Technological innovation has never caused long-term unemployment, and won’t now


One of the most common objections to the accelerated pace of technological innovation is that the increasing automation of human labour will lead to mass unemployment. But if innovation is inevitable and, to all intents and purposes, unstoppable under capitalism, we should have seen unemployment rising steadily over the past few hundred years. Does the idea of technological unemployment marry up with the historical data?

The last 150 years or so has seen a number of massive technology innovations, and an almost perpetual transformation of the world of work. Thinking about this, I wondered if any of these new inventions led to seismic changes in the overall levels of employment.

The best statistics I could find on unemployment in the UK go back to 1881. Here’s the data, as charted by the website Economics Help:

A cursory look at the chart is enough to bring the technological unemployment trope under some suspicion. After all, technological innovation has, if anything, accelerated since 1881. This progress hasn’t been continuous – it happens in fits and starts – but in general, jobs that are made redundant by new technologies don’t come back. So we should see an overall upwards trend in unemployment. Instead, it’s peaks and troughs all the way, with the rate of unemployment always seeming to fall back to under 5% after times of crisis (UK unemployment currently stands at around 4%).

So, what’s causing those peaks? Disruptive new technologies, or something else?

Prime suspects: technologies which should have caused lasting unemployment

Looking back over the period between 1881 and today, here are some technologies which, if long-term technological unemployment is real, should have caused unemployment rates to spike.

Electrification of industry

In 1881, the first electricity generating stations were built by Thomas Edison in Pearl Street in Manhattan and Holborn in London. Within two years electric motors were driving manufacturing machinery. Over the decades that followed, factory owners replaced their huge, centralised, inefficient and dangerous steam engines with many, many distributed electrical motors, making factories smaller, airier and more flexible. And with the steam engines went all the many jobs that revolved around keeping thousands of coal-fired engines running.

Looking at our unemployment data, the rate of joblessness did peak in 1885, 1893, 1904 and 1908. But looking at the four decades over which the vast majority of manufacturing in Britain was electrified, the unemployment rate oscillated around the 3-4% level. While electrification rendered obsolete the firemen, steam engineers and coal deliverers working in manufacturing, it greatly augmented the possibilities of the factory, leading to many more jobs for machinists and electrical engineers as production expanded exponentially.

Mechanisation of agriculture

The reaper and the threshing machine were brought together in the form of the combine harvester in the 1880s. These machines were more productive than a team of farm labourers, but needed large teams of horses to pull them. When steam-driven tractors were introduced in the early 1900s though, the amount of muscle power required for the harvest was dramatically reduced.

However, Britain’s agricultural heydey was long over by this time – half the meat eaten in Britain came from abroad by 1900, enabled by steam ships and railways, while tropical fruits like bananas could be imported on the new refrigerator ships – and so agricultural labouring jobs were already on the wane. The added effect of new technology could easily be absorbed by the economic activity of the rapidly-growing cities.

Computerisation of the office

When I entered the world of work, every company still had a typing pool. The introduction of the personal computer, which really took off with the Apple II in 1977, changed all that. Today there are just 30,000 typists employed in the UK, according to the ONS, down from an estimated peak of over a million. And yet, there was minimal impact on the overall employment rate: recent analysis of the UK Labour Force Survey suggests that there were 57% fewer typists in 2014 than there were in 1992, and 50% fewer secretaries. Yet over the same period, the official unemployment rate fell from a peak of over 10% to around 6%.

Robotisation of manufacturing

According the international federation of robotics, there were an estimated 1.63 million industrial robots in operation worldwide, doing the work of who knows how many human beings. GKN was the first UK company to introduce robots into their manufacturing processes. In 1966, Guest, Keen & Nettlefold, based in Wales, was the largest manufacturing business in the world. Over the following decades, Britain’s automotive industry was extensively robotised and automated.

While the introduction of robots was followed by a steady rise in unemployment during the seventies, the unemployment rate began dropping again long before the process of automation was completed.

The real drivers of unemployment: boom and bust, war and peace

If it’s not technological innovation that causes large rises in unemployment, where can we lay the blame? As the provider of our chart, Economics Help, explains:

“The big changes in unemployment are related to the state of the economic cycle. The 1920s and 1930s were a time of low growth and a major recession. The great depression saw the closure of many industries, leading to high unemployment – especially in manufacturing industries. It also shows that the UK economy was struggling in the 1920s – even before the great depression.”

Now, to simply write off the downturns of 1921 and 1929 as ‘times of low growth’ might seem a bit hand wavy. The new technologies of electrification, mechanisation and mass production (via Ford’s moving assembly line) were undoubtedly partly responsible, by causing a mass realignment of the economy – the companies that failed to adapt to the new technologies died off en masse, leaving the way clear for their forward-thinking rivals.

The economist Joseph Schumpeter, following Marx, coined the term creative destruction in 1942. In his vision of capitalism, it is the innovation of entrepreneurs that sustains economic growth:

“…the process of industrial mutation […] incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one. This process of Creative Destruction is the essential fact about capitalism. It is what capitalism consists in and what every capitalist concern has got to live in.” (Capitalism, Socialism and Democracy)

But this process takes place over long timeframes. The more proximate causes of the great depression can be laid at the feet of financial speculation, over indebtedness and institutional responses to the great crash. The great depression only ended when the war effort created a huge demand for labour.

While new technologies eliminate some jobs, the overall effect of technology is to increase demand and spending power, so that many more jobs are created by innovation than are lost

Following the war, the social contract in the US and Britain was upended and a strong commitment to full employment was made possible by the post-war economic boom. (A boom itself enabled by the new mechanical technologies which businesses were able to implement, often benefiting from government subsidies or fulfilling demand generated by government through New Deal-style programs.)

Why don’t new technologies cause long-term unemployment?

In 2015, Deloitte produced a report based on UK census data stretching all the way back to 1871. The conclusion was that, while new technologies eliminate some jobs, the overall effect of technology is to increase demand and spending power, so that many more jobs are created by innovation than are lost.

The report’s authors, having crunched over 140 years’ worth of employment data, were very upbeat. They found that not only are we doing less dirty and dangerous jobs than we used to, but that we’re much more likely to be working in caring or knowledge-intensive jobs. We also have much more money to spend on luxuries and grooming, hence the rise in service industry jobs.

Of course, we have to temper this optimism slightly. While overall the picture is positive, that doesn’t help the individual whose life, at least in the short term, is turned upside down because their job has been automated away. Whole communities can be devastated for generations when their major employer shuts down or relocates. The onus then is on social policy to fill the gap left behind. As we’ve seen, new technological developments never cause an immediate spike in unemployment rates. When big crises occur, they are almost always a result of the failure of economic policy to respond to economic cycles. Technological change generally creates many more jobs than it takes away.