I work at a factory that makes disposable plastic milk bottles. If you’ve ever bought milk in Ireland, the probability the container it’s in was made in my factory is greater than 50%.
The slowest, least efficient machine in the factory makes a little over 7,500 bottles a day; the two fastest, most efficient ones comfortably surpass 110,000 a day, and can go over 120,000 on a good day. A single operator is assigned to between one and three of these machines at a time.
When I was little, milk came in cardboard cartons and was delivered by a milkman. My parents tell me that in the 70s, milk came in glass bottles, and you left your empty bottles out to be collected. Glass bottles would break occasionally, so there was still need for new ones. This process has been automated since 1903; Google wouldn’t tell me how long this process takes to make a bottle, but it seems to be at least a couple of minutes., which is quite a lot faster than the traditional way, which would take several minutes for a single bottle to be made. With plastic bottles, output is much higher, with average production speed averaging no less than on bottle every six seconds, and the best machines in my factory doing well over a bottle a second.
It’s clear that makers of plastic bottles are much more productive than makers of glass bottles, who are in turn far more productive than traditional glassblowers, all thanks to automation. The same holds true across other industries as they became increasingly automated. Weirdly, the result of this has been the exact opposite of the positive effect everyone had hoped for.
People once thought that with more automation, people would work less, as they would only need to do a fraction of the amount of work to get the same results. The idea was that a factory owner before automation might hire 80 people to work a full day each, five days a week. With automation, everything becomes more efficient, and the owner can get the same results even if each person only works one hour a day, or even an hour a week if the automation is particularly good. Isaac Asimov predicted 50 years ago that by today, most work would be in the area of attending machines, and there would be a major problem boredom from all those highly-paid mechanics working one day a week. He wasn’t alone, either – George Jetson considered working an hour a day, two days a week, to be a heavy workload.
As we all know, that’s not how things went down. Factory owners don’t pay on output or results, they pay based on time worked. So when all this automation came in, the sensible thing was to let most of the workers go and have one person make what several had done previously – all for the same pay of course, because they’re still working the same number of hours. This increased efficiency allows the owner to drive prices down, increasing their own profit.
Thus we see that increasing automation leads to unemployment. Well, OK, you might say, but these are all low-skill jobs. Surely those who are made redundant can retrain, upskill, and land better-paid, more comfortable work?
Possibly. But consider the economic cost. A young, single person could probably survive on unemployment while they retrain in a country like Norway or the UK that has a good social security safety net. However, unemployment benefits aren’t huge, so this isn’t of value to someone with a mortgage or a car loan to pay off. Likewise, someone with young children can’t afford to pay for course fees and spend several months (or years!) in classes if their partner is a stay-at-home parent.
But let’s say all those people do get qualifications. What then? As anyone who graduated university, especially after 2006, can tell you, qualification != job. Take law. Most people, upon a moment of consideration, would admit that most lawyers probably aren’t making enough to afford a huge apartment in New York, but they still make good money, right? Sure, if you can get a job – which around 44% of American law graduates can’t. Why might this be? Well, a law degree is impressive, but a law firm or a company won’t hire a lawyer if they already have or don’t need one. With so many law graduates, there simply aren’t enough jobs to go around.
It’s the same for any profession and skilled trade. There’s only room for so many engineers, so many accountants, so many electricians, so many proofreaders.
All right, but let’s say people get different qualifications. If there are only a smaller number of qualified professionals and skilled labourers, surely everyone will be able to find work, right?
No, not really. See, even if only a few people are qualified as, say, accountants, supply can still outstrip demand. To be very crude, if one million people in a country have accounting qualifications, but there are only 800,000 accounting positions available, 200,000 of those prospective accountants will be out of work. Furthermore, managers will be able to reduce accountants’ wages due to all those unemployed accountants just itching for a job and available should their own accountant prove too expensive.
But hey, maybe I’m missing the big picture here. Surely there will still be demand for high-skill jobs that those former factory workers can do? Even if it takes a few tries, everyone will be able to eventually find a vacancy where they can earn money, right?
Probably not. Remember, perfect automation would eliminate low-skill work due to all human labour being taken over by machines. Thus, if perfect automation is achieved, the net result is that companies would have to hire back everyone they make redundant in order to maintain high employment. Granted, even if everone could find new jobs in an economy with no need for low-skill labour, most formerly unskilled workers would move on to companies with more need for professionals.
But would the need for professionals rise in step with the fall in need for unskilled labour? Probably not. Remember, these jobs have been taken by machines. One team of engineers can plausibly see to the machinery needs of several companies in similar areas – companies that, prior to perfect automation, would have employed many times the number of low-skilled workers as the engineering team needs. With the reduction in employees, particularly big companies might find themselves with less need of accountants (since they have fewer wages to pay, hence fewer transactions) and lawyers (fewer employees means fewer legal issues and less risk of someone suing).