The Neighbours' Villages Are Empty Too: What Is Left After They Empty

A question that can be checked

The last piece ended on a question: China's villages are empty — is it that development was too fast, or that the destruction was too thorough?

The two explanations predict different things.

If it is "development too fast," then the neighbours' villages, whether or not they were ever torn down, will empty and age too.

If it is "destruction too thorough," then once the people have gone from the neighbours' villages, the things that can still receive people — lineages, temples, associations — should remain in larger measure.

The first prediction is easy to check. The second is hard. I will set out what I found.

Japan: the people left, the shell stayed

Japan has a word, genkai shūraku ("marginal settlement"), for a village where more than half the residents are over 65 and even common village affairs are hard to keep going. A 2006 survey by the Ministry of Land, Infrastructure, Transport and Tourism counted 7,873 such settlements, 12.6%.

And the village organisations? Start with the neighbourhood associations (jichikai, chōnaikai). In the Ministry of Internal Affairs surveys, the nationwide membership rate by household was 78.0% in 2010 and 71.7% in 2020, falling steadily across the decade; but the number of associations themselves has stayed between 290,000 and 300,000 for thirty years, 295,838 in 2023. That is, the shell is still there and the people inside have thinned. A public-opinion survey in 1970 found that 90.2% of those aged twenty and over had belonged to one, but the method differed and the two cannot be compared directly.

One more detail: over the decade, the smallest municipalities (population under 10,000) saw the smallest fall in membership, only 3.1 percentage points; the largest fall, 7.6 points, was in cities of 200,000 to 300,000. In small places, people find it harder to leave. (From a summary of the Ministry's survey.)

Now shrines. The shrines under the Association of Shinto Shrines number about 78,000 (end of 2017). One compilation says that about 30% of their priests are kenmu — one person looking after several shrines — and it is by this device that shrines in depopulated areas have not disappeared one by one. In the Tajima region of Hyōgo there are 738 shrines and only 29 head priests, an average of 25.4 shrines per priest; the Hyōgo average is 10.7. (From a 2022 report in the Kobe Shimbun.) The shrine world itself admits the vicious circle: as ujiko (the households that support a shrine) decrease, income falls with them, priests are less willing to take the post, shrines have even fewer people to look after them. Some shrines have no priest at all, even a part-time one, and have become "inactive shrines."

So in Japan: the villages are empty and old; the organisations — shrines, associations — mostly keep their shells, but what is inside is growing thin.

Korea: older than we are

Korea's numbers are starker. In the 2020 Census of Agriculture, Forestry and Fisheries, 42.3% of the farm population were 65 or over, up from 21.7% ten years before, a doubling; in the same year only 16% of the national population were 65 or over. The provisional result for 2025 is 51.3%, over half for the first time. But that survey also folded in administrative records such as the farmland register, so it cannot be compared directly with 2020; another annual survey in 2024 gives 55.8%.

Put side by side, the scale is clear: in China's seventh census, 17.72% of the rural population is 65 or over. The two bases differ — "farm households" against "rural resident population" — so one cannot say directly who is older. But "the neighbour's countryside is much older than ours" can be said with confidence.

As for Korea's munjung and jongjung, the clan organisations, and what has happened to them in recent decades, I found no figures this round that I can cite.

Taiwan: the temples remain

The Ministry of the Interior's temple registration records 11,731 registered temples on the island at the end of 2008, of which 78.4% were Daoist and 19.5% Buddhist; at the end of 2002 there were 11,423. The increase was mainly because policy encouraged unregistered temples to register, not necessarily because new ones were built. (From a relay of the Ministry's statistics; I did not find the latest national total.)

Another kind of thing is called jisi gongye ("ancestral-sacrifice trusts"): immigrants of the Qing and Japanese periods, missing the ancestors of their home places, and their descendants pooled money to buy land, the income used for sacrifice and shared meals, the land mostly registered in the ancestors' names. The Ministry of the Interior still has a page on it today, and counties and cities still publish rolls of such corporations. The anthropologist Lin Meirong studied the Earth God temples around Caotun and treated them as an indicator of the settlement. (This is the study of one township only.)

What the ageing of Taiwan's countryside has to do with the number of ancestral halls and sacrifice trusts, I found no quantified data.

Side by side

What I read from this table: the neighbours' villages are just as empty and just as old. That supports the "development too fast" explanation: people leaving is something that happens everywhere.

But "the shell remains" does not mean "people are still inside." About 30% of Japan's shrines rely on part-time priests; among priests, more than 60% say the number of ujiko has fallen. So "the neighbours' organisations are still there and ours are gone" cannot now be said either.

That is all that can be said this round. For the most critical pieces — Korea's clans, Taiwan's ancestral halls, village-level organisations in China that can be compared — I have no figures to set side by side. This table testifies for no one's judgment and is not used to refute anyone's; my judgment does not rest on it.

Turning it around

First, the neighbours' villages were also "torn down." In December 1945 the occupation authorities issued the "Shinto Directive," cutting shrines from the state; that was the dismantling of a system, only by a different method from China's. Neighbours' Heavens wrote about it. So the premise "the neighbours were never torn down" is itself inaccurate.

Second, a shell such as the neighbourhood association may be only an administrative shell. That the number has not changed in thirty years may only be because it is treated as the lowest rung of administration, not because the people in the village want it. The number of shells does not directly show whether the people in the village can still gather.

Third, comparison across different bases is easy to overstate. Every figure I used in this piece is from a different year and a different basis, and can only be read for scale.

What I think

Having read all these figures, I come away with three things.

First, "empty" is not something to keep arguing about. All of East Asia is emptying; development really has been fast; there is nothing to quarrel over. I said in the last piece that I do not stop people from going out.

Second, "whether what is left can receive people once the village empties" is what should be compared. Japan's residents' associations and shrines have their shells, and the shells are thinning inside, but at least there is still a shell into which people can be put back. What I saw in the accounts of that township in central Hunan is that the shell itself was broken for thirty years, and an ancestral hall went up again only after more than thirty years. (One township's example does not stand for the country.) This difference I cannot measure with these few tables, and my judgment does not rest on them; the reason I have given in the last piece.

Third, the figures still to be filled in are already listed above: Korea's munjung, Taiwan's ancestral halls and sacrifice trusts, and village-level organisations in China that can be set alongside them. I will keep looking.

If in the neighbours' villages the shell is still there and the people are gone, and in our villages the shell is gone and the people have left too, which is easier to take up again?

References

Ratings: 🟢 original read; 🟡 second-hand relay (search summaries, media, official pages), original not read.

• Ministry of Land, Infrastructure, Transport and Tourism (Japan), surveys of the type "The State of Villages in Depopulated Areas," 2006 (7,873 marginal settlements, 12.6%) 🟡: https://www.mlit.go.jp/common/000029292.pdf
• Ministry of Internal Affairs and Communications (Japan), "On the Sustainability of Neighbourhood Associations' Activities" (answers from 1,741 municipalities; membership 78.0% in 2010 and 71.7% in 2020; 295,838 associations in 2023) 🟡: https://www.soumu.go.jp/main_content/000777270.pdf; National Diet Library ISSUE BRIEF No. 1306 (4 February 2025), "The Present State and Future of Neighbourhood Associations" 🟡
• Kobe Shimbun, "Acute shortage of priests: in depopulated areas, one head priest in charge of as many as 80 shrines," May 2022 (Tajima 738 shrines, 29 head priests) 🟡: https://www.kobe-np.co.jp/news/sougou/202205/0015299611.shtml
• Fuyutsuki Ritsu, Chūgai Nippō, "Opinion," 11 December 2019 (about 78,000 shrines under the Association of Shinto Shrines, about 30% of priests part-time) 🟡: https://www.chugainippoh.co.jp/article/ron-kikou/ron/20191211-002.html
• Shimane Prefecture Shrine Agency, "Results of the questionnaire survey on shrine operation in relation to depopulation and the declining, ageing population," September 2024 🟡
• Korea, 2020 Census of Agriculture, Forestry and Fisheries (farm population 65+ 42.3%); provisional 2025 result (51.3%, basis differs from 2020); another 2024 survey (55.8%), relayed from Nongmin Sinmun, News1, Seoul Economic Daily and others 🟡
• China's seventh census: rural population aged 60+ and 65+, 23.81% and 17.72% (see "How Many People Remain in the Countryside") 🟢
• Taiwan Ministry of the Interior, "Overview of Temple Registration" (11,731 temples at the end of 2008; 11,423 at the end of 2002), and the page on "ancestral-sacrifice trusts and deity societies" 🟡: https://www.moi.gov.tw/cp.aspx?n=160; National Religious Information Network statistical reports https://religion.moi.gov.tw/ChartReport/Index?ci=1&cid=2 🟡
• Lin Meirong, research on Earth God temples and sacrificial circles in Caotun Township, from surveys 🟡
• The "Shinto Directive" of December 1945: see the references of Neighbours' Heavens D02
• "When Did Rural China End?", "How Many People Remain in the Countryside," "The Ancestral Hall Is Back"