Robots Aren’t the Solution to Elder Care

James Wright


Paris Marx is joined by James Wright to discuss Japan’s efforts to develop robots to care for its growing elderly population, what the government hoped to achieve with that plan, and why it hasn’t worked out as planned.


James Wright is a research associate with Turning Institute and a visiting lecturer at Queen Mary University of London. He’s also the author of Robots Won’t Save Japan. You can follow James on Twitter at @jms_wright.

Support the show

Venture capitalists aren’t funding critical analysis of the tech industry — that’s why the show relies on listener support.

Become a supporter on Patreon to ensure the show can keep promoting critical tech perspectives. That will also get you access to the Discord chat, a shoutout on the show, some stickers, and more!



Paris Marx: James, welcome to Tech Won’t Save Us!

James Wright: Thanks so much for having me on. I’m a huge fan of the podcast.

PM: Thank you — the listeners will be like: Oh my god, another person who’s a fan of the show coming on as a guest. Paris does this all the time; Paris just wants to get praise from people! But I’m having you on the show, not because you told me you’re a big fan of the show, but because you have this new book called “Robots Won’t Save Japan.” Which I think is really insightful, especially when we think about some of the narratives that we’ve had around eldercare and narratives that reemerge because this is a live issue that is certainly becoming something that is more in the public discourse in the West, in recent years. As we’re looking at trends and demographics, and concerns around having enough labor to take care of old people. All these sorts of questions that are constantly being discussed.

I thought it would be great to have you on to talk about what you learned in looking at Japan, which is this country that is ahead of everyone else on that demographic curve. To see what we can look forward to and what we can learn, of course, as we seek to respond to these issues and not be blinded by techno- solutionism as we’re trying to think of how we’re going to approach that. As I’m saying there, I want to start by getting an idea of why this is an issue in the first place. Why when, we look at Japan, but also, beyond that over here as well, do we see eldercare as such a big issue?

JW: That really gets to the heart of a lot of the framing of the techno-solutionism, like you said, in Japan, but also in the EU, or in other countries in the UK. In the case of Japan, specifically, that demographic situation and the care crisis, is always used to introduce any media article or academic work about care robots or other care technologies, so I think it’s really great to dig into it. To try and understand where that framing comes from how accurate is it and what work is it trying to do? It’s pretty well known, like you said that, Japan has one of the oldest national populations, currently, it’s about 30% of the population are aged over 65 and that’s expected to reach 40% by 2050. At the same time, the total fertility rate is well below the population replacement level and it has been for many years now.

Japan’s overall population has started to shrink with a smaller proportion of children and young people. At the same time, alongside the demographic picture, there’s already a shortage of thousands of care workers. The most recent estimate that I’ve seen is that the shortfall is expected to reach almost quarter of million people by 2025. I mean, it’s just two years away. We have this narrative that Japan’s population is old, is getting older. Their care demands are already heavy, they’ll continue to rise, and that there just aren’t enough young people to provide the required amount of care. So that becomes a logical argument like: Okay, so if that’s the case, we need some solution that doesn’t involve people and that technology, and in the case of Japan, specifically, robots could be the best solution to this crisis.

When we look at it in a bit more detail and start to unpick some of that, sometimes, simplistic narrative that the aging population just equals too many old people and not enough young people to provide care. Actually, there are specific reasons why there is this shortfall of care workers. One is that Japan introduced a new care system — Long-term Care Insurance system in 2000 — which is a universal care insurance system. So everybody over the age of 40 pays a premium, and then once they reach 65, if they require some care service, they would pay a copay of means adjusted, between 10 and 30%. The rest of the cost would be covered by the insurance system.

In some ways that was great, because it socialized the cost of eldercare. But at the same time, it was like the government saying: We’re now responsible for eldercare. What that meant is that a lot of informal carers provided less care, more people were using formal care services. More women, especially who would have been informal carers, went into work. The really interesting thing is that since the year 2000, Japan’s population has fallen, the number of people of working age has fallen, but the number of people in work has risen and part of the reason is because more women have entered the workforce, partly as a result of their care insurance system.

PM: That’s fascinating because what you’re saying is that before this system came in, the way that most elderly people will be taken care of was informally, you might have had the daughter, or the husband’s wife taking care of the parents sort of a deal. It was not something that was in the formal economy, but it was something that was happening as social relationships, familial relationships. Then this new policy comes in from the Japanese government that creates more of a public system for eldercare, which then requires workers to go into the system, requires the construction of more of these care homes, but also allows women to go into the workforce. You have more of getting them into the formal economy more so, but then that also has implications for how this all works, the need for labor in order to make this work, and I believe you say a number of women who then go into the workforce also go into the care workforce. They’re just moving formally in that way.

JW: Exactly, but not enough. I think part of the expectation was when the Long-Term Care Insurance system was implemented, that all of these women who were providing informal care would become paid care workers, but that’s not really what happened. Because although the number of paid care workers has been increasing year-on-year, every year, it hasn’t made up for that shortfall in the loss, perhaps, of some informal care. The reasons for that are that care work in Japan, like many other places, is really badly paid. It’s often paid a minimum wage level, it can be extremely physically and emotionally demanding. It’s not particularly valued in society, seen as unskilled. It’s not really been a particularly attractive, necessarily industry for people to go into.

PM: That’s really important to understand and I’m sure it won’t be a surprise to any listeners in North America or the EU to hear about how care workers are treated in the expectations in that industry. One of the things that we often see is, especially in industries that are more feminized that the labor standards, and the expectations of pay and stuff are lower, because that’s how it’s been treated in our societies. I’m wondering, before we move on to talking about how you saw these robots implemented in these these care homes, and how they were developed and things like that. Why is the Japanese government, in particular, looking at a policy framework and a policy approach that encourages robotization of eldercare, rather than, any number of other approaches to it?

JW: I guess it helps to give a very general framing of the current political situation in Japan and where it’s come from. The Liberal Democratic Party is in power and it has been for most of the post-World War II, period and it’s conservative. In recent decades, there’s been a general aversion to immigration, though there have been migrant workers coming to Japan, but in relatively small numbers compared to countries in Europe or North America. Part of this political motivation for focusing on robots has been an aversion to having migrant care workers come in to do care work, which has happened in Europe, for instance. Also at the same time, there’s always a strong industry-lobby and there’s been this idea that by adopting robotics solutions for care could boost Japan’s robotics industry create this potentially massive export market to have Japanese care robots sent around the world to other aging populations.

At the same time, potentially solving the problem of care in Japan itself. There’s this anthropologist of Japan, Jennifer Robertson, who summed it up really well. She talked about part of the policy around humanoid robots being about retro-tech, so advanced technology in the service of traditionalism. She was looking specifically at humanoid robots in Japan, but how these, apparently, very technologically sophisticated devices actually end up reinforcing very gendered, ablest, and racialized stereotypes and traditional views of the family, harking back, nostalgically, to this golden-era of Japanese post-war, economic growth driven by industrial technologies.

PM: That’s fascinating and almost makes me think of the Jetsons, the robots are in there, but everything’s still very traditional patriarchal stuff like that.

JW: Absolutely! In 2007, Shinzo Abe, during his first term as Prime Minister, the Japanese Government published a document called Innovation 25, which imagined what Japan would look like in the year 2025, with all of these high tech devices, including robots. it definitely evokes that kind of Jetsons 1950s futurism and, actually, it was later published as a manga comic book in Japan, so that people could see a future. Although I do have to admit, with COVID and with a lot of remote learning and telehealth, and everything is not that far from what has actually happened.

There’s this idea of the government looking to the future, what the future should look like, often in quite a socially conservative way, but imagining really sophisticated technologies. The idea that you could develop care robots grew in the 2010s and actually in 2009, there was a major government funded four year research project to develop an international standard, ISO 13482, for personal care robots. That was laying the foundation and saying: Okay, if we’re going to have care robots, first, we’re going to start by establishing a standard for safety. That was followed in 2013, by another four year project, which was aimed at funding the research and development and commercialization of various different types of care robots. So that was the focus of my PhD.

PM: It’s fascinating to hear because, obviously, if we have any knowledge of Japan, from the West, knowing that it has big electronics companies like Sony, or Softbank, who are making these products. We’re familiar with that history for Japan. So the idea that it’s going to make these robots for us and this is going to be part of the future of their economy, just makes sense. Theoretically, that they would be working on these things and these are the types of products that they’re going to deploy for us because they’ve had these robots for so long. Another piece that’s really interesting there is you talk about the standards, and one thing that you mentioned in the book is that there’s a debate, or a fight, around standards because this is going to be really important to define what the future of these eldercare robots might be.

Where Japan and Germany are on one side of this and want one type of standard because they are really advanced in robotics, and whatnot, and want to keep control of that market. Then on the other hand, you have China and South Korea, that are more trying to get into this market that are trying to sell more of these things. They want a completely different standard, so there’s this fight over what this is going to look like, and thus who is going to win out. Where will the Germans and Japanese protect their market share and their control over this market? Or will the Chinese and South Koreans be able to work their way in by pushing a different type of standard?

JW: Absolutely. That was very much mindset when I was starting out my fieldwork research. I spent three months with the National Institute of Advanced Industrial Science and Technology, they were administering this big robot care development project. They saw standards as really important, strategically. Actually, I mean, my book is called “Robots Won’t Save Japan,” and part of the inspiration for that title is the fact that they are not one, but two previous books in Japanese with the title, “Robots Will Save Japan. That speaks to the huge amount of confidence and this idea that that robots would solve all of Japan’s problems.

One of those books focuses on the fact that Japan has lost a lot of its market share in producing electronics to China and Southeast Asian and other East Asian countries, in part because they’ve been played by the global standards framework, the way it’s worked, the organizations. They hadn’t managed to use this system, this global system of standards, in order to protect their market share and so it will been lost. That book had highlighted why it was really important for Japan to engage, proactively, in standards diplomacy to ensure robots, specifically which Japan had a big head start and could be protected into the future as their most important future industry. So I mean, it sounds very technical, but it’s really important to my interlocutors and the people that I was doing my research with.

PM: Definitely, it’s certainly an aspect of this that you wouldn’t expect to be something that is important, but setting the standards is actually incredibly important. The Chinese in particular, have been working hard in recent years in order to set and control some more of the standards that are being set, internationally, because that will help their market share and what they expect their their companies and what have you. In the way that the West and the United States have set a lot of those standards in the past for their own benefit. I want to move on to talking about these robots in particular. The best place to start, and it’s where you look in the book, is actually where these robots are developed because you went and spoke to some of these developers who were working on creating these types of robots. What were the things you found that were driving them? What did you find surprising about going to these places where are these robots are actually beingcreated that maybe you didn’t expect to see with their development?

JW: So like I said, I spent a few months with robotics engineers and programmers at this National Institute. They weren’t actually developing robots themselves, directly, they were administering this big funding project. They were helping private companies that were actually doing the development work. So these engineers were helping these private companies in a variety of different ways, so producing tools to help them doing project management, managing stage gates, in the production process, or the development process, and also helping them to test their robots. A large robot safety testing center that had been built as part of this earlier project to certify robots to the ISO standard.

But seeing their work, I got a sense of how they were approaching care as a engineering problem, I talked about how they have this algorithmic view of care. They see care as this series of individual tasks, which could be logistical actions, like moving somebody, you move, you push somebody in a wheelchair from point A to point B, or cleaning somebody, or feeding somebody, moving bodies through space, speech actions, communicating with somebody is an action. Then trying to figure out how all of these individual tasks could be done by robots.

During the time that I was there, over three months, there was only one time where, I was aware of, that the engineers actually visited a care home and interacted with the presumed end users of these robotic devices. There was very little interaction between the engineers on the one hand, and older adults, residents of a care home, on the other. There is this danger that if you just see care as a series of individual tasks, you just have this really flattened vision, or view, of what care is. Where all of the meanings motivations values attached to care are just erased and you end up with a series of physical or speech actions, and care workers and care recipients are just reduced to two dimensional characters.

PM: I think it’s such an important observation about how this all worked, because one of the other things that you mentioned as well — beyond being at the offices of this institute — was how when some of the people who made these robots went and visited the care home they wouldn’t talk to residents, they weren’t really interested in talking to many of the care staff. I think you said, they only spoke to the male staff as well, or had a meeting with only them. Which shows a particular kind of approach to this and a particular way that they’re imagining or people who they see as important, who they need to speak to, and others who they don’t. Which is a particular problem that comes up again, and again, when we see engineering and how engineering tends to approach some of these problems and the biases in that larger profession.

JW: Absolutely and it really reminded me of an earlier work, actually, by an anthropologist, Diana Forsythe, who was researching expert systems in the US in the 1980s, 1990s. I just saw so many parallels in the way that the engineers are so disconnected from the end users. Inevitably, you have, therefore, exactly the same problems coming up, which is that the end users, it turns out, aren’t interested in using these technologies, sophisticated technologies, that are then developed without their consultation.

PM: It makes perfect sense. I want to shift to looking at what actually happens in the care homes and how these robots are actually implemented and the real impacts that they have. Before we talk about the robots themselves, maybe you can give us an idea of what it’s actually like to work in one of these care homes or what these care homes are actually like in Japan. You said that these workers are often underpaid, often women, things like that, but what else did you notice about these care homes, in particular, and what it’s like to be at one of them?

JW: That’s difficult question. The care home, which I call Sakura, it’s what I translate as a publicly funded care home, that was residential with about 60 residents and about 25 staff. It had a daycare center where people could drop their elderly relative on the ground floor for the day to do activities and have meals. Then there was a residential care home on the second and third floor of the building. In recent years, the long term care insurance system has become more and more expensive and so the government has started to try and reduce costs. They’ve started to tighten the criteria for things like entering a care home. So the people who are entering care homes like Sakura, which is the most common type of care home in Japan, have typically some degree of dementia and quite often have a disability or they’re not able to walk, us a wheelchair.Typically, it’s mainly female residents because, for a number of reasons, women tend to live longer. In Japan, quite often, it’s the wife would look after their husband, typically, when they’re older, but the husband would die and then the wife would go to the care home.

Typically, I think the average age was around 88 in the care home in Sakura. What you have is mainly older women, mainly with fairly severe dementia, with a certain level of disability and then you have care workers, who most of them were born in Japan, and generally women, although there were more male care workers quite often who’d been laid off from manufacturing jobs. A couple of younger care workers who’d gone through a vocational school, but that was quite rare. It was mainly older women was probably the main the main group of coworkers. I describe in the book an average day, and the days were quite similar, because it’s quite regimented in terms of what happens when. Meals are served at certain times, everybody eats together. I also describe in the afternoons, there’s a recreation session, there are also volunteer groups coming in and out, obviously, this is before COVID. It’s very difficult, but I hope that gives us a sense.

PM: Abolsutely, that gives us a good overview of how it actually works, especially for anyone who hasn’t been to a care home themselves, and wouldn’t actually know how it works.

JW: Just to add, sorry — another really important thing was this constant communications between care workers and residents. So even if the care worker was brushing the teeth of the resident or lifting them, because all of the lifting was still manual. In Japan, it’s quite common, still, for care workers to manually lift residents rather than using a sling or device. Through all of the corrections, there’s this constant building of relationships and through communication. Which becomes important when we move on to talking about the robots.

PM: Absolutely, one of the things that you described in the book as well is how a lot of these workers have back issues because they’re doing a lot of that lifting. They’re lifting a lot of these elderly people who can’t get up on their own or have mobility issues, as you’re saying. And so this becomes one of the many reasons that robots are proposed, as a way to help these workers, or relieve some of these labor issues. When we look at the proposals for robots themselves, you look at three in particular that were tried in this care home. Do you want to explain what those robots were, how they are supposed to work, and how they are, theoretically, supposed to fit into the day to day of the care home?

JW: So under the big care robot device development project that I was researching, they had several different kind of categories of care robots: lifting devices, mobility maids, monitoring systems, communication robots, and also toilet and bathing aides to try and cover the breadth of different care actions, care tasks. The three robots that I looked at were Hug, which was lifting robots are quite boxy looking, very heavy device, with a robot arm. You would maneuver the resident over this robot arm and press a button and the arm would raise and lift the person up. You could then wheel this Hug robot around, press another button, and the person would go down and be lowered onto a chair, wheelchair, bed, toilet, etc.

The idea was that using this device, you wouldn’t have to manually lift the resident. The second mobile was Paddle, which is a Seal-type robots it’s cute. It’s meant to be a seal pup. It has some AI software that enables it to be taught its name, so if you call it by a certain name, after a while it will respond — by crying and barking like a seal and wiggling its tail. It also has haptic sensors, so it responds to touch and it has a temperature control system, so it’s a body temperature. Also, the third robot was Pepper, which was this humanoid robot that was used to do recreational exercise sessions. So it would stand at the front of the room and talk to the residents and say: We’re going to have this exercise session and play some music and then model different upper arm exercises, upper body exercises, that residents would follow along to.

PM: Yes, because Pepper doesn’t have lower body, either [laughs]. The residents do have lower bodies, but often they’re not in use because they’d be sitting down, and it’s more difficult for them to do so. Especially for the type of people who would be in one of these homes. Pepper obviously does not have legs, just kind of base.

JW: I should have mentioned that! [both laugh]

PM: I feel like if anyone has looked into these robots before, or encountered any of them, it’s probably Pepper because Pepper has been promoted a lot — not just in Japan, but in North America and Europe as well — as one of these robots that is going to have this communicative function or do a whole range of things. One of the robots that I mentioned to you before we started recording, and that’s mentioned in your book, of course, is Robear. Which is another one that was proposed back in the 2010s. That was also supposed to be a lifting robot. I remember this was one of the examples of potential automation that was used frequently in a lot of these discussions in the mid 2010s.

As you write in the book, this really never went anywhere, despite all of the talk that there was around this being one of the potential robots that were going to automate eldercare or help to automate eldercare. What did you find when these robots were trialed in the care home because you’ve discussed to us what these robots are theoretically to do or what they’re kind of set up to do, but in practice, what were they actually doing when they were implemented into the care home?

JW: If I just go through them: so Hug, which was this lifting robot. You would think that with care workers who suffer from back pain. And at this care home where I did my research, it was like 80% of the care workers said that they had some kind of back pain. The idea was that this was the lowest hanging fruit in terms of care robots. You have a proven technology; you can prove it’s safe to use. It’s going to reduce injuries in care staff. It’s going to improve their working conditions — what could go wrong? But when they actually used this robot, almost immediately, there was such a negative reaction from members of the staff. They said, you know, this looks scary, the residents aren’t gonna like it, when they tested it out with some of the residents and started to lift them up.

A couple of people said it was uncomfortable. Other people said, it’s fine. But the coworkers said, you know, they said, it’s uncomfortable. So, you know, it’s not safe to use it, we don’t want to use it. And they also talked about how they wanted to care with their own hands, which is quite a kind of an ideal in Japan, the idea that you have this kind of touch is an important aspect of providing care. And they said, you know, using robots with older people is kind of disrespectful of them, you know, these are our elders, we should be respecting them and not just moving them around, like a forklift truck moving goods or something.

PM: That makes perfect sense because, especially you have this elderly person, you’re taking care of them, I guess that touch is particularly important, especially, when you’re in this more institutionalized environment. Where you might not get so much of that because you’re not in a in a home setting or anything. So I think that makes perfect sense. Then, as you say as well, it’s not just that this Hug robot that can be uncomfortable, or people have negative reactions to it. those negative reactions are not just among the residents, but among some of the staff as well, even just on perceptions of how people might respond to it. They don’t really like the idea of using this. But as you say, they already have non-robotic solutions that they could be using, if they wanted to things like slings and things like that, that would help with the lifting, but many of them don’t use those either. It just seems ill prepared to actually suit the realities of what’s going on in the home.

JW: Absolutely, that was a major issue. Once you dig below just the initial reactions to how it looks. It was also the fact that the spatial arrangement of the care home The fact that you couldn’t just leave Hug out, you couldn’t leave it in the corridor or in somebody’s room because somebody could trip over it. They could hurt themselves, like bumping into the edges, press the button by mistake. It always had to be wheeled to the specific person who needed to be lifted, it took some time to put them onto the machine to move them.

Then you have to move the robot back to where it’s being stored. This was a common theme with the other robots as well, it wasn’t just the case that they do that action, and then they magically go back to where they’ve installed. All of these robots entailed additional work, the care workers, which sometimes the engineers who developed the devices didn’t really view that as work. I mean, it was an invisible tasks that took additional time for care workers today.

PM: The engineers obviously weren’t going into the homes very much to see how these robots were used in practice, as you describe, they were quite detached from the realities of working in or living in a care home. That’s important as well, we know that — especially in these homes, where they have to take care of you said, about 60 residents — there is a certain amount of time that they have for particular tasks. The question is, do you dedicate that time to wheeling around a robot and setting up a robot? Or are you using that to actually spend tim,e the bit of quality time you have with that resident, that they are going to appreciate?

JW: Exactly. One of the other robots, Paddle, was seen as being the most hands-off, easy to use robot. This is the seal cub-shaped robot, which careworkers responded really positively to initially. They thought it was really cute; a lot of the residents also found it cute. They were quite happy to have it there in front of them to touch it, to stroke it. It’s very tactile, it has fake fur, but soft and warm. The first few days, there was a very positive reaction, and actually, Paddle was first developed in like, 1999, it’s been around for a really long time. The inventor of paddle has created successive versions of Paddle that have tried to improve the design. The issue came after a few days, when a couple of the residents started to develop an attachment to Paddle.

One older man would pick up Paddle and put it in front of him and figured out that there is a zip underneath so would try to unzip Paddle and remove the skin. There’s a robotic body underneath. Obviously, the care workers felt that that was dangerous, so they had to stop him from interacting with Paddle, which was fairly straightforward because he was in a wheelchair, he was in the corner of the room. They could just relocate Paddle to a different table where he couldn’t get to it. But another woman, she would talk to Paddle, with the other people on our table, and then she put it onto her lap and wheel herself back to her room, and put Paddle to bed. She would start talking to it and quite often crying, having a really emotional reaction to it. Eventually refused to eat meals or to sleep unless Paddle was there beside her.

Other studies of Paddle have, although many have been very positive, some of them have found similar very emotive reactions, which have been interpreted positively or negatively. Also attachment, emotional attachment, as with other social robots. In the case of the care home in Sakura, that was seen by care staff as been quite a negative thing. Their conclusion was that we have to carefully monitor Paddle and make sure that people have a fixed amount of time playing with Paddles, basically. Again, it seems like it would be an easy thing to just give residents, but in reality to care staff had to spend quite a lot of time and effort monitoring its use, and quite often, they just put it out of reach somewhere.

PM: I’m sure it disrupts the flow of the care home unit. You said the time is quite regimented. Then, of course, in some of the cases where people have these particular attachments to them, or don’t react to it in the way that it is programmed or designed to be reacted to, then all of a sudden that creates work to have to monitor it to have to make sure that it’s being used properly to have to make sure that residents are responding to it in the way that you would expect and then having to deal with the consequences if that’s not the case.

JW: Exactly. There was similar issues again, with Pepper, which was this humanoid robot that’s from SoftBank Robotics. Pepper, for anyone who’s not familiar with it is a white plastic and metal humanoid, I think it’s about four-foot tall and it has a touch screen on its chest and a podium dome base that it can be wheeled around on. That one was seen by the care staff as being the most robot-y a robot. I mean, it was like people’s perceptions of what a robot would look like, humanoid, high-tech looking. Having different technologies, a touchscreen, and apparently being able to have a conversation because Pepper can talk.

But again, there were problems, perhaps, because it was so sophisticated, there were even more problems in terms of Pepper, sometimes, not working correctly, always needing a Wi Fi connection, which they didn’t really have. Sometimes breaking down or not working as expected. The great marketing line for Pepper was that it would be: “The world’s first emotionally engaged robot.” It could tell your emotions by reading your facial expressions, that was one of the big selling points. Although, actually, in reality, it was almost never used, it took a lot of work for pepper to learn how to read emotions.I think you had to feed it many image, so in most commercial applications, that functionality was never used.

What the care staff typically used it for is they would wheel it to the front of the room, a big room where all of the residents would gather and it would do it spiel of running a recreation session. What the care workers found was that, actually, if they just wheeled Pepper to the fun, explain what Pepper was to the residents, which they had to do on a daily basis. The residents wouldn’t really do what Pepper was telling them to do, so they wouldn’t really do their exercise. You had to have a human staff member stand next to Pepper and copy Peppers’ actions and echo what Pepper was saying, because Pepper is really short. People at the back couldn’t really see what Pepper was doing properly. Pepper also has a very high pitched voice in Japanese, so they couldn’t really hear what Pepper was saying, properly.

There was a interesting reaction. Residents were interested, the care staff enjoyed having this robot there. But when you look, not just at pepper, but at all of the robots, you see this trend of as we’ve discussed, like these invisible tasks, extra labor that the care workers have to do. Also, cutting into, like you said, that interaction time, the time to communicate with the residents. So instead of having a conversation with the residents there instead trying to figure out: What’s wrong with pepper? How do I reboot it or moving Hug around the corridors. These were workers who are really hard pressed at times, I mean, they’re running around corridors, responding to nursing calls and doing other things. Even if it only cut into their time to do care by a few minutes per hour, that was really felt in how they would interact with the residents.

PM: One of the things that stands out with what you’re describing is that these robots are creating more work than expected, taking more time in order to operate. At the same time, they’re performing fewer functions than people were probably led to believe, or at least not doing them, as well, as was promoted or expected through the marketing materials. This is particularly important, because the idea is that these robots are going to replace some of the need for human labor in the eldercare system. Of course, what you’re seeing in your work is that that’s probably not the case.

One of the things that stood out to me was that the workers in Japan, were not so worried about the robots replacing our jobs, as we often hear the narrative being in Europe and North America. I guess, in part, because as you’re saying, there is a shortage of eldercare workers who are around there who are able to do this work. But then the other piece of that, that you described, is that while they might not be taking away jobs, they are still acting in a way to deskill and devalue some of these labor, if these robots are used and implemented in the way that they are imagined. Can you expand on that a bit and explain why that is a concern?

JW: I think it was perhaps best typify by a care worker standing next to Pepper and literally just copying what Pepper was doing., the physical actions of Pepper, repeating what Pepper was saying. I mean, it was a skilled task interacting with residents, talking to them communicate with them, it was a skilled task before they started using Pepper. After they started using pepper, they just copied what Pepper was doing, it became a manual task. With Hug, similarly, there was a skill to communicating with the residents — there was an element of touching contacts, which was seen as very important and valuable in care — that was replaced by operating the Hug machine and moving it around and storing it and cleaning it, maintaining it.

With Paddle, again, it could stand in for a communicative tasks, communicating with a resident, talking to a resident or having a conversation. Now, you could just give the resident Paddle and leave them to it. In each case, there’s definitely a case to be made that is deskilling those tasks. The devaluation comes in with the fact that these robots are really expensive, so if we’re saying that the robots can’t really replace care labor, but they can deskill it. This implies that if you have a limited pot of money for your care system and you’re going to scale up the use of these robots. It is this zero sum game of care economy, as it is, in the long-term care insurance system. Unless you want to spend a massive amount of money, you’re going to have to reduce the amount that you’re paying to human care workers.

I hypothesize that the only way that you could make these robots work at scale — if you’re going to scale this up across Japanese care homes — you’d have to have much bigger care homes with more residents and perhaps less skilled, and definitely less well paid care workers. The ironic thing is that these robots were originally intended, in part, to replace the need for migrant care work, but in reality, what it points to is a synergy of migrant care workers that perhaps don’t speak much Japanese, with robots or other devices, where they can get around the need to speak Japanese. If you can reduce care work to manual tasks that can be done by a migrant coworker that doesn’t speak Japanese, that could be paid less money.That it seems like that may be one direction that Japan is heading.

Actually, despite the anti-migrant rhetoric and policies for a long time from the LDP party in Japan. Actually, in recent years, they’ve started to deregulate a lot of migration routes to Japan. They’ve started setting targets to bring hundreds of thousands of migrant workers from China and Southeast Asia to Japan. But it seems the tide is turning in that particular debate, so this is one possible future that I suggest could play out in Japan. That’s a technology enabled migrant care system, which I suggest is not really a great solution, because there’s a strong sense that migrant care workers could be exploited in Japan.

PM: Absolutely. You lay that out really well. The concerns that come with it and what it would potentially mean. One of the things that you noted in the book was that the workers were seeing these robots being trialed, were seeing them not really working out so well, and then finding out how much they cost. They were saying: Why would we spend all this money on these robots when we could just be paid better? Which just seems natural, but of course, people don’t want to pay workers well, as we well know, as we’re still seeing, fights play out right now around that.

I would want to shift more toward the narratives that around this because there are a lot of narratives around robots being central to how we address this demographic crisis, that range all the way from corporate visions reflecting their desire for profit to more progressive visions of highly automated societies where human labor is far less necessary. We have more leisure time and all this sort of stuff. What does your research tell us about those narratives and what responding to demographic aging and an increased need for eldercare will actually require?

JW: I think it’s pretty difficult. I wouldn’t want to give a blanket answer because doing ethnographic research is really important to focus on what the technology is, in a particular context, how it’s actually used. Those narratives are really important and they helped shape their technological development and how it’s received. It’s really important to focus on the specific context. To be perfectly honest, when I started my PhD, I was quite naively optimistic about robots, I did think that they would provide a really helpful solution for many issues around care work. It was through the process of doing the actual field work, and seeing them in use, that changed my mind about these specific robots. You can point to these wider issues around deskilling for sure, but I’ve also seen other technologies that have been helpful, or potentially helpful. Things like using machine learning to predict when older people are likely to fall, with falls being a major cause of injury and death for older people. It really depends on what is the specific technology and how is it being used.

PM: I think when I see this my concern is often that we have these narratives around how robots are going to be so useful to addressing eldercare and how it’s going to require far less labor. We get distracted by that and not really recognizing the realities of what happens when these technologies get implemented, and how they don’t tend to work out as expected. Then, that leads us to not really be investing in or looking at the real solutions to the problem because we just falsely believe that we can let the technology sort this out. We don’t need to be worried about the labor aspects of this, or how we’re actually going to have enough workers to do this, or what creating a good working environment would be in order to get people to do this work that is going to be really necessary.

Thinking about these bigger questions, because the robots and the technology will address it for us. I feel like one piece where this really came through in the book, to me, was you write that in 2018 85% of Japanese businesses didn’t renew their Pepper leases and Pepper was supposed to be not just a robot that was going to be useful and eldercare, but many different commercial applications. Then once these businesses actually used that they were like: Yeah, not so much. But at that same time, as that was happening, Pepper was still being treated as this revolutionary development in Europe and North America, as these narratives were essentially being rolled out. I guess that’s where my concern is with these things.

JW: I’d absolutely agree and that is the main danger of, especially, a government buying into the hype, and actively funding it to the tune of hundreds of millions, or even over a billion, dollars as in the case of Japan with care robots. The idea that it would distract from actually investing that money in more effective and more equitable approaches to care. Also, in making very difficult, perhaps, political and economic decisions about the future of care by always imagining that robots are going to save us. I think the framing, in general, of technology as a solution, as a salvation, is not helpful for many reasons. I mean, not just distracting from the important political decisions that need to be made, but raising expectations and hype without addressing really the substantive challenges. So I agree with you.

PM: Thank you [laughs]. James, this has been a really fascinating conversation to dig into what has been going on in Japan, what these eldercare robots have actually meant for the way that people are working on and what they are likely to mean into the future. Yes, we face these demographic aging issues, crises, as some would define it. These are going to be important conversations, I’m sure that the discussion of robots and electronics is going to come up again and again, as one of the ways that we address these issues. I think it’s important that we have these conversations, that we learn from the Japanese example, and what has already been seen there, as we continue to discuss these things in Europe in North America, which is a bit further behind on that graph or curve, as we were saying earlier. Before we close off our conversation, anything else that we should know about this that we maybe didn’t get to in my questions?

JW: One thing that I would point to, and it’s interesting for me observing what’s happening in Japan, despite the fact that care robots haven’t really been widely adopted by care homes, I think the current figures probably around 10% of care homes are using some robotic device, which is well below the expectations for these development projects. But the government is really doubling down on tech solutionism. One of the major flagship R&D projects at the moment is Japan’s moonshot Research and Development Program, which they’ve invested about a billion dollars and has so many project objectives that just read like science fiction.

So for instance, making cybernetic avatar robots that overcome limitations of body, brain, space, and time by the year 2050. Also, regrowing lost limbs, developing cryosleep for interstellar travel, even brain to brain communication — even beyond Elon Musk’s wildest dreams [laughs]. I don’t know, it’s fascinating for me — beyond looking at care robots, specifically — is also to look at these other sci-fi imaginings of the future and in the case of Japan, the really fascinating thing is the growing disconnect between the rhetoric, which is just going way out, and the reality of what is actually possible and the actual technologies that are currently available. So, I don’t know what that says about what lesson can be learned from that, but it’s really fascinating.

PM: Definitely and it’s a widespread problem that is not just occurring in Japan, unfortunately, as these sci-fi visions distract us from reality in many cases. I believe you spoke to an engineer in the book who said that these ideas of eldercare automation and robots taking over are pretty much sci-fi as well, not something that is really happening in reality. James, it was fascinating to speak to you about all this to get your insight on all of it, really insightful conversation. Thank you so much!

JW: Thank you, Paris.