Poof! Work's Disappearing Act is a Cheap Trick
Despite promises to save time and energy, most labor-saving innovations transfer work rather than eliminate it.
I love a self-checkout.
It eliminates the anxiety of potential small talk. It can save some time. It's fun to scan things!
Of course, companies love self-checkout for another reason. For every customer that opts to check themselves out, that's one fewer customer that a paid cashier needs to check out. As more people opt for self-checkout, a store schedules fewer cashiers. As they schedule fewer cashiers, they can reduce the number of cashier positions they hire for.
At my local Target, there is typically only one checkout lane staffed by a cashier anymore. And there is often a line of five to eight people deep for the four self-checkout stations. Whatever amount of shrink that Target sees at the self-checkout, it must be offset by cuts to payroll.
Cashier work is eliminated little by little until only the barest minimum survives—for now.
From Paid Work to Unpaid Work
Of course, cashier work isn't actually eliminated. Paid cashier work is eliminated. But it's transferred onto the customer, who becomes an unpaid cashier.
Self-checkout usage has increased by 53% in the last 5 years, according to research by a firm that studies retail customer behavior. And by the end of 2022, 48% of all cash registers in the US were self-checkout. Some estimates find that almost 40% of transactions are done by self-checkout today.
Companies have all sorts of messaging around self-checkout: convenience, safety, less waiting, lower prices. But at the end of the day, the move to self-checkout is about profit. Fewer paid cashiers mean lower payroll.
This trend toward moving work from paid staff to customers began in earnest in the 1970s. We get ATMs instead of bank tellers, self-serve gas instead of pump attendants, and eventually, apps instead of travel agents. But as Helen Hester and Nick Srincek argue in their new book, After Work, all of this "automation" isn't really eliminating any work:
Often mistaken for ‘automation’, these technologies are in fact labour-displacement machines used to shift work from a waged worker to an unwaged user.
Hester and Srincek look closely at the "automation" trend in reproductive labor—cleaning, cooking, child-rearing, education, etc.
Before the invention of the modern washing machine, for instance, washing clothes was a time-intensive and physically labor-intensive task that included fetching water and scrubbing clothes on a washboard. Clothes washing was such an onerous task that it was not a frequent occurrence. People had fewer clothes and went much longer between washings. Households that could afford to outsource this task did—most often to poor Black women.
The home washing machine and dryer promised to save money, time, and physical labor. But, as Hester and Srincek point out, it didn't save money, time, and physical labor so much as transfer and reconfigure that labor. The in-house machines also served to dramatically increase expectations. Instead of washing clothes once a week, clothes washing became a daily task for many families. And again, what had been paid labor for some became unpaid labor.
After Work details many similar shifts, describing how one social or technological innovation that promised to eliminate work, in fact, just transferred or reconfigured it. But let's move out of the realm of reproductive labor and into the realm of (cringe) productive labor.
Do More In Less Time!
Time-saving apps promise to help us automate, eliminate, or optimize our work. From project management apps to design apps to social media apps, we're bombarded by the same value propositions remixed to fit their target market. But do these apps actually deliver? Is there a real time savings? Or do these apps merely transfer or reconfigure work in a way that leads to profit for some and more work for others?
The boom in artificial intelligence apps promises to save us time. And while I won't dismiss the possibility of some time savings, I can't help but notice the work these apps create, too.
First, these apps don't seem to work well yet without a very specific type of instruction. So people who use them are investing time and energy in learning how to better "prompt" an AI app to deliver what they're looking for. Is it really easier or more efficient to learn how to prompt an AI than it is to learn some basic copywriting or research skills? Regardless, "how to prompt" is a market opportunity. Even if substantial time savings can be achieved, many people are spending money on products that promise to teach how to write better prompts.
Second, to use the product of an AI app, there needs to be another step in the work. Facts must be checked. Details must be massaged. The tone must be adjusted. Assuming that one cares about those things, those tasks are work that should be accounted for.
Finally, these apps are both the product and the creator of heightened expectations. The proliferation of social media platforms created a demand for more and more social media content. Creators responded by sinking more and more hours into making posts. AI apps seem to help alleviate this demand by generating gobs of content on demand. However, fast content production only serves to increase the expectation that creators will post gobs of content.
Where Does the Work Go?
If work is rarely eliminated and instead transferred to something or someone else, where does that work go?
The case of AI provides one set of answers. On one hand, work goes “down” the labor value chain. When so-called higher-skill work is shunted off to the AI, then some work is transferred to so-called lower-skill workers. Instead of hiring a copywriter at, say, $80 per hour, you hire a virtual assistant for $25 (or $5) per hour. The labor gets cheaper—but it doesn’t disappear.
On the other hand, work can also go “up” the labor value chain.
recently told me that she’s putting In loads of unexpected work sifting through applications for an open role. The cause? Automated applications. Job seekers today have a variety of choices for automating the process of applying for roles. These applicants may not have to spend as much time combing through job ads and filling out applications, but employers like Kate put in extra time on the other end.And yes, I'll be writing more on this particular subject!
In addition to work moving up or down the wage ladder, some work that appears to be eliminated is merely substituted for other work. Here, we can look at the boom in independent online courses. An expert or coach who typically works with people one-on-one wants to “scale” their business to earn more money with less work. So they decide to build an online course.
They transfer some of the work time they’d spend with a client to creating the course. However, once the course is created, they transfer additional work time to marketing, selling, teaching, and maintaining the course. They may or may not be making more money as a result, but they’re unlikely to be working much less than they were before. The work wasn’t eliminated. It was merely reconfigured into a different set of tasks—which may very well be tasks that aren't what that business owner wants to be doing.
Further, they've transferred work that was once their responsibility to the customer's responsibility. The course creator can serve more customers in less time, but the customers are doing more work on their own to get (hopefully) the same benefits.
"Labor-Displacement Machines"
As Hester and Srincek note, technology rarely eliminates work but, instead, displaces it. And most often, the displacement creates the opportunity for wealth accumulation: marketing opportunities, product opportunities, payroll reduction opportunities, etc. When wealth accumulation is the goal, we almost always get negative externalities that fly under the radar.
It's pretty easy to see this when it comes to things we traditionally think of as technology (e.g., new inventions or apps). But social and economic technologies also act as labor-displacement machines.
The nuclear family is a social technology that displaced collective reproductive labor and replaced it with individual reproductive labor—sequestering upper and middle-class women in the home and providing justification for paying women (and immigrant and Black women especially) far less than men when they did work outside the home. Platform business models are economic technologies that transfer what was the waged work of journalists and creatives to the unwaged work of "users." Business strategy based on core competencies is an economic technology that serves to transfer labor from in-house roles to specialized contractors.
Making Better Choices About "Saving" Work
Displacing, transferring, or reconfiguring labor doesn't have to lead to negative outcomes. However, when we don't notice the way work shifts around in a system, we can't notice the entirely predictable negative outcomes of that shift. When we don't see that the task we've "eliminated" for ourselves via delegation or automation has created a different kind of task in its place, we can't weigh the true costs of our choice.
The good news is that we can anticipate how work will be transferred or reconfigured. Before making a decision to use a new "labor-saving" technology, we can figure out where that labor is actually going. Once we know where the labor is going, we can determine who is doing it and/or how it's being done. That information can help us make better decisions about our own work and how our choices may impact others.
Great point about work displacement.
I like buying on line, fare card scanners, even eyeball scanners at the airport, but I don't like self checkouts, and some aspects of the experience I hate, for example the way I'm assumed to be a thief by default.
I live in a small neighborhood on a small island. Our local supermarket is community owned. I know all the cashiers by name, and they know me. One even subscribes to my Substack. I like being asked in person if I'd like to round up to support this month's local charity. Lonliness is an epidemic, and for some of my neighbors these might be the only humans they talk to for days at a time. That was certainly the case for my father. I'm an introvert, neurodivergent, and I hate small talk. But I like a human at the checkout.
Great explanation for why so many professors don’t shift to independent online courses and why many life coaches / therapists don’t end up marketing themselves.