Technology Isn’t Replacing Workers but It’s Driving Income Inequality

If you ask the majority of the human population, they’ll tell you that companies are imposing automation and aim for robots to take over an awful lot of human jobs; however, robots can never be as good as humans. On the other side, if you ask companies, they’ll tell you that robots are in fact creating human jobs.
So, who is right?
No one. New research shows that technology has had an insignificant impact on replacing workers but it’s been a major driver of income inequality.

Reading Time: 4 minutes

technology replacing workers

Illustration: Milica Mijajlovic

Is technology replacing workers? 

While the tech community is widely acceptive of introducing machines to the labor market, workers are not so thrilled about this process. They’re intimidated, skeptical, and hostile. And they’re not to blame – there’s been a negative framing, claiming robots are there to take our jobs. Who would want that kind of future?

In this case, the only difference between those who are scared and those who are not is that latter have more information. To balance things out, we’ll provide you with all the relevant information on the topic so you feel less intimidated and more confident in the future of your work. 

According to the research paper “Tasks, Automation, and the Rise in U.S. Wage Inequality” published in October 2022, industrial robots (automation) have reduced US employment by only 0.5% over the last four decades, which is completely insignificant in the grand scheme of things. 

However, there are quite a few key findings we would like to point out in this article. 

But before that, let’s explain how the scholars Daron Acemoglu (Institute Professor at Massachusetts Institute of Technology) and Pascual Restrepo (Assistant Professor of Economics at Boston University) gathered this data and why it’s relevant. 

Initially published in Econometrica journal, the study analyses statistics from the US Bureau of Economic Analysis to understand how human labor was used in around 50 industries from 1987-2016, but also to which extent was machinery and software adopted in the same time span. 

Moreover, this has been the subject of research for the two authors for quite some time, so there were able to use their previously compiled data about the adoption of robots in the US from 1993-2014. 

Not only that, but they used US Census Bureau metrics to find valuable data about workers active on the labor market from 1980-2016, from 500 demographic subgroups and their employment status and inflation-adjusted hourly wages. 

Having all available information at their disposal, the two scientists were able to estimate what impact automation has had on workers, leading to some groundbreaking findings, as presented below. 

Who is the most threatened? 

The wealth gap in the 21st century is only getting bigger, and this research might uncover why, as it proves the correlation between income inequality and automation. 

These are the most important findings from the paper “Tasks, Automation, and the Rise in U.S. Wage Inequality”: 

  • Automation accounts for 50-70% of the changes between more- and less-educated workers’ income inequality from 1980-2016. 
  • As a result of automation, the income of men without a high school degree has dropped by 8.8%, whereas more-educated male workers have seen a significant increase since 1980. 
  • Even though women are generally paid less than their male colleagues for the same position, along with fewer women being in higher positions than men, automation has reduced the wages of women without a high school degree by 2.3% since 1980. 

One interesting factor that might have added to the growing income inequality, as the scholars noted, is the declining prevalence of labor unions. 

byu/antonos2000 from discussion

As mentioned, automation is the most obvious in the use of self-checkout machines, call-center systems, assembly-line technology, and other devices. Or, in the context of occupations, worker groups specialized in routine tasks have suffered the consequences of rapid automation in their industries the most. 

Some may argue that automation isn’t bad at all and that it has, in fact, improved their workflow significantly. That’s also correct. 

So, the real question is, who benefits from automation?  

According to the literature on automation and jobs, Technology disproportionately benefits highly and less-educated workers but also produces significant complementarities between high-tech tools and labor. (MIT

What impact has automation had on the workforce?   

To prove this point, one frequently cited example is a self-checkout kiosk. Namely, you packing your own bags doesn’t improve the quality of the service in the slightest. Sure, some of us prefer to avoid human contact on the days we feel especially grumpy or when we’re having a few extra bags of snacks so we don’t want to be judged by cashiers or other customers in line. 

But the truth is, cashiers don’t care about what you’re purchasing at all. You may feel like a special snowflake, but to them you’re just one of many customers they encounter. It’s pretty amazing how they manage to stay kind and chat with you, isn’t it? 

Anyway, many people believe that self-checkout kiosks are convenient but keep in mind that most customers get stuck in the process and end up asking for help. That’s why professor Acemoglu refers to this type of tools as “so-so technology” or “so-so automation”. 

Source: Massachusetts Institute of Technology

In other words, automation of this sort has large distribution but fewer productivity gains, whereas technological innovation that increases industry productivity isn’t largely distributed. 

byu/antonos2000 from discussion

What can we learn from Amazon?   

Another extreme example of an automation process would be Amazon, America’s 2nd largest employer, which takes great pride in its optimistic attitude: “Robots are not taking human jobs, they’re creating them.”  

According to them, automation helped the company to scale up and hire around a million Americans. Therefore, they’re planning to invest even more in robotic innovation, making it one of the company’s priorities. 

Moreover, they claim automation doesn’t have a negative impact on workers and that humans are undoubtedly a necessity. In a way, they’re acting as if they’ve found a perfect balance where everything runs smoothly and cost-efficiently. 

Why are they so confident? 

By the words of Amazon’s Vice President of Robotics Scott Dresser, published initially in Politico, the key is to Involve employees in maintaining and coordinating the robot systems that now handle much of the fulfillment process. 


You don't need a four-year degree to be a robotic maintenance specialist. That’s part of how we design our systems — to be simple and easy to maintain, so they're relatively easy to train someone on within a matter of a few weeks to be proficient and maybe become an entry-level maintenance technician.

Scott Dresser. Source: Politico

From the above, we can conclude that Amazon’s key takeaway is that workers should keep retraining and upskilling, no matter the industry, no matter the education level. The problem with this philosophy is, those with lower income will still see a lower increase, even if they keep ‘striving’. 

Should we try to give an alternative answer to the main question: Are robots there to take our jobs or to create them? 

It would look something like this: Robots are here to be our competition, meaning that we need to constantly adjust and improve our skillset, not only for what was our main job description a few years back, but also to be able to work around the machine, keeping up with its improvements as well. 

Now it seems about right. 

A journalist by day and a podcaster by night. She's not writing to impress but to be understood.