Vote-Weight Disparity Passes 2x in 16 Districts — Japan's Last Redistricting Started at 1.999

2017 redistricting0 districts at 2x or more

Calculated on 2015 populationWidest gap 1.956x

2020 census23 districts at 2x or more

Under the districts then in forceWidest gap 2.096x

2022 redistricting0 districts at 2x or more

Calculated on 2020 populationWidest gap 1.999x

January 202512 districts at 2x or more

Estimated from resident registry

January 202616 districts at 2x or more

Estimated from resident registry

The rule for drawing districts is that each must hold at least as many people as the smallest district and fewer than twice that number; the 2022 plan came in at 1.999 times. It starts with 0.001 of headroom. Because seats are apportioned on the large-scale census held once a decade, the next correction has to wait for the 2030 results. The figure of 16 districts as of January 2026 is NHK's estimate based on the resident registry.

The disparity in the weight of a vote in Japan's House of Representatives elections is widening again. Working from the resident registry as of January 2026, NHK estimated that 16 districts now stand at 2x or more, four more than the 12 a year earlier.

Read that as the system failing to keep up and you misread it. This widening is what the design produces.

Look at the most recent redistricting. The plan recommended in 2022 by the House of Representatives Electoral District Delineation Council had a widest gap of 1.999 times[2]. The most populous district, Fukuoka 2nd, held 547,664 people; the least populous, Tottori 2nd, held 273,973. Districts at 2x or more numbered zero.

The standard is "less than twice." The plan was placed 0.001 short of it.

Why does it come out so close to the line? Trace how districts are drawn and two constraints appear.

The first is that the population used for the test is frozen at the moment of redistricting. The drafting policy the council adopted in February 2022 requires each district's population to be "at least that of the district with the smallest population among the prefectures with the fewest people per member, and less than twice that number." The population here is that of Japanese nationals, taken from the 2020 census with foreign residents subtracted. Whether the standard is met is judged on a single day's snapshot.

The second is the interval between corrections. Apportionment of single-member seats among prefectures is fixed by law to the Adams method, applied on the large-scale census conducted once every ten years, and has been used since the 2020 census. The recommendation must come within a year of the census results first being published in the official gazette. The recommendation was issued on 16 June 2022, the law promulgated on 28 November, and it took effect on 28 December[3].

Judge on a snapshot, correct once a decade. Put those together and the previous cycle shows the result.

The 2017 redistricting, measured on 2015 population, had a widest gap of 1.956 times[4] and zero districts at 2x or more. By the time those same boundaries met the 2020 census, the widest gap was 2.096 times[5] and 23 districts[5] had crossed the line. Zero to 23. So 140 districts across 25 prefectures were rearranged, and the count went back to zero.

What is happening now is the second lap of that round trip. Reset to zero in December 2022, and at 16 three years later.

Vote-weight disparity does not arise because the system is broken. It arises because two things that move on different clocks are joined by a single mechanism.

People move every day. District boundaries move once a decade. Nothing in the present arrangement bridges that difference. So the gap starts opening the day after a redistricting and keeps opening until the next census. Correction is not late; its period simply does not mesh with the period of the phenomenon.

How much headroom is taken then matters. Because the standard is "less than twice," a plan need only clear it. The 1.999 of 2022 satisfies the law. It also leaves almost no room for movement. Redraw at around 1.7 and the same migration would take far longer to push a district past 2. Whether boundaries are drawn flush against the limit or well inside it decides how many of the next ten years are spent in what courts have called an unconstitutional state. The drafting policy asks only for the former.

Onto this maps the question of how to treat regions losing population. The ten prefectures that lost a seat are all losing people. The policy goal of halting regional decline and the principle of apportioning seats in proportion to population do not point the same way. If the first does not succeed, the second will take further seats from those regions at the next revision. Failure in regional revitalisation thins regional representation as a consequence.

So what can be done? Three routes are available. Shorten the interval, inserting a review between censuses using the resident registry, so that fewer years are spent in the widened state. Set a tighter ceiling on the gap at the moment of redistricting, capping it at, say, 1.7 rather than 1.999, so that fewer districts cross the line within the decade. Or change how proportionality itself is treated, which runs straight into the constitutional demand for equal weight per vote. The first two are technical choices; the third is a choice about values. The debate at present largely occupies the first two.

When a system keeps returning to the same place, the thing to doubt is the design rather than the operation. Setting an indicator flush against its threshold and calling it met has the same shape as the problem taken up in designing outcome indicators（このサイトの記事）. How to build institutions on the assumption that people keep moving is inseparable from the migration described in population decline and the concentration in Tokyo（このサイトの記事）. The figure of 16 is not a spike this year; it is a value on the way up to the next redistricting.

Further Reading

• 日本の国会議員 — 政治改革後の限界と可能性 (Japan's National Legislators: Limits and Possibilities After Political Reform)（外部サイト、新しいタブで開きます） (Shinsuke Hamamoto, Chuokoron-Shinsha, April 2022). An empirical study of how the behaviour of legislators and the balance of power within parties changed after single-member districts were introduced. It offers material for asking how far a technical choice such as districting reaches into the nature of representation.

References

Schedule for Redistricting of Single-Member Districts Based on the 2020 Census, and Drafting Policy and Related Materials — Ministry of Internal Affairs and Communications (2022). House of Representatives Electoral District Delineation Council

On the Redistricting of House of Representatives Single-Member Districts — Ministry of Internal Affairs and Communications (2022). Ministry of Internal Affairs and Communications

Outline of the Act Partially Amending the Electoral District Delineation Council Act and the Public Offices Election Act (Adoption of the Adams Method) — Ministry of Internal Affairs and Communications (2016). Ministry of Internal Affairs and Communications

Vote-Weight Disparity of 2x or More Rises to 16 Districts, NHK Estimate — NHK (2026). NHK NEWS WEB

House of Representatives Electoral District Delineation Council — Ministry of Internal Affairs and Communications (2026). Ministry of Internal Affairs and Communications