Death To Machines?

Robert Skidelsky

At the start of the Industrial Revolution, textile workers in the Midlands and the North of England, mainly weavers, staged a spontaneous revolt, smashing machinery and burning factories. Their complaint was that the newfangled machines were robbing them of their wages and jobs.

The rebels took their name, and inspiration, from the apocryphal Ned Ludd, supposedly an apprentice weaver who smashed two knitting frames in 1779 in a “fit of passion.” Robert Calvert wrote a ballad about him in 1985: “They said Ned Ludd was an idiot boy/ That all he could do was wreck and destroy,” the song begins. And then: “He turned to his workmates and said: ‘Death to Machines’/They tread on our future and stamp on our dreams.”

The Luddites’ rampage was at its height in 1811-12. An alarmed government sent in more troops to garrison the disturbed areas than were then available to Wellington in the Peninsular War against Napoleon. More than a hundred Luddites were hanged or transported to Australia. These measures restored peace. The machines won: the Luddites are a footnote in the history of the Industrial Revolution.

Historians tell us that the Luddites were victims of a temporary conjuncture of rising prices and falling wages that threatened them with starvation in a society with minimal welfare provision. The Luddites, however, blamed their misfortune on the machines themselves.

The new knitting frames and power looms could weave yarn into cloth much faster than the most skilled artisan weaver working in his own cottage. Caught between fixed costs (the hire and upkeep of their domestic appliances) and falling prices for their products, tens of thousands of families were doomed to become paupers.

Their plight evoked some sympathy (Lord Byron made a brilliant speech in their defense in the House of Lords); their arguments, however, did not. There could be no rejecting progress: the future lay with machine production, not with old-fashioned handicrafts. Trying to regulate trade, Adam Smith taught, was like trying to “regulate the wind.”

Thomas Paine spoke for middle-class radicalism when he said, “We know that every machine for the abridgment of labor is a blessing to the great family of which we are part.” There would, of course, be some temporary unemployment in the technologically advancing sectors; but, in the long run, machine-assisted production, by increasing the real wealth of the community, would enable full employment at higher wages.

That was the initial view of David Ricardo, the most influential economist of the nineteenth century. But in the third edition of his Principles of Political Economy (1817), he inserted a chapter on machinery that changed tack. He was now “convinced that the substitution of machines for human labor is often very injurious to the class of laborers,” that the “same cause which may increase the net revenue of the country, may at the same time render the population redundant.” As a result, “the opinion entertained by the laboring class, that the employment of machinery is frequently detrimental to their interests, is not founded on prejudice and error, but is conformable to the correct principles of political economy.”

Just consider: machinery “may render the population redundant”! A bleaker prospect is not to be found in economics. Ricardo’s orthodox followers took no notice of it, assuming it to be a rare lapse by the Master. But was it?

The pessimistic argument is as follows: If machines costing $5 an hour can produce the same amount as workers costing $10 an hour, employers have an incentive to substitute machines for labor up to the point that the costs are equal – that is, when the wages of the workers have fallen to $5 an hour. As machines become ever more productive, so wages tend to fall even more, toward zero, and the population becomes redundant.

Now, it did not work out like that. Labor’s share of GDP remained constant throughout the Industrial Age. The pessimistic argument ignored the fact that by lowering the cost of goods, machines increased workers’ real wages – enabling them to buy more – and that the rise in labor productivity enabled employers (often under pressure from trade unions) to pay more per worker. It also assumed that machines and workers were close substitutes, whereas more often than not workers could still do things that machines could not.

However, over the last 30 years, the share of wages in national income has been falling, owing to what MIT professors Erik Brynjolfsson and Andrew McAfee call the “second machine age.” Computerized technology has penetrated deeply into the service sector, taking over jobs for which the human factor and “cognitive functions” were hitherto deemed indispensable.

In retail, for example, Walmart and Amazon are prime examples of new technology driving down workers’ wages. Because computer programs and humans are close substitutes for such jobs, and given the predictable improvement in computing power, there seems to be no technical obstacle to the redundancy of workers across much of the service economy.

Yes, there will still be activities that require human skills, and these skills can be improved. But it is broadly true that the more computers can do, the less humans need to do. The prospect of the “abridgment of labor” should fill us with hope rather than foreboding. But, in our kind of society, there are no mechanisms for converting redundancy into leisure.

That brings me back to the Luddites. They claimed that because machines were cheaper than labor, their introduction would depress wages. They argued the case for skill against cheapness. The most thoughtful of them understood that consumption depends on real income, and that depressing real income destroys businesses. Above all, they understood that the solution to the problems created by machines would not be found in laissez-faire nostrums.

The Luddites were wrong on many points; but perhaps they deserve more than a footnote.

© Project Syndicate