By Tadaaki Fujisawa
Half a year has passed since the Yoshihide Suga administration was launched. The Cabinet’s support rate has trended below its nonsupport rate in the public opinion polls conducted by media outlets. The survey conducted by the Sankei Shimbun and Fuji News Network (FNN) is the one exception. Sankei and FNN suspended their public opinion poll after they discovered erroneous data entry and they relaunched it in January this year. Have they been able to ensure the poll’s reliability, though?
Sankei-FNN survey findings differ from those of other polls
According to the joint public opinion poll conducted by Sankei and FNN on March 13–14, the support rate for the Suga Cabinet is 51.4% and the nonsupport rate is 42.8%. In the poll conducted by the Mainichi Shimbun and the Social Survey Research Center on March 13, the Cabinet approval rate was 36% and the nonapproval rate was 55%. The Sankei-FNN poll findings for Cabinet support and nonsupport differ considerably.
The same can be said of the findings of the Sankei-FNN joint polls conducted in January and February. With the relaunch of their poll, Sankei and FNN started asking the following follow-up question to respondents who gave a vague answer to the question asking about Cabinet support: “Which is closer to your view: ‘I support the Cabinet’ or ‘I do not support the Cabinet’?” In this way, they press respondents to answer. This could result in steering public opinion in a certain direction.
In June last year, it was discovered that fabricated data had been mixed in the Sankei-FNN polling data. According to the media outlets, fabricated data had been entered in the data for all 14 public opinion polls they conducted between May 2019 and May 2020. A total of 1,886 pieces of data – which represents 12.9% of all data – were fabricated.
This was a huge problem, and the Japan Association for Public Opinion Research issued a statement, which said “Presenting fake data from entirely unscientific public opinion polls as public opinion is unforgiveable . . . This could even lead public opinion in the wrong direction.”
Looking through the 14 Sankei-FNN public opinion polls, it can only be said that the two outlets welcomed then-Prime Minister Shinzo Abe’s maneuvering on constitutional amendment and guided public opinion in that direction, through such poll findings as “76% say constitutional amendment should be discussed” (May 11–12, 2019) and “over 60% call for ‘debate on constitutional amendment’” (Aug. 3–4, 2019).
In the case of these 14 public opinion polls, the polling company that had been contracted to perform the polls subcontracted the project to another company without permission, and the subcontractor’s representatives overseeing the polling call center input fabricated data. In June last year, FNN retracted all the problematic public opinion poll findings and all related broadcasts. The Sankei Shimbun announced that it would retract all articles on the 14 polls and it said it “would suspend public opinion polls until it had confirmed and introduced solid polling methodology.”
On Jan. 15 this year, Sankei and FNN restarted their joint public opinion poll, saying they “had formulated measures to prevent data fabrication.” However, they are not attempting to explain the matter in good faith to the people of Japan. For example, they have not removed the fabricated data and re-issued the results of the public opinion polls.
In February, the Broadcasting Ethics & Program Improvement Organization (BPO), which reviews FNN’s public opinion poll reports, released an opinion statement that “there were grave violations of broadcasting ethics” in the FNN reports. BPO strongly criticized the TV station, saying “FNN has not disclosed the content of the fictitious data or described how that impacted the Cabinet and political party support rates.”
Doubts remain as to whether Sankei and FNN have sincerely reflected on how best to conduct public opinion polls, an indicator of public sentiment that influences the future of politics.