Sat, Aug 17, 2019 - Page 8 News List

Many pollsters, one public opinion

By Yang Tai-shuenn 楊泰順

As a professor of political science, I am often invited to analyze public opinion polls, so I am often among the first to see the results of surveys. I have been doing this for years, and based on my experience, poll results in Taiwan are full of strange surprises. A look at the slew of polls released last week on prospects for the presidential election on Jan. 11 next year would leave one wondering which one really reflects public opinion.

The Formosa online news service appointed Beacon Marketing and Research to conduct a survey on New Taipei City residents’ support for prospective presidential candidates. According to this survey, in a duel between President Tsai Ing-wen (蔡英文) of the Democratic Progressive Party and Kaohsiung Mayor Han Kuo-yu (韓國瑜) of the Chinese Nationalist Party (KMT), Tsai would be the winner with a support rate of 41.5 percent against Han’s 36.2 percent. If Taipei Mayor Ko Wen-je (柯文哲) were added as a third candidate, the three contenders would have about equal support.

These results are similar to those of a poll conducted at about the same time by the Green Party Taiwan, except that the Green Party poll gave Tsai a bigger lead over Han at 7 percentage points and put Tsai 5 percentage points ahead of Han in a three-way contest.

The Taiwan New Constitution Foundation, which was established at the end of last year, appointed Trend Survey to do its survey. The results, announced on Aug. 7, showed that Tsai would trounce Han by 51 percent to 31.2 percent in a two-way vote and she would still be far ahead if Ko were added to the mix.

The Chinese-language Apple Daily appointed Statinc to conduct a poll whose results were announced at about the same time. This poll found that there would not be much of a gap between Tsai and Han in a two-way race, but if Hon Hai Precision Industry Co founder Terry Gou (郭台銘) were to join the fray with Ko as his running mate, Gou would be ahead of both Tsai and Han. Interestingly, if only the results from landline interviews are counted, Gou’s lead over Han goes up to 6 percentage points. Ironically, was it not Gou who suggested some time ago that a purely landline-based poll would be unfair to him?

At the same time, the Taiwan Competitiveness Forum appointed Taiwan Real Survey to conduct the same kind of poll. This survey put Han ahead of his rivals in a three-way and a four-way contest, but had him leading by the smallest margin in a race between Tsai, Han and Gou.

Within the space of one week, these companies announced completely different results for a three-way race, including wins for Tsai, Gou and Han, or a dead heat for all three. The results put the candidates in just about every possible order.

Given that each of the commissioning organizations has a clear-cut political standpoint, it is reasonable to wonder whether the results have some political purpose other than reflecting public opinion. However, credibility and scientific methods are the basic principles that polling organizations depend on for their existence, so are they willing to sacrifice their credibility to meet clients’ demands?

These companies should bear in mind that surveys on electoral support rates will eventually be put to the test by the actual election. All will be revealed on the day the votes are counted. Are they not worried that the election results will cause the public to question the way the polls were conducted? Maybe they assume that people will have forgotten by then.

This story has been viewed 1480 times.

Comments will be moderated. Keep comments relevant to the article. Remarks containing abusive and obscene language, personal attacks of any kind or promotion will be removed and the user banned. Final decision will be at the discretion of the Taipei Times.

TOP top