Carlos Lastres
Carlos Lastres1,586 words

Japan's Over 65s Have Smartphones. 0.3% Use Them to See a Doctor.

The Japanese government asked people over 65 in four countries what they do with a phone or a computer. One of the answers is 0.3%.

That is the share of Japanese respondents over 65 who said they use the internet for a medical consultation. In Sweden it was 22.2%. In the United States, 20.4%. In Germany, 7.2%.

Japan is the oldest country in the world. It has 0.3%.

The survey is the Cabinet Office's tenth international comparison of the lives and attitudes of older people, published inside the 2026 white paper on the ageing society. It was in the field from mid September to early November 2025, with 1,267 valid responses in Japan, 843 in the United States, 766 in Germany and 955 in Sweden.

I read it expecting a story about older people struggling with technology. That is not what is in it.

They have the phones and they use them

Asked whether they use a mobile phone or smartphone to contact family and friends, 84.1% of Japanese respondents said yes. Sweden was 91.3%, the United States 90.3%, Germany 80.9%.

Japan is ahead of Germany. It is within seven points of Sweden. And it moved: the same question in 2020 got 75.7%.

So the device is in the hand, it is charged, and it is being used every day to talk to people. Whatever is happening further down the list, it is not that Japanese seniors are offline.

Here is the rest of the list. Japan first, then the United States, Germany and Sweden.

ActivityJapanUSGermanySweden
Contact family by phone or smartphone84.1%90.3%80.9%91.3%
Look things up or shop online37.5%80.4%58.0%84.2%
Use social media16.9%61.3%45.0%68.4%
Online banking or financial transactions10.0%73.9%31.7%94.2%
Government procedures online9.2%33.7%26.1%71.4%
Online medical consultation0.3%20.4%7.2%22.2%

Read down the Japanese column. It starts at parity and falls off a cliff.

Messaging, 84.1%. Then shopping and search, 37.5%. Then banking, 10.0%. Then government, 9.2%. Then healthcare, 0.3%.

Every other country's column declines gently. Sweden goes from 91.3% to 94.2% and back to 22.2%. The United States from 90.3% to 73.9% to 20.4%. Japan's column is not a decline. It is a wall.

The wall is where the forms are

The activities are not equally hard. Messaging is one input and one button. Banking is identity verification, an account number, a transfer limit, a confirmation screen and usually a hardware token or a one time code delivered somewhere else. A government procedure adds a login system, a form with legal language and an attachment upload. A medical consultation adds an appointment slot, an insurance card, a consent flow and a video call.

So the honest reading is that the wall sits exactly where the interface gets complicated. That is unsurprising. What is surprising is that it sits there in Japan and not in Sweden, where the same people have the same eyes and the same hands and the same age.

The difference is not the user. It is what the user was handed.

I have used Japanese banking and government interfaces. Enough of them to say the following without hedging: the ones I have used are among the least usable production software I have encountered anywhere, and it is not for lack of care. They are meticulous. They are covered in explanatory text. Every regulatory requirement has been discharged visibly, in a box, with a checkbox next to it.

They were built to survive a compliance review, not to be completed by a person. Those are different design goals and they pull in opposite directions. A compliance reviewer wants every disclosure present and every consent recorded. A user wants three fields and one button. Nobody in that process was accountable for the second thing, so the first thing won, every time, at every screen.

Sweden had the same regulations, roughly. It had different accountability for the outcome.

What the demographics do to this

The same white paper carries the population figures. As of 1 October 2025, Japan's total population was 123.22 million and 36.22 million people were 65 or over. That is 29.4%.

The line I would put in front of any product team here is the next one. The 75 and over population is 21.27 million, or 17.3% of the country. The 65 to 74 population is 14.95 million, or 12.1%. The older half of the older half is now the bigger group, and it is the group where vision, fine motor control and working memory have changed most.

By 2070 the projection is that one person in 2.6 is 65 or over, and roughly one in four is 75 or over.

Put the two datasets together and the sentence comes out like this. In the country with the largest proportion of over 65s in the world, 10% of them do their banking online and 0.3% see a doctor that way, and it is not because they lack a phone.

The AI numbers say the same thing

The Ministry of Internal Affairs and Communications published its 2026 white paper on information and communications on 24 July 2026, with a four country survey fielded in January and February 2026. Generative AI use among Japanese respondents aged 60 to 69 was 33.5%. In China it was 86.5%, in the United States 51.0%, in Germany 45.2%.

Japanese 15 to 19 year olds were at 91.7%, the highest of any Japanese age band. So the country has both ends of the distribution: near universal adoption among teenagers, a third among people in their sixties, and the widest internal spread of the four countries surveyed.

There is a smaller finding in the same white paper that I have not stopped thinking about. Asked what generative AI is to them, respondents in Japan, the United States and Germany all most often said a machine or a tool. The second most common answer in Japan was a dictionary or an encyclopedia. In the United States and China, the second answer was a friend. In Germany, a counsellor.

A dictionary is something you consult when you already know what you are looking for. It does not initiate, it does not follow up, and it does not help you with the thing you did not know to ask. If that is the mental model your interface has taught people, they will use it for lookups and nothing else, which is precisely the usage pattern the Japanese data shows.

Roughly a third of them want to learn

The Cabinet Office survey asked what new knowledge or skills people over 65 would like to acquire. Around 30% in every country named digital skills, covering computers, smartphones and social media. In Sweden it was the single most common answer.

That is the finding that removes the last comfortable excuse. The demand is there. It is not materially lower in Japan than anywhere else. What differs is what happens when someone acts on it and opens the bank's app.

What I would actually do

I do not think this is solved by adding a large text mode, and I want to be blunt about that because it is the usual response. A font size toggle addresses the easiest 5% of the problem and lets everyone stop.

Four things, in the order I would do them.

  1. Test with people in their seventies before launch, on their own device, in their own home, with their own reading glasses and their own connection. Not a lab, not colleagues, not a proxy. Most teams here have never once watched this happen and it changes what they believe within twenty minutes.
  2. Count the steps to completion and cut them. Not simplify the steps. Remove them. Every field that can be derived, inferred or asked later should not be on the screen.
  3. Give the compliance text a home that is not the middle of the task. It has to be present. It does not have to be between the user and the next button.
  4. Treat abandonment as the primary metric, per step, by age band if you can get it. A flow with 90% abandonment at step four is not a support problem. It is a design defect with a location.

Japan is going to be the first developed country to find out what happens when a third of your customers are over 65 and your transactional interfaces only work for the other two thirds. Every other ageing country is ten to twenty years behind on the same curve and will buy whatever solves it.

Which is the part I would put to a Japanese board. The 0.3% is not only a social problem. Nearly a third of the country is in this group, about a third of them say they want the skills, 84% already carry the device, and they are being turned away at the confirmation screen. There is no acquisition campaign anywhere with those numbers.

Sources

← All posts