Technological progress boosts productivity and has made societies wealthier, but the impact of new digital technologies could be different from anything seen before. Some experts predict a future with robots and other forms of automation increasingly replacing workers, contributing to stagnant income, and worsening inequality. Yet it is difficult to pinpoint the net impact of advanced technologies on labor. There is anecdotal evidence that robotics and automation reduce manufacturing employment and wages, but evidence from the service sector remains scant. Collaborative research by APARC experts is now starting to fill this gap.

The researchers — including Karen Eggleston, APARC deputy director and director of the Asia Health Policy Program (AHPP), Yong Suk Lee, the deputy director of the Korea Program, and University of Tokyo health economist Toshiaki Iizuka, a former AHPP visiting scholar — set out to probe the impact of robots on services provided in nursing homes in Japan. Their study, one of the first investigations of service sector robots, offers an offset to the dystopian predictions of robot job replacement.

Published by the National Bureau of Economic Research, the study suggests that robot adoption has increased employment opportunities for non-regular care workers, helped mitigate the turnover problem that plagues nursing homes, and provided greater flexibility for workers. It is also published in AHPP's working paper series and is part of a broader research project by Eggleston, Lee, and Iizuka, that explores the impact of robots on nursing home care in Japan and the implications of robotic technologies adoption in aging societies.

[Subscribe to APARC's newsletters for the latest analysis from our experts.]

Japan has been on the front lines of a demographic crisis, grappling with a declining overall population, increasing proportion of seniors, and aversion to large-scale immigration. It has also been an early adopter of robots to address the shortage of care workers relative to a growing demand for long-term care services. Japan’s experience is especially instructive as more countries face aging populations, helping shed light on how demographics interact with new automation technologies.

In a VoxEU.org article, Eggleston, Lee, and Iizuka describe their study, its findings, and its implications. Examining the relationship between robot adoption and nursing home staffing in Japan, they find that robot-adopting nursing homes had between 3% and 8% more staff than their non-adopting counterparts. The increases in staffing occurred entirely among the non-regular employees. Nursing homes with robots also appeared to have higher management quality and were better able to reduce the burden on care workers. The results suggest “that the wave of technologies that inspires fear in many countries could help remedy the social and economic challenges posed by population aging in others.”

The Financial Times Magazine has recently featured the study by Eggleston, Lee, and Iizuka, calling it “groundbreaking in several ways but perhaps most clearly for setting its sights not on manufacturing but on the services sector, where robots are only just beginning to make their mark.” The great value of the study, the article notes, is that it lays the foundation for an empirical debate “on a subject that will be deluged with human emotion as robots continue their march into the services sector.”

You can also listen to a Financial Times podcast that features the new study (the segment starts at 4:52).

[Subscribe to our newsletters to stay up-to-date with the latest from our researchers.]

Lee studies the impacts of AI and robotics across multiple industries, including manufacturing, retail banking, and nursing homes. A trend he sees across most sectors following the adoption of robotics or AI is a positive increase in productivity. This has impacts for both the short-term and long-term relationships between humans and their robot coworkers, or “co-bots.” While it is true that the introduction of automation and robots initially replaces a significant number of workers in sectors such as manufacturing, over time, that impact reverses and there are job gains in many cases.

“The impact of robots often evolves over time from replacing human workers to augmenting them,” Lee explains, “and productivity gains [can] create opportunities for existing and new occupations.” This happens in a variety of ways. In some cases, the use of robotics and automation in one area frees up time, labor, and resources to employ more people in other, higher-skilled areas. In another situation, increases in productivity brought on by automation allow for greater company growth than would not have been possible otherwise. This, in turn, spurs the need to expand the workforce.

Alternatively, supplementing the labor of a small workforce with robotics and AI can also spread limited resources much farther. Lee and Eggleston’s studies of the impacts of robots on nursing home care in Japan repeatedly show that the use of robots positively increases the quality of service that oftentimes-understaffed care facilities can provide to the elderly and infirm. This can range from monitoring the physical condition of patients and reliably delivering medications to providing mental and emotional support to elderly residents through the use of robotic humanoid companions. Such innovative use of tech fills critical gaps that a human-only workforce would struggle to meet in a staffing shortage like Japan faces.

Looking to the future, Lee shares this perspective: “When the automobile was invented, we suddenly had a new demand for drivers. Now we’ll have to see if [automation] creates demand for other new occupations.” It’s an area of innovation and research he, Dr. Eggleston, and other Stanford researchers will be closely watching with their human eyes in the years to come.

Read the original article by Stanford HAI here >>

The world’s population is aging at a faster rate and in larger cohorts than ever before. In countries like Japan that have low fertility rates and high life expectancy, population aging is a risk to social sustainability. Developing policies and healthcare infrastructure to support aging populations is now critical to the social, economic, and developmental wellbeing of all nations. As the COVID-19 pandemic has repeatedly shown, accurate projections of future population health status are crucial for designing sustainable healthcare services and social security systems.

Such projections necessitate models that incorporate the diverse and dynamic associations between health, economic, and social conditions among older people. However, the currently available models – known as multistate transition microsimulation models – require high-quality panel data for calibration and meaningful estimates. Now a group of researchers, including APARC Deputy Director and Asia Health Policy Program Director Karen Eggleston, has developed an alternative method that relaxes this data requirement.

In a newly published paper in Health Economics, Eggleston and her colleagues describe their study that proposes a novel approach using more readily-available data in many countries, thus promising more accurate projections of the future health and functional status of elderly and aging populations. This alternative method uses cross‐sectional representative surveys to estimate multistate‐transition contingency tables applied to Japan's population. When combined with estimated comorbidity prevalence and death record information, this method can determine the transition probabilities of health statuses among aging cohorts.

In comparing the results of their projections against a control, Eggleston and her colleagues show that traditional static models do not always accurately forecast the prevalence of some comorbid conditions such as cancer, heart disease, and stroke. While the sample sets used to test the new methodology originate in Japan, the proposed multistate transition contingency table method has important applications for aging societies worldwide. As rapid population aging becomes a global trend, the ability to produce robust forecasts of population health and functional status to guide policy is a universal need.

Read the full paper in Health Economics.

Learn more about Eggleston’s research projects >>

This interview was originally produced by the Oliver Wyman Forum.

Coronavirus has dramatically increased the use of technology as governments, healthcare providers, and businesses tackle the pandemic and its devastation. But even before the crisis, Japan, a country long at the forefront of robot production and usage, had begun to use this technology in many of its nursing homes.

About 60 percent of the country’s nursing facilities now use robots. The proliferation of machines has had a relatively minor impact on turnover or wages of caregivers because of strong demand for care, an aging working population, and government subsidies for robot implementation, according to research by Karen Eggleston, the deputy director of the Shorenstein Asia-Pacific Research Center and director of the Asia Health Policy Program (AHPP), Yong Suk Lee, a center fellow and the deputy director of the Korea Program, and Toshi Iizuka, professor at the University of Tokyo and former visiting scholar with AHPP. Robot-adopting nursing homes, the researchers found, had between eight to 11 percent more staff than those who didn't adopt robots.

Caregiving is a physically demanding task. Staff frequently lift residents in and out of bed, and many suffer from back pain. Many of the robots deployed in Japan either help caregivers perform physical tasks or facilitate movement by the residents themselves.

The research couldn’t be timelier. Nursing homes have taken a heavy toll from the coronavirus. The disease has claimed the lives of more than 28,000 residents and workers of care facilities in the United States – approximately 35 percent of all deaths in the nation as of May 11. By contrast, Japan’s overall death toll stands at a little over 900 in early June.

Professors Eggleston and Lee discussed the implications of their research in a Zoom interview with Partha Bose, a partner at Oliver Wyman and a leader of the Oliver Wyman Forum, as well as Jilian Mincer, managing editor of the Oliver Wyman Forum, and Dan Kleinman, the Forum’s digital editor.

[Like what you're reading? Sign up for our newsletters to get content like this delivered directly to your inbox.]

From a labor economics point of view, Japan has been struggling with staffing in these care facilities. What made it much more acceptable for robots to be used in their situation versus other sectors?

Karen Eggleston: Japan has an extreme demography that it’s dealing with. The demand for long-term care is going up quite a bit while the overall population is declining. Although they're relaxing some immigration, there are issues with that. Some of the policy goals were to support robotics and to understand how it complements or substitutes for specific tasks in long-term care — to bring down back pain among care workers, for example — and to explicitly set a target for percentage of healthcare providers and long-term care clients who say it's acceptable to have a robot involved in their care.

They went into it well aware that robots weren't going to completely push out the workforce, but it's all a question of what type of tasks they can be involved with and how they can get an early read on that and start developing appropriate robots and re-engineering the care processes to meet that surge in demand.

What kinds of tasks are robots being used for in Japan right now?

Yong Suk Lee: There are these wearable transfer-aid robots that can actually help care workers lift persons and move around. There are similar types of robots that are non-wearable. And there are robots that directly assist the elderly in their care: They can use these to move around, and related to that, bathing activities, going to the bathroom, and so on.

The main type of robots are monitoring robots. They’re basically a camera system. They signal to the nurses or caregivers in an aid station if there seems to be abnormal movement so that they can actually go there — especially during the night when there's less staffing to actually go and check how the residents are doing. Those are the highest in terms of rate of adoption. And then there are those cute communication types of robots to help patients with dementia communicate with their families and caregivers.
 
What sort of facilities are using these robots? Do they tend to be urban facilities, or can they be anywhere in the country, like rural areas?

Lee: Based on preliminary results, adoption is higher in urban areas, but it's not significantly different. In China, it could differ drastically because there's a huge urban-rural divide in China of public health systems or public service in general. Robots are capital intensive. In South Korea, there are private homes that are wealthier and those could adopt new technologies earlier, but still, adoption in general is not widespread due to sufficient immigrant labor providing care. There could be an urban-rural divide for sure. Because of government subsidies in Japan, it equalizes distribution across regions.
 
Eggleston: We find more part-time or irregular nurses in urban areas. That may seem counterintuitive, but when you think about it, having that concentration of human capital in urban areas might facilitate that kind of part-time work and so on. There are differences we can see in our data between urban and rural homes, but we don't see large differences in terms of robot adoption and use.

Are there kinds of conditions that robots are better suited for than others?

Lee: In general, most robots are related to mobility issues. The biggest consequences of the elderly staying in homes are pressure ulcers on their skin because of their extended time in bed and low levels of mobility. Robots could provide a major contribution if they help residents move about and reduce pressure ulcers.

Communication robots are helpful for patients with dementia. The adoption of those isn't high compared to monitoring robots, but I think it's becoming more accepted and especially helpful for certain types of patients.
 
Do you think robots will be helpful for medical care?

Eggleston: That's the hope. For example, night monitoring reduces the probability of a severe fall which requires hospitalization and so on. There are contentious issues with nursing homes about physical restraints for patients, which are not allowed. And so, adding robots might deal with some of the outcomes. Both the producers and users of robots are hoping this will have a significant impact on the quality of care. 

As you look at how the coronavirus has affected long-term care facilities, do you wonder what the outcomes might've been had some of these facilities had robots?

Lee: Yes. What our findings indicate is that robots are not replacing workers in Japan. They're allowing firms to adopt more nurses — the skilled type of caregivers, which is an important finding directly related to the quality of care. Allowing critical personnel to actually focus more on patients. If there were certain technologies in place, caregivers could have spent their time more efficiently. I believe that's going to be a discussion going forward in the US and in many other countries that have suffered drastically. 

Eggleston: Particularly the communication and monitoring robots would help to some extent. They can save caregivers from having to go room-to-room and enable communication between people at the facility, and also with their households.

We do know that there's a potential there, and it might affect future adoption in nursing homes in the US and elsewhere. But given the huge financial hit the industry has taken as a whole, it might be a while before that plays out.

Given the coronavirus’s prevalence in nursing homes, people may be wary of taking jobs there. Can these robots be used in a recruiting capacity for nursing homes?

Lee: Certified nursing assistants in the US are not well-paid and it’s a physically demanding job. Now there's an extra concern of, "What will happen to me when I work here?" A lot of nursing homes had enormous difficulty recruiting people. They were paying extra for nurse aids, but they weren't able to recruit given the situation.

Potentially, nursing home facilities that have the capacity to adopt robots may be able to advertise this as not only being able to improve the quality of care for the residents but providing better work conditions for the caregivers. What we're finding in our research is those that adopt these robots tend to have better management practices.

What has surprised you most in your research?

Lee: We didn't find that robots replaced care workers. They’re being used to supplement the workers and maybe have better outcomes in quality of care. This is having an overall net positive effect both on jobs and productivity.