Economists are now predicting the rise of the robots will impoverish us. Can that be true?

I’ve just read a frightening new working paper by authors including Jeffrey Sachs, outlining scenarios under which the rise of robot competition could actually make people worse off. 

Sachs – who made history by becoming a Harvard Professor at age 28 – is a heavyweight in the field of economic development, so it’s worth listening when he writes “technological progress can be immiserating.”

The paper acknowledges that such predictions have been made before, and proved wrong. There were some details in the history of Ludditism that I didn’t know, particularly the role of the state in defending novel production methods.

“Concern about the downside to new technology dates at least to Ned Ludd’s destruction of two stocking frames in 1779 near Leichester, England. Ludd, a weaver, was whipped for indolence before taking revenge on the machines. Popular myth has Ludd escaping to Sherwood Forest to organize secret raids on industrial machinery, albeit with no Maid Marian. More than three decades later – in 1812, 150 armed workers – self-named Luddites – marched on a textile mill in Huddersfield, England to smash equipment. The British army promptly killed or executed 19 of their number. Later that year the British Parliament passed The Destruction of Stocking Frames, etc. Act, authorizing death for vandalizing machines. Nonetheless, Luddite rioting continued for several years, eventuating in 70 hangings.”

The model constructed by Sachs and his co-authors has no role for hangings. It simplifies the economy into a technology sector producing “goods” and a residual sector staffed by humans, producing “services.”

The model tries to answer the question:

“Will the reduction in the cost of goods produced by more advanced robots compensate workers for the lower wages?”

The team runs the models several times and gets a range of different answers depending on assumptions. But the news is certainly not all good.

“A second prediction of our model is a decline, over time, in labor’s share of national income.”

The model has ‘retention of code’ as a central feature. They argue that over time, useful code builds up so that new code is less and less necessary, leaving less and less work for people engaged in its production.

Code is defined as “not just software but, more generally, rules and instructions for generating output from capital.”

It assumes over time code becomes more durable, driving unwanted “high tech workers” to go and work in the services space, where they drive down wages.

“The price of services peaks and then declines thanks to the return of high-tech workers to the sector. This puts downward pressure on low-tech workers’ wages and, depending on the complementarity of the two inputs in producing services, low-tech workers may also see their wages fall”

The ‘retention of code’ is a key feature of the model. When the researchers ramp up the coefficient on that, the model has gloomier and gloomier predictions.

The mechanism by which this works is because each more poorly compensated generation can add less and less to the economy’s capital stock:

“The long run in such cases is no techno-utopia. Yes, code is abundant. But capital is dear. And yes, everyone is fully employed. But no one is earning very much. Consequently, there is too little capacity to buy one of the two things, in addition to current consumption, that today’s smart machines (our model’s non-human dependent good production process) produce, namely next period’s capital stock. In short, when smart machines replace people, they eventually bite the hands of those that finance them.”

But is code different to any stock of knowledge? Humans have invented designs for thousands of perfectly functional cars, yet there’s work being done on inventing new and better ones at a fantastic rate. Computer code may accumulate, but “rules and instructions for generating output from capital” sounds like management. I don’t see managers being replaced by computers soon.

The model also has no room in it for the rapid expansion of the service sector. I’ve written about this before, and I think it is a central to an economy operated by the fanciful and idiosyncratic species we call humans. If our needs are met cheaply, we will invent new things to want.

Nevertheless, the paper adds to the rich debate over what might happen in an economy where humans are not directly engaged in the tasks most important for their survival.

I’ll leave you with the working paper’s dystopian predictions:

“Will smart machines, which are rapidly replacing workers in a wide range of jobs, produce economic misery or prosperity? Our two-period, OLG model admits both outcomes. But it does firmly predict three things – a long-run decline in labor share of income (which appears underway in OECD members), techbooms followed by tech-busts, and a growing dependency of current output on past software investment.”

“Our simple model illustrates the range of things that smart machines can do for us and to us. Its central message is disturbing. Absent appropriate fiscal policy that redistributes from winners to losers, smart machines can mean long-term misery for all.”

Time to start getting ready for when the robots take our jobs.

When the federal Department of Industry starts investigating when robots will be taking our jobs, you know the possibility has gone from remote to real.

A lot of jobs are at risk – half a million, according to the article – and they’re not “bad” jobs.

“The challenges presented by more automation are not limited to low-skilled positions, as robots are increasingly replicating the tasks of medium and high-skilled workers.”

job automation
Source: Department of Industry

I’ve written before about what will eventually happen to employment after the Robot Revolution. I think new skill-sets will rise to the top: people-skills and creative skills. The inspiration angle and the emotion angle will be our edge when robots are doing the physical and routine thinking work.

In the mean time, automation will make things cheaper. More and more goods will be like water.


Water is cheap. So cheap we don’t even think about it. Water is plentiful. You can easily get more than you could ever use.

Its abundance makes it easy to forget that it is incredibly important. And many other goods and services are much like water. Energy, definitely. You no longer need to pay a fortune for firewood and paraffin. Electricity comes into the house at far less than our willingness to pay.

You could argue clothing and food have already gone that way too. At certain very popular stores, you could buy a complete outfit, including shoes, for under $40. You can meet your daily energy needs for a couple of dollars.

We don’t talk much about how awesome this is. But it is incredible. It’s why absolute poverty doesn’t exist in the same way any more. Being poor is still a huge disadvantage, but has more to do with access to other needs like healthcare and housing and opportunity and with the challenges that poses to decision making, than simple starvation.

When people complain, saying things becoming cheap strips them of their value, they don’t realise the alternative.

Indulge me for a moment longer, let’s imagine prices going the other way – from free to expensive – and instead of using water as our example, let’s use air.

Air is abundant and cheap, and we barely think about it. Should we charge for it? Would that make people value it? Charging for air would be great for GDP. The whole population would be customers of the various air providers. Maybe they’d pre-pay, maybe they’d be on a plan (don’t go over your cap!). The government would set up a means-tested scheme to provide free air for certain groups. Still, it would be a big bump to the economy, and there’d be a lot of jobs in it. Jobs! Given how politicians love to promise jobs, I’m surprised charging for air isn’t on their radar.

So I hope I’ve convinced you, via these examples, that cheaper goods and fewer jobs is not necessarily bad. It can be good, in part.

But it will not be uniformly good for social outcomes. Here’s how the Department of Industry sees it.

“The comparative advantages of being human — the ability to solve problems intuitively, improvise spontaneously and act creatively — as well as the unlimited needs and wants of humans suggest that the displacement of jobs due to automation is unlikely to be long term.”

Note their use of the words “long term.”

The advance of the robots will not be uniform, and there will be times when lots of people get put out of work all at once. At these times, the returns to capital will be higher, and the returns to labour will be lower. In these times, panic will rise. Even though a future where humans are all out of work is laughably implausible, it might not seem that way if you and everyone you know just got the sack.

We will need policy settings that will help at these times.

The policies could try to prevent the robots from taking the jobs, but that means forfeiting the benefits in terms of cheaper goods. It would also be incredibly hard to implement.

So what is the best way to make sure we’re ready for a bump in unemployment? What’s the best way to make sure people are ready to get back into the workforce?

I’d argue there are two big things we should do:

1. Invest in education now. Education protects against long-term unemployment according to the data,  probably by making people ready to re-learn. This is not the time to be making university educations more expensive. Quite the reverse. It’s also the time to be investing in making sure nobody falls through the cracks. Proper implementation of needs-based school funding, in the manner suggested by David Gonksi, would be a good way to make sure that people are adaptable when the time comes.

2. Stop starving the beast. The federal deficit is growing, and since the government has failed to implement a range of spending cuts, and is opposed to tax hikes, it will probably keep growing. In the future, we may need to exploit the federal government’s ability to provide Keynesian stimulus via the  “automatic stabilisers” that are welfare payments. Bulking up the budget is necessary, and we probably need to push up the tax-to-gdp ratio. It may seem like an economically sensitive time to be lifting taxes, but that’s not the case if you put land tax on the agenda. Unlike taxes on income and companies, land taxes are not a tax on productive activity, plus they tend to be progressive (rich pay more, poor pay less).

With these policies in place, we should be much better placed to welcome the robots as our servants, not our rivals.