That is, rising productivity makes us richer. As Paul Krugman points out, average wages in an economy are generally determined by average productivity in that economy (with a few variations and exceptions, which I’ve discussed elsewhere). Broadly speaking, the higher productivity is, the higher wages are. We have to work less in order to stand still, not harder nor longer.

Having managed to get the direction of travel wrong on the effect, a notable achievement, Coote then uses this as a leaping-off point for a standard NEF argument: that we should all be working shorter hours, because it’s nice to have more time to be human in.

They’re right, of course: we should take some portion of our greater wealth from mechanisation and productivity rises in greater leisure. The thing is, we already do – and we have been for a couple of centuries.

The usual reference here is to Keynes’s Economic Possibilities for our Grandchildren, where he tells us that by about now we’ll all need to work only 15 hours a week, spending the rest of the time basking in our miraculous prosperity.

But what Keynes missed – and the NEF misses today – is that what we’ve chosen to cut isn’t our working hours, but housework.

Today we study time use by looking at four possible uses of the passing hours. We can’t get someone else to sleep for us, so there’s personal time. Then there’s paid working hours; unpaid household working hours; and leisure, which makes up the remainder.

The big movement of the past century, roughly since Keynes wrote in 1930, has been the reduction in household working hours – largely thanks, as Ha-Joon Chang and Hans Roslin have both pointed out, to the magical wonders of the washing machine.

With that machine, we automated and roboticised the work of a household – as well as with reliable cookers, microwaves, vacuum cleaners and the rest. The result, by at least one estimate, was that time spent running a household fell from the equivalent of a full-time job to that Keynesian 15 hours a week.

What did we do with that spare time? Some went into leisure, and some of it (especially for women) into doing more paid work. The result was that female working hours rose, male working hours fell, and male and female household hours both fell.

The net effect is that working hours for all have fallen and leisure hours have risen. (There’s a partial exemption for working parents, who have diverted many of the time gains to spending more time than ever looking after their little princes and princesses.)

Given that this has been a consistent pattern for a couple of centuries as we automate things, it’s reasonable that this process will continue as the robots take ever more of our jobs from us.

At which point, another piece of economics that the NEF doesn’t get comes into play.

It is axiomatic that human wants and desires are unlimited, and the resources available to meet them scarce. If we get into a situation where this is not so, then it’ll all be very interesting – but it won’t be economics. But it allows us to see what we’ll all be doing as the robots write the newspapers and man the tills – not sitting on the dole, but something else entirely.

It’s often said that new technologies provide us with new jobs to do. But this isn’t entirely true. Instead, those new technologies do the old jobs we used to – while we go off and satisfy some other human need or desire.

The real output of the tractor, for example, was the NHS. Only once we didn’t need 95 per cent of the people growing the food could we have 10 per cent of everyone playing doctor and nurse.

And so it will be when the robots do the heavy work and the algorithms do the brain work: we’ll all go off and do something entirely different. Including taking ever more of our greater wealth as leisure.

The only possible limit to this will come if we find human desires and wants are not unlimited – that we reach a point where industry produces all we want without any need for human labour.

At which point, of course, everyone has all they want – which doesn’t sound like a problem. It’s also Marx’s precondition for the arrival of true Communism, but without the need for the Gulags this time round.

Economics, and the knowledge of it, is important. With it one can avoid the NEF’s error of thinking, the idea that we need a Fat Controllerish plan to force everyone to work shorter hours. We’ve been reaping the fruits of a better life for the centuries since capitalism started, and there’s no reason at all to think we’re going to stop.

In other words, the 20- or 30-hour working week will arrive when we decide that’s how we wish to spend our wealth. As the old economics tells us, not the new.

Tim Worstall is Senior Fellow at the Adam Smith Institute.