← Back to the latest articles
QUESTION 70Politics, Polling & Democracy

By NOW IN QUESTION EditorialSourcing & correctionsReport an error

Takaichi's approval rating: 57% or 49.1%—what can a poll say about public opinion?

Two October 18 reports measure the same cabinet differently. Neither discrepancy nor popularity is a substitute for explanation.

NNN and Yomiuri reported 57% approval for Japan's Takaichi cabinet on October 18, 2026; Kyodo reported 49.1%. Japanese-language X records displayed related terms. Before choosing a preferred number, readers need to distinguish fieldwork, measurement and the political meaning of approval.

Two reports published on October 18, 2026 gave different approval ratings for Japan's Takaichi cabinet: 57% in the NNN–Yomiuri survey and 49.1% in Kyodo's. [1][2]

The 7.9-percentage-point gap crosses the psychologically powerful line of 50%. It does not, by itself, show that one result is genuine and the other fabricated.

What was measured, and when?

Survey Fieldwork Cabinet approval Reviewed publication
NNN–Yomiuri October 16–18 57% October 18, 10 p.m. Japan time
Kyodo October 17–18 49.1% October 18, 5:47 p.m. Japan time

NNN's account describes a nationwide telephone survey of eligible voters with 1,019 respondents: 402 reached by landline and 617 by mobile. The reviewed Kyodo report also describes a nationwide telephone survey, but does not provide enough detail to reconstruct its sample composition or calculation. [1][2]

The publication dates match; the fieldwork periods do not entirely match. Nor is this a time series in which one survey fell from 57% to 49.1% within hours. It is a comparison of separate surveys.

The Japanese-language X evidence

The Japan page of Trends24 displayed the Japanese terms for “cabinet approval” and “roughly unchanged at 57” in several October 18 evening UTC snapshots, reviewed on October 19. [3] Public X information surfaced in Yahoo's real-time search also included Japanese posts referring to the October 18 report of 57% and posts discussing the 49.1% result. [4]

This supports limited Japanese-language circulation, not an inference about users' nationality. We did not obtain every national trend, comprehensive post counts or the distribution of opinion on X.

Why might surveys differ—and what is unconfirmed here?

A poll is not a device that reads everyone's mind simultaneously. It estimates opinion through decisions about contact, participation, questions and aggregation.

The American Association for Public Opinion Research emphasizes disclosure of sampling, wording, mode and weighting, and consistency of measurement when tracking change. [5] Those are reasons to examine how a number was produced, not reasons to dismiss polling.

Sample composition, question context, nonresponse and processing are therefore relevant possibilities. This review has not established which factors produced this particular gap. Saying “it must be older respondents” or “it is because they used phones” would invent an explanation.

Without the underlying design, it is also inappropriate to declare the entire difference to be within a specific margin of error. Sampling variation is not the only possible source of uncertainty. Conversely, disagreement between estimates is not proof of manipulation.

A trend is not a contest between pollsters

To assess change, begin with comparable measurements within the same series, then examine whether other series move in similar directions. Selecting the highest or lowest result to suit a political preference defeats that purpose.

An average can be useful only with appropriate attention to timing and methodology; it does not automatically become the “true” rating. This article does not manufacture a new approval estimate by averaging the two figures.

Approval does not endorse every decision

The following is this article's political-philosophical analysis.

Support for a cabinet may express expectations, policy judgments or a comparison with alternatives. A single answer cannot identify every reason behind it, still less establish assent to each policy or ministerial decision.

High popularity cannot replace a government's explanation of a particular action. Low popularity does not prove every policy mistaken. Our account of Kazuo Yana's resignation and appointment responsibility also separates individual evaluation from institutional accountability.

What does the 50% line decide?

The contrast between 49.1% and 57% is striking partly because readers import the logic of electoral victory into an estimate of sentiment.

But a survey and an election perform different tasks. One estimates attitudes; the other allocates authority under established procedures. Neither a poll nor a majority can, by itself, decide the truth of a factual claim or eliminate the rights of a minority.

Above 50%, a government still owes explanations to people who reject it. Below 50%, supporters' reasons do not become worthless. Democracy needs both counting and the exchange of reasons.

The people outside the number

People who decline an interview, hesitate to respond or find the available answers inadequate remain members of society. Survey research tries to estimate their views responsibly, but publishing a percentage does not erase social complexity.

The appropriate response is neither to reject measurement nor to treat it as an oracle. Ask pollsters how the information was made and ask officials why their decisions are justified. These responsibilities are connected without being interchangeable.

Respecting public opinion means recognizing both what a number can represent and what a number cannot decide.

Sources and verification

  1. NTV News NNN / Livedoor: Takaichi cabinet approval at 57% (October 18, 2026, 10 p.m.; Japanese) (survey organization's own account in syndication)
  2. Kyodo / Nikkan Sports: Approval falls below 50% (October 18, 2026, 5:47 p.m.; Japanese) (a separate independent survey, not corroboration of NNN's numerical result)
  3. Trends24: Japan X snapshots (October 18 evening UTC records reviewed October 19; third-party source)
  4. Yahoo real-time search: disapproval / cabinet approval (public search displays reviewed October 19; results change)
  5. AAPOR: Best Practices for Survey Research (primary methodological guidance, reviewed October 19)

Information checked: October 19, 2026. Fieldwork and publication dates are distinguished above. The causes of the numerical gap remain unestablished. This article does not endorse a party or candidate.

SHARE

Share this article

Carry the question into another conversation.

DISCUSSION

Discuss this question

Write in Japanese or English. Comments are translated automatically and shared across both versions of the article.

0/2000 characters

Loading comments…

DISCOVER MORE

What to read next

Explore the same question through other stories and ideas.

Articles attracting attention on NOW IN QUESTION.

  1. 01QUESTION 25What is Todai Kokugo Kenkyukai, and why do people search ‘controversy’ or ‘suspicious’?
  2. 02QUESTION 13Why Did Surgery at Kyoto University Hospital Not Stop?
  3. 03QUESTION 18Who Can Stop a Festival That No One Organizes?—Yajū Day and the Protocol Crowd