Pilar Pulgar, a 51-year-old supermarket cashier, feels her “legs tremble” when she sees a customer using an automated checkout. Estefanía García, 28, is “scared” by the automation at the new distribution center where she works, after years employed at a warehouse that used hardly any automation at all. José Luis, a 42-year-old sales representative, says his company regularly threatens him with the prospect of being replaced by a machine: “They convey this to us quite explicitly, to make us feel like we’re expendable.” Other workers in danger of being replaced by automation are more optimistic. “It’s hard for me to imagine how what I do could be automated,” says Beatriz Espinilla, 39, who is in charge of fitting car doors on an assembly line. “It requires very fine, detailed work,” she says. “There’s some uneasiness, but I’m not that worried,” adds Diego Martin, a 36-year-old train conductor.
Automation affects millions of workers around the world. Major developments in technology in recent years, along with new advances in technology currently under development for the future, are setting off alarms, as many worry about an imminent crash in employment rates. This has not happened yet, however, and experts who spoke with EL PAÍS believe it probably never will. History, as Mark Twain once quipped, never repeats itself, but it does often rhyme. And this is as true for labor markets as anything else. Experts predict that today’s professions will face a significant loss of specific job tasks, as has always happened during periods of major technological advancement. This, they say, will require more retraining programs for workers whose jobs become obsolete, but it will not result in an overall loss of employment opportunities.
Historically, new trades have compensated for the destruction of old ones. “This debate already took place during the first and second industrial revolutions; in various periods of major technological change, these developments have proven to be positive, both in terms of overall wealth and in terms of the net effect on labor. Hopefully, this time it will not be so different,” says Miguel Sánchez, an economist in the Research Department of the International Labour Organization (ILO), and the co-author of the report, World Employment and Social Outlook: Trends 2023.
“The eschatology of ‘the end of work’ sells, but it’s unfounded, because neither statistics nor history support such a conclusion. The evolution of society shows how, after every technological revolution, there is an accompanying increase in productivity (in both the medium and long term) through the creation of more and better jobs,” says Juan José Fernández, a professor in the Department of Private and Business Law at the University of León, in Spain. “We have been having this debate for a decade now, and even with the leap in digitalization brought about by the pandemic, the destruction of employment due to automation just has not happened,” adds Arturo Lahera, an expert on the subject and a professor in the Department of Applied Sociology at the Complutense University of Madrid.
This assessment is echoed by Federica Saliola, head of the Social Protection and Jobs Global Practice at the World Bank. “In Germany, Singapore and South Korea, where robot density [the number of operational industrial robots relative to the number of workers] is high, you see that overall employment rates also remain high. Automation goes hand in hand with innovation, and innovation creates new jobs,” she says. Stefano Scarpetta, Director of Employment and Social Affairs at the Organisation for Economic Co-operation and Development (OECD) agrees: “A lot of jobs are going to change, but we are not heading for a drop in the total number of workers.” In its 2019 report on the subject, the OECD predicted that in 10 years, 14% of current jobs will be fully automated and around 32% partially automated. The organization also claimed that this process will not entail a drop in the number of workers worldwide, which has been growing steadily in recent decades (with the exception of the pandemic period).
Another report on the future of work, published in 2020 by the World Economic Forum, notes that professions where the demand for employees is falling the most are: data recorders, secretaries, accountants and assembly-line workers. The professions that experienced the greatest loss in labor demand in the U.S. between 2007 and 2018, according to the same report, were: computer operators, secretaries and typists. Thus, according to these estimates, between 2020 and 2025, 85 million jobs will be lost to automation globally, while 97 million new jobs will be created in new trades. In terms of tasks, the World Economic Forum estimates that in 2020 around one-third were performed by machines, and that in 2025 this figure will increase to about one-half. This growth “is largely due to increased digital connectivity and the adoption of new technology, which are generating more jobs for data analysts, scientists, and artificial intelligence or digital marketing specialists,” explains Sam Grayling, a researcher at the World Economic Forum. Professor Fernández adds: “The total number [of jobs] remains the same, and is even increasing, which shows that there are plenty of new occupations that can replace those that are disappearing.”
Raquel Sebastián, senior researcher in the Department of Economic Analysis at the Complutense University of Madrid, and a specialist on the subject, says that in Spain, between 1998 and 2019, automation has caused an increase in inequality and further polarized divisions of labor: “Occupations at the bottom and top of the wage spectrum are growing, while occupations in the middle are decreasing. The spread of technology, driven by falling prices in the sector, has led to a process of routinization: the most routine jobs, and those in the middle of the wage spectrum, have been replaced by machines.” This, Sebastián warns, leads to a worsening of Spain’s already high rates of inequality: “Robots will not only displace the middle class in the distribution, but will also cause — though increasing wage dispersion – a significant growth in economic inequality.” Saliola agrees, and warns that this phenomenon is global.
Resistance to new technology
Regardless of how groundbreaking a technology is, it will always face obstacles that moderate its impact. Pulgar, for example, says most of her customers ignore the self-checkout stations: “Most people don’t want to use them. I think some people are too lazy to scan barcodes, some are elderly and don’t know how to, and others refuse to do it because they know it’s a problem for workers. In the long term, when people get used to it, as happened at gas stations where now everyone fills up their own tanks, maybe we’ll disappear.” José Luis, a telemarketer, believes the company where he works is moving towards “troubling” levels of automation: “Workers talk less and less with each other, or with managers; everything is very impersonal. And while technology does more and more tasks, we feel more and more pressure, and have fewer breaks.” Social rejection of automation is a key factor in the transportation sector as well, according to Diego Martin, the train conductor, who is also a specialist in the automation process in his sector. “In Japan, they already have the technology to automate their entire rail network, but resistance from the population’s has helped mitigating its implementation. Airplanes fly practically on their own, but it’s hard to imagine anyone wanting to get on a plane without a pilot.”
Professor Lahera holds the media responsible for the spread of catastrophic predictions regarding the impact of new technologies on employment. “Whenever a new technology appears, the media always emphasize the consequences it will have on the labor market, but fail to give the same attention to the topic when these forecasts are not fulfilled.” For example, he says, “Amazon has just announced the closure of several supermarkets with automatic checkout services, which charge you directly as you leave the store. Delivery drones are also running into more and more problems, and autonomous driving was supposed to be widely available by the end of the last decade, but I have yet to see a truck without a truck driver. But the case that I find most interesting is the metaverse: there is starting to be disinvestment, which doesn’t mean that it can’t recover, but it does mean that its implementation will be more difficult than expected.”
The latest technology to monopolize the public conversation is artificial intelligence (AI). Scarpetta, from the OECD, says that AI will have a major impact on the evolution of the labor market, but he does not anticipate a significant overall reduction in jobs: “Unlike other technological developments, it will mostly affect higher-skilled jobs, but I think that over the long run, employees with less training will experience the greatest impact.” Professor Sebastián emphasizes that AI technology requires a significant workforce. “AI requires a large number of workers to provide the information and data it uses to interpret reality,” she says. Another barrier to AI, says the economist Miguel Sánchez, is that the technology is often more expensive than maintaining real employees. “Today, in most countries, labor as a factor of production is less costly than investment in new substitute technologies.”
Training new skills
Even if automation is not gutting the labor market, that doesn’t mean there won’t be curves ahead for workers, especially for those who perform repetitive tasks, like moving boxes. “There’s a big difference between what we do in this warehouse compared to the other one I worked in,” says García, the 28-year-old who works in a highly automated distribution warehouse in Spain. “Before, we worked together, assembling the box, putting the seal on it, checking if it was all right, closing it.... Now we do that work alone. We also used to move boxes from person to person, but not anymore.” Thanks to these developments, she says there are now fewer people in her workplace dedicated to these tasks, “but there are more workers who attend to the machines, making sure everything is working correctly,” she says. The big challenge is for these workers, whose tasks are becoming less and less necessary, to learn the new skill necessary to run the machines. “Because I’m young, I see that I still have room to adapt to another sector, or to learn new skills where I am now, but if I were older, I think I would be a bit more overwhelmed by all this,” García says. Pulgar, who is 51, says she does find it more stressful: “If they automate all the checkouts in the supermarkets, the only job that’s left is stocking, which is an area where companies tend to hire more men because they’re stronger. Let’s hope that doesn’t happen.”
Sebastián believes that retraining these workers is fundamental to “counteracting the disruptive effects of automation on the Spanish labor market,” and says that “the highly differentiated effects of automation corresponding to levels of education make it necessary to create specific training programs for those with less education, so that they can acquire the new skills demanded by the labor market.” In the near future, he adds, “it will be essential for those who currently perform routine and manual tasks to be able to develop new conceptual tasks,” he adds.
This training, experts say, should not be focused only on adolescents and young people, because there are millions of adults and workers who are nearing retirement whose skills will soon become obsolete. “Technology is unpredictable, but what is certain is that workers will have to be more adaptable in the future,” says Saliola. “We have to embrace the idea of spending a lifetime learning, as workers are going to have multiple careers, not just different jobs. We also have to connect the needs of companies with training programs.” Saliola says governments will have to make “significant” investments to achieve this goal. At the same time, he defends the importance of supporting workers through this transition, especially those who need to learn new skills to find a place in the new technological ecosystem. “Governments must support people in job transitions, including by investing in education, encouraging retraining and upskilling, and providing adequate social safety nets,” Grayling adds.
Professor Lahera is committed to challenging the stereotype that people aged 45 or 50 cannot adapt to new work scenarios. “We have to design a process that does not erase their knowledge, but builds on what they already know. The problem is not that they are unable to adapt, but that companies haven’t found the right strategies. This has to change.” He says this is especially difficult in Spain, because of the country’s production landscape, which is run mainly by small and medium-sized companies “with fewer training programs than large ones.” Sebastián says there is also an intrinsic risk to these ideas: “One trap is that if we never become obsolete, we can continue working, and since there will be fewer and fewer jobs with physical costs, we open the gates to lengthening working life and increasing the retirement age.”
Espinilla, the 39-year-old car-factory worker, says that she is “delighted” to continue learning if necessary. “I think my colleagues and I are showing that we are willing to train and develop in the direction the company needs us to go. All I ask is that we be included in the change, that we be part of that future. We have always given our best and we will continue to do so. They can count on us,” she says.
Technology improves, productivity and wages don’t
Despite recent technological developments, productivity in the global economy has been stagnant for years. “The expected increases in labor productivity from investment in new technologies, which would have led to higher incomes and more leisure time, have not materialized,” says the expert from the ILO. “This is because there simply has not been an improvement in productivity. In fact, for the past decade we have been in one of the lowest periods of labor productivity growth in history. Some believe that the current technologies are simply not disruptive enough to radically transform the growth model. Others think that we are going through a phase in which the fruits of these technologies have not materialized, but that they will in the future.”
There is another parallel debate about why these technological improvements have not led to an improvement in the conditions of the working class. “Today, we see income disparities both at the corporate level, with a few companies dominating entire sectors, and at the household level, with a substantial increase in inequality in most countries,” says Sánchez. “This suggests that only a limited number of actors benefit from these new technologies.”
Lahera agrees, adding: “There is a concentration of productivity gains from these technologies in certain sectors, and the improvements really only have an impact on capital and shareholders. In this new labor scenario, the gains are not trickling down to wages, and that’s why wages have been stagnating in Europe for a decade.” He says that one decisive reason for this is the diminishing power of unions: “In the 1970s, managers made between six and 10 times as much as workers. Now they make a hundred times more.
The threat of obsolescence
- In the car factory where Beatriz Espinilla (39) works, there is a robot that transports the parts to the assembly line operators. “I’ve been here for 10 years and it was there when I got here. I know that they used to be carried by a forklift. Since I’ve been here, as far as I know, other positions that have been automated are the more dangerous ones, the ones that involved work that ends up causing injuries,” she says. The possibility of being replaced by a robot isn’t something she loses sleep over. “We’re more concerned about the semiconductor shortage crisis,” she says. Espinilla is fine with technology “helping,” but she’s hoping it can be incorporated “with balance.” “I think people are always going to have to supervise. There are so many processes, way more than it seems, in which we do detailed work that’s very difficult for a machine to replicate. I can’t imagine how a robot would fit a car door, which is what I do now. Maybe it will happen in the very distant future, but I don’t think I’ll be around to experience it,” she says.
- Despite the uncertainty that automation brings to her sector — logistics and distribution — Estefanía García (28) sees some important advantages as well. “It’s more productive this way, and there is less margin for error. We make fewer mistakes, and we exert less physical effort... Just by the fact that we don’t have to lift heavy objects, we make fewer repetitive movements, you really notice that in your body at the end of the day.” García is especially grateful that she hardly has to bend down at all anymore, since the boxes are placed at arms-height. “But of course,” she adds, “you see that with all this, there’s a need for fewer people. It’s a change that’s normal not to like. It can give you better quality of life while you’re still working there, but in the long run, it’s very possible that they’ll end up not needing you. What we used to do with 10 people, they can now do with just two. García says that the incorporation of technology is “good” and that it should be implemented right away, but at the same time, she asks “that no one be left out.” “I know it’s really hard to do this, but for example, a tax on machinery when it replaces workers, as some countries have already proposed, would be great.”
- Pilar Pulgar, 51, believes that using an automatic checkout in a supermarket is a “really cold” thing to do. “I don’t like it at all,” she says. “Besides, as a customer, it doesn’t even save you any money, so I don’t understand why you would choose that option. It’s not like it’s much faster either,” says Pulgar, who is from Madrid and has worked as a cashier for 20 years. “The first thing I thought when I saw one of those automatic checkouts was that we were going to disappear quickly, but we’re still here. I think our staff is more or less the same as before.” She criticizes the fact that this automated checkouts do not offer any personalized attention: “I don’t know, I think it shows that some companies only want profits, that they don’t care if a machine replaces a human being. In the end, big companies have the advantage of knowing that people aren’t going to go to small stores because the small stores can’t compete with their prices.” She dislikes how technological innovations that imply savings for the company do not translate into a drop in prices or an improvement in working conditions. “I’m confident that customers who like me will continue to prefer me over a machine,” she says
- Diego Martín, 36, is a train engineer and the general secretary of Spain’s rail conductors’ union (SEMAF). From his position, he knows firsthand what the chances are that trains will end up operating on their own: “This has always been something we’ve heard, but in the railroad industry, it’s very hard to make big, rapid changes.” This, he says, is due to the connections between train networks, which are managed by European regulations that change very slowly. But there are closed-circuit rail networks, where automation is making inroads: “In the Barcelona Metro there are lines that operate automatically and there will be some in Madrid soon as well,” he says. Even so, Martín is confident that trains will never operate entirely on their own: “There will always have to be personnel on the train. If it gets stranded between two stations, 50 kilometers from each one, you need someone who knows how to get it going again. When high speed rail came on in the 1990s, everyone said that in 10 years, everything would be high speed. And 30 years later there’s still a lot of conventional track, and it still carries the most passengers. It’s like the autonomous car, which seemed imminent, but look at all the legal and social problems they’re having,” he adds.
Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition