Yes, Automation Has Eliminated Jobs

Our collective amnesia about how the world used to work

A lot of people dismiss the potential effects of automation on the job market by saying, "We've been automating for over a hundred years and there are more jobs than ever," or words to that effect. They argue that we haven't eliminated any people from the labor force to date, despite all the automation and technology we have developed and deployed. This is simply the “lump of labor” fallacy, they assure us.

But that's simply not true.

We have animated away a huge portion of the labor force: people under the age of sixteen.

What people forget is that for most of human history, people under the age of sixteen were not just an important part of the labor force—they were an essential part. Child labor was critical to both the agrarian and the hunter-gatherer economies. In preindustrial times, it was all hands on deck. Everybody needed to work in some capacity to ensure that there was enough food and other resources for everyone precisely because there was very little mechanization and people lived much closer to the margins.

Obviously, children don't have the same capabilities as adults. However, their skills were put to use wherever they could. Child labor was very important on farms, for example, and on traditional family farms, children—including ones as young as age five—essentially were the labor force. Similarly, in foraging and gardening cultures studied by anthropologists, children perform all sort of daily activities from the time they are old enough to walk, even though adults still provide for them. No matter your age, everyone cooperated and there was always something to do.

Today children hardly work at all (with family farms being one of the few exceptions, along with some other minor ones that can be dismissed).

We may think of this as strange due to our modern cultural conditioning, but given the fact that children were an essential part of the labor force for all of human history, it was not considered at all strange that they would participate in the economy as it transitioned to industrial labor. In fact, it would have been more strange had they not done so. In preindustrial times, the population pyramid was heavily skewered toward young people (as it still is today in developing countries), so the idea that everyone under the age of sixteen would simply sit around all day and not work would have been seen as absurd. It would have led to catastrophe. And what would those children do in any case? They still had to eat, after all.

And yet, that's where we are today, and by far the biggest contributor was automation and mechanization.

I thought of this while reading Oded Galor's book, The Journey of Humanity where he pointed this out (all quotes from this source unless noted otherwise; emphasis mine):

[C]ontrary to popular belief, child labour, was neither an innovation of the Industrial Revolution nor a significant factor in the industrialization process. Nor, in fact, was child labour eradicated as a result of the legislation against it.

Child labour has been an intrinsic element of human societies throughout history as the challenges of a subsistence existence demanded that young children perform a plethora of back-breaking tasks, both domestic and agricultural. But when the Industrial Revolution broke out, the prevalence of the phenomenon had reached an unprecedented magnitude. Families' earnings in urban areas were barely above subsistence, and children as young as four were sent for employment in the industrial and mining sectors.

Child labour was particularly prevalent in textile factories where delicate hands were advantageous for unclogging machines. The dismal, abusive, and hazardous working conditions that children experienced over this period, along with educational deprivation, reinforced the cycle of poverty. (pp. 78-79)

While this is highly anecdotal, if you listen to the comedic American history podcast The Dollop, whenever they discuss someone from the nineteenth century, especially men, it seems like they enter the workforce in some sort of capacity between the ages of twelve and fifteen. That may be a source of humor on a comedy podcast, but the fact is, it was a reality for a lot of people in that era. It was quite common, for men at least, to be in the workforce at ages younger than sixteen, meaning that there were plenty of jobs for them to do. There aren’t anymore. As Galor notes: "Between the years 1870 and 1940, the proportion of American boys aged fourteen to fifteen who were in work dropped from 42 to 10 per cent. Similar patterns were recorded among girls and younger children." (81). Today, that proportion is basically zero.

Why? The common explanation is that we banned the practice, which of course we did:

The first effective legislation to limit child labour was passed in Britain in 1833. This Factory Act banned the employment in factories of children under the age of nine, limited the working hours of children aged nine to thirteen to nine hours per day, and prohibited night shifts for children below the age of eighteen.

In 1844, Parliament passed a new act, limiting children ages nine to thirteen to six and a half hours of work, so that they could devote three hours a day to schooling, restricted the working hours of children aged fourteen to eighteen to twelve hours per day, and imposed safety requirements in the operation and cleaning of the machines by children.

In subsequent years, Britain passed additional measures that consistently raised the minimum age of employment and compelled factory owners to pay for the education of their younger workers. (p. 80)

And while these laws were certainly important and necessary, the fact is that automation had largely eliminated most of those jobs anyway. That is why the removal of these workers from the labor force didn't really affect economic growth at all:

...rapid technological change in the course of industrialization and its impact on the demand for educated labour gradually reduced the profitability of child labour for parents as well as industrialists in two ways. First, the new machines reduced the relative productivity of children by automating the simpler tasks that children were capable of, thus magnifying the difference between the earning capacity of parents and children and reducing parental benefit from child labour.

Second, the rise in the importance of human capital in the production process induced parents to invest their children's time and energy in education rather than work, and led industrialists, keen for their workforce to be better equipped with the relevant skills, to support laws which limited and ultimately prohibited child labour. (pp. 79-80)

What happened when we removed this segment of workers from the labor force? Did the economy collapse? Were there labor shortages? Was there runaway inflation? Did work go begging?

Nope, not at all. In fact, even before the laws were passed, child labor had already been on the decline for a long time. For the first time in human history, the labor of people under sixteen labor was not required at all for the economy to function:

...while [legislation] may well have been a contributing factor, child labor was in fact in a downward slide in Britain well before this state intervention. In the British cotton industry, the proportion of workers younger than thirteen, dropped from nearly 13 percent in 1816 to 2 percent in 1835, before any significant enforcement of the new labour code. A similar trend occurred in the linen industry.

Technological advancements played an important role in phasing out child labour long before legislation did, partly because machines, like Richard Robert's self-acting spinning mule, had already reduced the need for child labour in many sectors. And while the silk industry was exempted from the restrictive child labour legislation, due to its struggle to compete with foreign producers who had access to cheaper raw material, the proportion of child workers in silk factories fell, from nearly 30 percent in 1835 to 13 percent by 1860. If this trend is representative, it is not inconceivable that, even without legislation, child labour would have diminished.

In fact, in the latter half of the nineteenth century, public funding for education relieved employers of the full burden of financing their employees' schooling, effectively decreasing the 'tax' on child labour. Yet the number of children employed in factories never returned to its turn-of-the-century level. In the period 1851-1911, the proportion of boys aged ten to fourteen employed in factories fell from approximately 36 per cent to less than 20 per cent; for girls the proportion declined from nearly 20 per cent to nearly 10 per cent.

Similar trends were recorded in most developed countries. Legislation appears to have played only a subsidiary role in these processes while the main factors that reduced the employment and the exploitation of children were the widening parental-children income gap and changing attitudes toward education. (pp. 80-81)

And what ultimately caused the widening parental-children income gap was automation. Put in economic terms, these children became zero marginal product workers at any wage factory owners could afford to pay thanks to technological progress, much like horses after the widespread adoption of the internal combustion engine. Of course, unlike horses, children could still enter the workforce—just much later as adults. And while the vaunted "service economy" means that children could theoretically once again become a larger part of the current labor force, the impetus for this is based exclusively on corporate greed and the desire for lower wages, and not on economic necessity.

Simply put, we automated away all of the jobs of people under the age of sixteen. Child labor was effectively eliminated by automation. Yet, in present-day discussions surrounding automation, you never hear this fact. Why is that? The fact that we automated away what had at one time been a significant portion of the labor force (and barred the remainder from working) with no deleterious impact on economic growth (in fact it increased) is very relevant to today’s discussions surrounding automation.

Today's economy functions just fine with virtually nobody under the age of sixteen working full time. In fact, even eighteen-year-olds aren't all that valuable to the economy anymore. The jobs that are available tend be low-paying and dangerous (which is why those laws are still necessary). Instead, people have to spend many more additional years in formal education and training just to secure a reasonably well-paying job, and the educational requirements and competition for these jobs keeps going up. In practicality, people are being held out of the workforce longer and longer—sometimes until their late-twenties—with no deleterious effects on the economy. In addition, retirement has removed old people from the other end of the labor force. This is even seen as beneficial, as it allows new entrants into the workforce. In previous eras, by contrast, people would work as long as they were able.

So, then, given that we effectively automated away all the jobs for people under the age of sixteen, why do we assume it's simply impossible to automate away the need for labor for people over the age of sixteen?

Out of the entire US population (the third largest on earth), only around half are in the labor force. Yet the modern economy functions more-or-less adequately with only half of us working—the lights are on, water is flowing from the taps, the trains and busses are running, food is on the shelves, public order is maintained, fires are put out, and injuries are tended to. Why do we simply assume that it’s impossible for this number to decrease further? (and in, fact, some have argued that a significant portion of existing jobs are bullshit) During the pandemic, we even coined a term for this phenomenon: essential workers. Doesn't that term imply that a lot of people’s work, in some sense, is already nonessential? Why do we assume that we cannot reduce this proportion of workers any further given that history? And what happens if we do?

Of course, women of childbearing age were much less represented in labor force in the past, but, even though they did not earn an income, they were hardly not working. To say so is to devalue unpaid domestic labor, something conventional economists have done for a long time. In fact, you could argue that automation had the opposite effect in this case: by automating so many domestic tasks like cleaning, washing and cooking, and outsourcing other domestic tasks to the market, it freed up women to enter the labor force in a way they could not before.

It's true that the portion of working age adults in official statistics is much higher (~70-80 per cent), however those statistics are based on the population aged 25-54. This is now considered the “prime working years.” If you are outside of that age range, and not in paid employment, it's assumed that you are not seeking a job and that unemployment is “voluntary.” Again, that is not the historical norm! This statistic implies that we've drastically reduced labor not only for people under age sixteen, but from people under age of 25 and older than age 54. Yet we are consistently told that automation had no impact on the labor force!

So, I would argue that, yes, the workforce has, in fact, already changed tremendously because of automation and technology. And yet these changes are simply ignored because they were so gradual and took place a long time ago before the majority of us were born. Creeping normality and shifting baseline syndrome strike again.

In fact, you could argue that the primary purpose of school is not really education, but to have someplace to keep children occupied during the work day while their parents go to work to earn an income. That's why universal education only came about after this portion of the workforce was effectively eliminated. Seen from this perspective, public education is really a subsidy for employers and businesses. It frees up parents from the responsibility of looking after their children so that they can go to work for wages and make money for business owners. This could be the reason why employers went from opposing universal public education to supporting it:

[I]ndustrialists were reluctant to fund the education of their potential workforce, as there was no guarantee that these workers would not take their newly acquired skills and find employment elsewhere. Indeed, in 1867, the British iron magnate James Kitson testified to an official commission that individual manufacturers were holding back on funding schools because they feared that their competition would reap the rewards...

As it became increasingly apparent that skills were necessary for the creation of an industrial society, previous concerns that the acquisition of literacy would make the working classes receptive to radical and subversive ideas were jettisoned and capitalists began lobbying governments for the public provision of that education.

Industrialists in Belgium, Britain, France, Germany, the Netherlands, and the United States became actively engaged in influencing the structure of their countries' public education systems and encouraged their leaders to amplify investment in mass education. Ultimately, national governments caved to pressure from the industrialists, and increased their expenditure on elementary-level education. (p. 75)

Galor argues that this support for education was because of the higher demands for human capital: "Put simply, industrial societies around the world, even those who resisted other aspects of Western modernity, supported the provision of public education predominantly because they realized the importance of general mass education in a dynamic technological environment, both for business owners and for the workers themselves." (pp. 74-75)

But, did they? Was enlightened self-interest by businessmen really the driver? Surely, it does not take eighteen years of full-time, eight-hours-a-day instruction to impart the necessary skills for reading, writing and arithmetic that are required by employers, either then or now. The fact that schools operate on pretty much the same schedule as businesses is surely not a coincidence. This makes the idea that it's only about education suspect. Yes, skills are important in an industrial economy, no doubt. But the structure of modern education is clearly about more than just providing basic skills. Most people forget a lot of what they learn in school anyway and learn what they need to on the job, just as they always have.

There were other reasons to adopt universal education aside from human capital. Public education was a key method of nation-building in the nineteenth century. By now, most people are aware that the American educational system was based on the Prussian model. The Prussian education system was designed to weld together the disparate German-speaking nations of the former Holy Roman empire into a single unified nation-state called Germany (Deutschland) under the aegis of Prussia. Students would be taught to read and write the standard version of the language as opposed to the regional dialects of German they spoke at home. In addition, they would be taught to obey their headmasters making them ideally suited for work in the factory, the mines, the civil service, or the military, rather than in the fields as the economy transitioned from agrarian to industrial.

So it's easy to see why American educators in the nineteenth century would be keen to adopt this model given that the United States was a collection of disparate immigrants from even more far-flung areas of Europe (and farther abroad) that needed to be stitched together into some kind of unified, coherent whole¹. This unification process was an obsession of nineteenth century American intellectuals who saw America as an experiment that might fail if some kind of cultural homogeneity could not be enforced. This is to say, early primary education was as much about nation-building as it was about education. It wasn't simply about employers wanting employees with more skills.

The Social Fallout

Given this history, some of our most recent societal changes start to make more sense.

Instead of supporting mass education, employers have been consistently withdrawing their support and starving it of funds since the neoliberal revolution. At all levels of education, from primary public schools to universities, businesses have withdrawn their prior support, including at the university level where the advanced skills and training supposedly desired by employers are imparted. Why is that?

Could it be that employers just don't need that much skilled labor anymore thanks to automation? And could it actually be to the benefit of businesses and politicians to have less educated people? How else could this massive withdrawal of support for education be explained? Could it also be that education has accomplished its nation-building purpose and is no longer needed? Perhaps, from a political standpoint, education is not even desirable anymore for elites given that a divided and fractious population cannot unite against the ruling class and is easier to control and manipulate. College achievement rates have stagnated or even started to drop—especially among men—but employers don't seem all that concerned².

Galor above alludes to the fact that employers once feared mass education would impart radical and subversive ideas in workers’ heads—euphemisms for socialism. Is it a coincidence that right-wing parties (particularly here in the United States) relentlessly attack education? That “school choice” has been used to hollow out funding for public education? That teachers unions are portrayed as enemies of the state? That universities are pilloried as places of “indoctrination?” That the elimination of the Department of Education has been a key objective of the Republican Party for years? Is it also a coincidence that more educated populations correspond to support for more equality and inclusive social policies?

A while back, I wrote a review of a book called The Fates of Nations by Paul Colinvaux, an ecologist. That book had some interesting insights about education. Colinvaux argued that more education does not necessarily produce more jobs, and, as a result, increasing education in societies that were not suitably economically developed would inevitably lead to social chaos, not growth. In his analogy, it's the equivalent of an animal producing too many offspring for its ecological niche. Using Colinvaux's terminology, the most desirable niches are shrinking, meaning that more people must necessarily occupy the less desirable ones. According to US Bureau of Labor statistics, the fastest-growing “niches” will primarily be low-paying ones (except for nurses and computer programmers).

Schools today appear to play more of a filtering role than an educational one, determining who gets to rise in the “meritocracy” and who will be left behind from a very young age³. American education is based on the "freemium" model, where the majority receive an inferior product and only those who can afford to pay get any sort of quality or attention. From this—at least in the United States—it looks like we are going back to the era where only a small minority from privileged families will be able to receive the kind of education necessary to earn a decent living while more and more of us become superfluous. Higher education costs have been almost entirely offloaded to the individual with no penalty for corporations. Economists Anne Case and Angus Deaton warned that the major differentiator between those suffering from "deaths of despair" and those who weren't was a college degree.

Simply put, the "all hands on deck" phase of the economy is gone. Thanks to globalization and the universality of spoken English, businesses are able to recruit elites from all over the world. They no longer have to rely on domestic workers to staff the professional managerial class. University education in America now is as much for foreigners as it is for native-born citizens, which is why there is no impetus to even try to make it affordable for Americans anymore (plus foreign nationals from wealthy families can afford to pay more).

This may be part of the backlash against immigration waged by far-right political parties around the world. Immigrants are portrayed as "stealing" low-wage, low-skilled jobs, which are becoming more and more of the national economy. This is portrayed as a failure to properly integrate, however, this alleged "cultural mismatch" was curiously not as much of a concern when jobs were more plentiful and education was cheap. As we saw in our review of Slouching Towards Utopia, the movement of people was actually higher in the previous era of globalization than it is today. The difference was the amount of well-paying jobs then as compared to now.

Meanwhile, the professional managerial class is depicted as a menace by populist politicians, as it is staffed by both the "winners" of the meritocracy as well as highly educated foreigners imported from abroad The resulting class and income divide allows these educated people to be portrayed as some kind of malignant, foreign entity which adheres to strange codes of conduct (i.e. “wokeness”). The “PMC” can be scapegoated to divert anger from those who pay their salaries—the globalized capitalist elite class who fund the campaigns of these so-called “populists.” Playing people against each other is the oldest trick in the book.

At both the higher and lower ends of the economy, employers complain of worker shortages even as ordinary citizens deal with failing schools, government budget cuts, crumbling infrastructure, soaring tuition costs, and more and more onerous requirements for the remaining stable and secure jobs. This is fundamentally the result of the hourglass-shaped social structure caused by neoliberal globalization, which in turn results from forces such as automation.

There is also a connection between educational requirements and lower rates of childbirth. South Korea is an extreme example, where intense competition for jobs is accompanied by some of the lowest birth rates in the world (and high suicide rates). Certainly, if jobs were as easy to come by as they were during the nineteenth and most of the twentieth century, it's hard to imagine this would be the case. It’s now expected that children be a drain on their parents’ resources until at least age 18. Only after then can they potantially earn an income and be self-sufficient, whereas in past eras—as The Dollop illustrates—it was not at all uncommon for children to start earning a living as young as age fourteen. The investment requirements for children have now become so extreme and so onerous that people simply don’t have them anymore if they have no desire to. Why compete in a game you can’t win?

Yet, rather than deal with this reality, populist politicians demand a return to the “traditional family” and wage war on family planning. Isn’t the reality of job market the fundamental issue here, and doesn't that prove that the job market has indeed changed drastically due to forces like automation?

As long as the debate is centered on the idea of all jobs disappearing overnight—an absurd proposition based more on science fiction than reality—we will never have an intelligent conversation about the real, demonstrable effects of automation on the workforce. This will lead to the kind of demagoguery we are already witnessing around the world. I think it's important to remember the history of what has already happened.

Circling back to the point I started with, isn’t it at least possible that what happened to children and teenagers during the nineteenth and early twentieth centuries thanks to automation is now happening to the rest of us? Most children and teenagers, at least, can live with their parents until they turn seventeen. The rest of us don’t have that option (although we are seeing children living with their parents longer and longer). Is that really what’s behind a lot of social and political conflict that we're seeing today? And if so, what can be done about it?

1. It was also used to destroy native cultures, as the tragic story of Indian Schools illustrates. Native children were forbidden from wearing tribal dress, speaking their native language, and were forced to assimilate, sometimes through force.

2. The Powell Memorandum written by Lewis F. Powell—considered the opening shot of the neoliberal counterrevolution—goes to great lengths to depict higher education as indoctrinating students to hate the "free market." In those days, university education in California was basically free.

3. For example, the School to Prison Pipeline.