Voters were nearly equally divided on whether the Constitution should be revised, according to a survey by The Asahi Shimbun.

The survey, conducted ahead of Constitution Day on May 3, showed 45 percent backed constitutional revisions, up from 43 percent in last year’s survey, while 44 percent said changes were unnecessary, down from 46 percent.

The gap between the two camps was the narrowest since the 2013 survey.

The Asahi Shimbun began the annual survey on how the public views proposals for amending the Constitution, a key agenda pushed by Shinzo Abe, who was prime minister from December 2012 to September 2020. The ruling Liberal Democratic Party, led by Prime Minister Yoshihide Suga, is also seeking revisions.

The Constitution, drafted during the U.S.-led postwar occupation of Japan, has never been revised since it went into force in 1947.

War-renouncing Article 9 of the Constitution has long been a target of those who want changes, including Abe.

According to the survey, 61 percent supported the article as it is, more than double the 30 percent who were in favor of revising it.

In the 2020 survey, 65 percent were opposed to revising Article 9, while 27 percent backed an amendment.

Fifty-five percent of voters in their 30s were in favor of constitutional revisions, the largest group by age. Voters in their 70s or older were the least enthusiastic about amendments, with only 35 percent backing revisions.

In the survey, respondents were asked to choose up to three reasons from among eight provided.

Among pro-revision voters, the most common reason selected for their stance was “a lack of sufficient provisions concerning defense,” at 58 percent, followed by the Constitution is “getting old,” at 46 percent, and “it does not reflect the national characteristic because it was imposed by the United States,” at 35 percent.

The three most common answers chosen by respondents against revisions were: “the Constitution brought peace to Japan,” at 71 percent; “there is no major problem with the Constitution that needs to be revised,” at 41 percent; and “the Constitution has taken root among the public,” at 40 percent.

Asked if they thought about the supreme law in their daily life, 68 percent of all respondents chose “not at all” or “not so much,” while 30 percent chose “often” or “sometimes,” the survey showed.

It also found that 61 percent of respondents in favor of revisions do not give much thought or pay close attention to the Constitution, while 39 percent said they do.

The survey also showed the text of Article 9 and asked the voters for their views.

In the 2013 survey, 52 percent were opposed to revising Article 9, while 39 supported changes.

But in the following surveys, including the latest one, more than 60 percent of respondents rejected amendments to the article.

Asked to choose from among four answers to the question of whether momentum is building for revisions, 19 percent chose “significantly” or “to some extent.”

In comparison, 76 percent picked the remaining two answers: “not at all” or “not so much.”

The ratio of respondents who think momentum is gathering for constitutional revisions was 22 percent in 2019, 21 percent in 2020 and 19 percent in 2021.

The comparable figures for those who feel otherwise was 72 percent in 2019, 76 percent in 2020 and 76 percent in 2021.

The LDP is pushing to add an article to the Constitution that allows the Cabinet to issue a directive under a state of emergency and temporarily restrict people’s rights during contingencies, rather than waiting for the Diet to enact a law.

When voters were asked about the emergency article for the first time in the 2019 survey, and given three answers to choose from, 55 percent picked Japan should “respond to emergencies without revising the Constitution.”

The figure in the latest survey was 54 percent.

Thirty-three percent chose “the Constitution should be revised to deal with emergencies,” up from 28 percent in 2019.

Six percent in the latest survey selected “there is no need to revise the Constitution in the first place” in relation to emergencies.

The survey was conducted between early March and mid-April by sending questionnaires to 3,000 eligible voters selected randomly across the nation. Valid answers were received from 2,175 respondents, or 73 percent of the total.