13.8% of People Entering Japanese Prisons Are 65 or Older — But per 100,000 Population the Over-70s Run at 4.7, a Quarter of the 20–64 Rate

Share of prison admissions aged 65 and over

2005All 4.8%　Women 6.2%

2024All 13.8%　Women 21.6%

Prison admissions per 100,000 population (2024)

Ages 20–6418.7　（women 3.6）

Ages 65–699.3　（women 2.3）

Ages 70+4.7　（women 1.5）

The white paper attributes the rising share to most other age groups declining. In 2024 there were 2,049 admissions aged 65 and over, of which 1,370 were over 70, about 2.3 times the 2005 figure.

The FY2025 White Paper on Crime carries the figures on older prisoners.

In 2024 there were 2,049 prison admissions aged 65 and over, up 2.0% on the year[1]. Those over 70 numbered 1,370, about 2.3 times the 597 recorded in 2005[1].

The share of all admissions accounted for by the over-65s stood at 13.8% in 2024, down 0.4 points on the year but 9.0 points above 2005[1].

On why it rose, the white paper is explicit: the share has trended upward because most other age groups have been declining. That is a statement about the denominator, not the numerator.

Swap the denominator for population and the order reverses. Prison admissions per 100,000 population run at 18.7 for ages 20 to 64, 9.3 for 65 to 69, and 4.7 for 70 and over[2]. The over-70s enter at a quarter of the 20–64 rate.

Over the twenty years in which the share nearly tripled, admissions of the over-70s rose 2.3 times. The over-70 population grew over the same period. Divided by population, the figure still sits below every younger band.

The share is higher among women

Women aged 65 and over accounted for 337 admissions, down 0.3% on the year, and 21.6% of all female admissions[1], well above the overall 13.8% and 15.4 points higher than in 2005[1].

Women over 70 numbered 253, up 0.4%, and 4.6 times the 55 recorded in 2005[1]. Against 2.3 times for the over-70s overall, the female figure has doubled that rate of growth.

Here too the population figures matter. Female admissions per 100,000 run at 3.6 for ages 20 to 64, 2.3 for 65 to 69 and 1.5 for 70 and over[1]. The 21.6% share is high because female prisoners are few overall. With a small denominator, the same change in numbers moves the rate much further.

The offence mix differs sharply by age band

The paper records that theft accounts for a higher share among older admissions than in other age bands, and markedly higher among women over 65 than among men over 65[3].

The other bands have their own shapes. Fraud accounts for about a fifth of admissions in their twenties, and stimulant offences for about a third among those in their forties and fifties[3].

Each band carries a different distribution. Fraud in the twenties, drugs in the forties and fifties, theft among the old. Within one category called prison admissions, different things are happening.

Numbers peaked in 2016 and have been flat since

Admissions of older prisoners have not risen steadily. After peaking at 2,498 in 2016, the highest since 1989, the figure has run between 2,000 and 2,200 from 2017 onward, standing at 2,049 in 2024[1].

For close to a decade since the peak, the number has been flat in the low two thousands. What has shifted is the composition: the over-70s have exceeded the 65-69 band consistently since 2017[1]. The count holds while the ages inside it rise.

Dementia screening runs inside the institutions

For those aged 65 or over at the start of their sentence, penal institutions run a dementia screening test and refer those flagged for examination by a doctor. In 2024, 2,086 were screened, 284 went on to examination and 132 were diagnosed with dementia[1].

That is 6.3% of those screened. Of those screened, 13.6% went on to examination, and 46.5% of those examined received a diagnosis.

A programme for older prisoners came with the new custodial sentence

Treatment has changed too. With the introduction of the unified custodial sentence under the 2022 amendment to the Penal Code, a welfare programme for older prisoners was established, aimed at helping them understand the characteristics of ageing and maintain the physical and mental health needed for life in society[1].

The unified sentence merged imprisonment with and without labour, shifting the frame from compulsory work to treatment aimed at rehabilitation. A programme for older prisoners was placed inside it. Inside the prison, arrangements are being built for an ageing population.

A rate can rise for two reasons

Either the numerator grows or the denominator shrinks. The paper states plainly that it is the latter. Reading 13.8% as evidence that crime among older people is rising inverts what the source says.

Numbers have run flat in the low two thousands since the 2016 peak, and the rate per 100,000 is a quarter of that for ages 20 to 64. The share still climbs because admissions in every other age band keep falling.

The same misreading recurs whenever volume becomes the subject. The 20% take-up rate for public assistance（このサイトの記事） turns on the same distinction: unless you separate what is moving the numerator from what is moving the denominator, the remedy points the wrong way.

The figure at the exit has stopped falling

One number has not moved, though. The one at the exit.

Reimprisonment rates among released prisoners show that the overall two-year reimprisonment rate stood at 13.8%, up 0.8 points on the year[4]. That comes after a downward trend from the 23.4% recorded in 1999, with the figure first dropping below 16% in 2019 at 15.7%[4]. This is the year it turned back up.

The direction reverses by age band

By age, the under-30s have been consistently lowest, while the 50-64 and 65-and-over bands have run consistently higher than both the under-30s and the 30-49 band[5]. In 2023, the under-30s fell 0.9 points while the 30-49 band rose 0.6 points and both the 50-64 and 65-and-over bands rose 1.3 points[5].

The younger band fell and the older bands rose. The overall 0.8-point rise is the average across those opposing movements.

Theft carries the highest rate

By offence, theft stood at 19.7%, the highest of any category for the past twenty years, up 0.3 points on the year and 9.4 points below its 2004 level[5]. The others run 15.6% for injury and assault (up 4.1 points), 12.8% for stimulant offences (up 2.2 points) and 8.4% for fraud (up 1.4 points)[5].

Theft is the offence that accounts for the largest share of older admissions, and theft carries the highest reimprisonment rate. The reason people come in and the reason they come back sit in the same place.

Full-term and parole differ by more than double

Type of release matters too. Those released at the end of their term stood at 20.8%, against 9.6% for those released on parole[4]. More than double.

Parole comes with arrangements for where the person will live; release at full term does not. Parole is harder to obtain without a settled address, so part of the gap reflects who gets selected. The fact that those released without arrangements return more often still stands.

The five-year figures show the same shape

Extending the window does not change the direction. For those released in 2020, the five-year reimprisonment rate was 10.4 points below the 2001 cohort overall, 13.2 points below for those released at full term and 7.4 points below for those released on parole[4].

Over the long run the rate has come down. The 2023 rise is one year inside that trend. But the gap between full-term release and parole has not narrowed across those twenty years.

Keep the entrance and the exit apart

The entrance figure is not high once population is taken into account. Only the exit figure, the rate of return, remains high. Merge the two into a single subject called crime among older people and the response tilts toward enforcement. Keep them apart and what is needed becomes the housing and daily life that follow release.

A welfare programme for older prisoners now sits inside the institutions. Dementia screening runs there. The arrangements inside have changed to match an ageing population. Whether the arrangements outside have changed by as much is not something these statistics answer.

When 132 people diagnosed with dementia walk out

As the shared structure of programmes that fail to reach people（このサイトの記事） sets out, what is provided and what is used are different things. 93.4% of adult guardianship applications are filed to manage or close a bank account（このサイトの記事）, which shows how narrow the entrance to decision-making support already is.

What awaits the 132 people diagnosed with dementia when their sentences end lies outside the white paper's statistics. No published figure follows whether someone diagnosed inside was connected to support outside.

Further Reading

• Keimusho Shika Ibasho ga Nai Hitotachi (People Who Have Nowhere but Prison)（外部サイト、新しいタブで開きます） by Joji Yamamoto (Otsuki Shoten). A former member of the House of Representatives who has written continuously about the gap between welfare and criminal justice, drawing on his own time in prison. Useful for seeing the individual paths behind the statistics.

References

FY2025 White Paper on Crime, Part 4 Chapter 8 Section 2: Crime by Older People, Corrections — Ministry of Justice, Research and Training Institute (2025). Ministry of Justice

FY2025 White Paper on Crime, Part 5 Chapter 3: Trends in Reimprisonment Rates among Released Prisoners — Ministry of Justice, Research and Training Institute (2025). Ministry of Justice

FY2025 White Paper on Crime — Ministry of Justice (2025). Ministry of Justice