After the Lower House election took place in February, some people on social media called into question the surprisingly good results of the newish Team Mirai party, which managed to secure 11 seats out of a total 14 candidates. There were claims that something fishy was going on because the numbers from one constituency to another were suspiciously alike. As it turns out, there was nothing unusual except maybe an acute case of cognitive dissonance.
Which isn’t to say the election overall was entirely on the up-and-up. Since last year, several mainstream media outlets have run stories about “mizumashi,” or vote-padding, in Tokyo’s Ota Ward. On January 16, Yomiuri reported that in last summer’s Upper House election, there was evidence that Ota Ward’s election committee “inflated” the number of invalid ballots by 2,600 in order to cover up a discrepancy between people who showed up to vote and the number of ballots submitted, a followup to an Asahi Shimbun article that appeared last July 20 that said pretty much the same thing but also mentioned that perhaps mistakenly counted absentee ballots were to blame.
In a March 2 report, TBS Radio attempted to explain the problem. In the average general election, about 5 percent of ballots are deemed invalid for various reasons. In Japan, a voter has to write the name of the candidate they want on the constituency ballot and indicate the name of the party and/or candidate they prefer on the proportional ballot. If they fail to write or indicate a name or write an invalid name, the ballot is considered void.
When election officials in Ota Ward counted the ballots in last summer’s Upper House election they found that the difference between the number of people who showed up to vote and the number of ballots was about 2,600. Asahi said that according to a retired election official, the people in charge of the vote in Ota Ward inflated the number of ballots in order to make the numbers add up. He stressed that these extra ballots were counted as being invalid or void, which means that the results didn’t change at all. And since it is fairly well known that 5 percent of votes are usually deemed invalid in a given election, the idea was that no one would question these numbers.
However, it is illegal to tamper with vote numbers, and the retired official who spilled this intelligence said he found out about the mizumashi while dining with current election officials, who, over drinks, revealed what they had done. They seemed to think it was no big deal and implied that they had been taught this method by their superiors. However, the retired official was quite shocked and posted his reaction on social media. That’s when the media picked up the story, though nothing came of it at the time.
The matter came up again in the wake of the recent Lower House election. Both Tokyo Shimbun and Asahi Shimbun ran followup stories about the Ota Ward vote-padding, presumably in a bid to stimulate conversation as to whether something similar took place in February. Tokyo Shimbun ventured the idea that the Ota officials may not have actually thought they were doing anything illegal, since the ballots they added were invalid and so had no effect on the outcome. An expert in elections told Asahi Shimbun that officials who count ballots after the polls close are under considerable pressure to submit their results as soon as possible, so any discrepancy, like the one described in the Ota Ward case, would delay those results.
This time, however, the authorities took action. According to Asahi, police have sent charges to prosecutors against 3 election workers in Ota Ward for last summer’s vote-padding and one more instance of election wrongdoing. The police believe about 30 employees were involved in the crime, but only 3 so far have admitted to the charges. The police also believe that similar improprieties were carried out in at least 7 elections in Ota Ward since 2016.
The police claim that the suspected officials “forged” 2,500 constituent ballots and 2,700 proportional ballots in last summer’s Upper House election, and also “removed” 18 invalid ballots from the Ota Ward results of the 2024 Tokyo gubernatorial election. There is also evidence that 70 ballots were removed from the 2022 Upper House election.
Apparently, it was the Ota Ward office that, having discovered the discrepancy, reported its suspicions to the police. The Asahi reports that as of 3 a.m. on July 21, the morning after polls closed, election officials concluded their count and discovered the discrepancies. As already pointed out, there is an expected number of invalid ballots for various reasons in a given election. It is also often the case that voters do not deposit their ballots in the ballot box and instead take them home, either by mistake or on purpose, which could also account for a discrepancy. These are understandable explanations, but in this case the discrepancy was much larger than the margin of error, and so the officials would have to come up with another explanation. Instead they just made up a number in order to account for the missing ballots.
One theory is that officials may have miscounted the number of absentee ballots, but in any case they artificially inflated the number of invalid constituent ballots by 2,500 at 3:30 a.m., and the number of proportional ballots by 2,700 at 4:50 a.m. Two of the employees then destroyed any documentation that recorded these changes. It’s believed the chief administrator of the Ota Ward election committee was told that the discrepancy was greater than 2,000 ballots and that the numbers were made up, but the administrator did not report this information to the Ota Ward office.
In a related aside, it’s perhaps notable that many voters in Osaka were quite ticked off about local elections that were called at the last minute to take place alongside the Lower House election in February. For some reason, the mayor of the city of Osaka and the governor of Osaka Prefecture, both members of the Japan Innovation Party, called special elections for their respective posts despite the fact that they are up for election in April 2027. Some constituents called the special election, announced less than ten days before it was to take place, a waste of money and time, since it cost the region ¥2.8 billion.
What makes it interesting is the number of ballots that were subsequently thrown out. In the poll for mayor, a whopping 13.77 percent of the ballots were deemed invalid, compared to 5.1 percent in the previous mayoral election in 2023; while 10.29 percent of ballots for the governor’s race were voided (1.98 percent in 2023). One reason a ballot is voided is that the voter purposely wrote in a joke candidate, or someone who is bound to be disqualified, as a way of protesting the election itself. Of course, both incumbents in this case won easily, which was the purpose of the election, but that goes without saying, doesn’t it?

This is a really well researched piece. The detail about election officials casually revealing the mizumashi over drinks is honestly one of the more surreal moments I’ve read about in Japanese civic governance. It says a lot about how normalized certain practices can become within institutions. The part about Osaka’s unusually high invalid ballot rates is also worth digging into further. Do you think the media will continue to follow this story closely, or will it quiet down once the prosecution phase begins?
Not sure. We would hope Tokyo Shimbun might get into it in more detail since they are nominally a local paper. As for the Osaka story, we definitely would like to know more, but it may already be old news.