Taiwan Today

Public opinion polls have become a hot topic in Taiwan politics. Are the major political parties capable of putting professional ethics above party interests when conducting their own surveys?

December 2, 1995 passed with­out incident. Why bother to record the fact? Well, that par­ticular Saturday was election day in Taiwan, a firecracker of a festival which in the past has generated some unsightly scenes. In the island's third leg­islative election, however, the only real news lay in the results.

The ruling Kuomintang (KMT) suf­fered its biggest ever electoral blow, barely holding onto a majority of 85 seats, down from a previous 96, in the 164-seat Legis­lative Yuan. The pro-independence Democratic Progressive Party (DPP) gained 4 seats to win a total of 54. The New Party (NP), which advocates reunification and the restoration of harmonious ties with Mainland China, while emphasizing the need for greater social justice and a crusade against corruption at home, was consid­ered the biggest winner, tripling its repre­sentation to win 21 seats.

What most surprised people about this election, however, was the number of can­didates surveys reported as likely to win, but who nevertheless lost, and vice versa. Many of the DPP’s top pro-independence activists and party elders, for example, were either voted out of office or failed to gain a seat. New Party candidates, on the other hand, who were tipped to win only a modest number of seats, confounded pessimistic expectations. All of this has again put political opinion polling in the spotlight, and raised questions about the reliability of such surveys, particularly when conducted by political parties.

To construct an impartial and accurate survey is of course a diabolically difficult process. There are so many factors, any one of which, if misjudged, can undermine a poll's result. The purpose of the survey has to be established, the questions must be designed impartially, samples have to be judiciously selected, with careful attention paid to the actual method of conducting the poll. Finally, someone has to collate and interpret the data before the results can be presented and publicized. When an expe­rienced and independent organization is in charge, the results are usually reliable. But in Taiwan, where political parties fre­quently conduct and then publicize their own surveys, the situation is less clear-cut.

Opinion polls have been part of the Taiwan political scene for a long time, but they really came into their own in 1994. That year was a milestone for democracy, with the provincial governor and assem­blymen, small-city mayors, city and county council members, and the mayors of Taipei and Kaohsiung all up for election. Most observers believed it was only natural for Taiwan to embrace opinion polling as part of the process of develop­ing into a genuinely democratic society. Lin Tung-tai (林東泰), secretary general of the non-profit ROC Public Opinion Sur­vey Association, certainly thinks so. "The fact that even the DPP has begun to believe in electoral public opinion polls shows that Taiwan's democratic movement has matured," he says. "If they'd continued to turn their backs on polling and gone on just staging street demonstrations, there'd be nothing to talk about on the democracy front here." According to Lin, for a long time the DPP did not trust public opinion polls. "They thought it was just some kind of numbers game played by the KMT," he says.

The DPP decided to use public opinion polls a year ago because the party had come to recognize that such surveys provide a scientific method of assessing electoral prospects. "Polling is a sort of map for can­didates," says You Ying-lung (游盈隆), executive director of the DPP election strat­egy committee. "It tells us where a candidate stands, if his public support is running ahead or behind that of other candidates in the same district, and shows how he can devise better campaign strategies."

You Ying-lung points out that such surveys played a vital role in last Decem­ber's legislative elections, because the campaign atmosphere was much less inflammatory and sensational than was the case in any previous election. "In such a cool campaign atmosphere," he says, "how can you best estimate your chances of winning? The answer is opinion polls."

What factors contributed to this low­ key atmosphere? People in general were thought to be more interested in the upcoming first direct presidential election, slated for March. And the economic down­ turn that began to be noticeable last June has deprived many candidates of the kind of corporate financial support necessary to mount an extravagant campaign. The opening up of cable television has also made voters less willing to sit outdoors and listen to candidates detailing their plat­ forms. Most would rather stay home to watch them debate with each other on the various TV call-in talk shows. Lin Yu-siang (林鈺祥), deputy director of the KMT's policy research department, thinks low-­profile campaigning is an inevitable step in the development of party politics. "In the future," he says, "we're going to see far fewer campaign banners fluttering in the breeze."

Times have changed, in other words, and it is no longer enough to rely on traditional methods of evaluating the election cli­mate. According to Lin, in the days when Taiwan was largely an agriculture-based society it was easy to figure out how many votes a candidate was going to get, through the KMT's system of local voter-mobiliza­tion networks. "You simply asked city or county councilmen, who then would ask district chiefs, who then would ask village chiefs," he says. "Everyone knew every­one. It wasn't a secret who you were going to vote for. Those estimates were highly accurate. I could even buy votes if I wanted to, and once I was sure you'd voted for me, then I'd pay you. Nowadays, people keep their doors shut. How can I know who you're going to vote for? Even the agricultural villages are becoming urbanized, and when a place is turning urbanized you can no longer control who votes for who."

In one sense, however, people are less secretive than before. "Taiwan society is becoming increasingly liberal, so people in general do not turn down the chance to be interviewed for a survey," says the DPP's You Ying-lung. "What's more, people are not so afraid of coming clean about their political orientation." He was surprised to find that even the most sensitive ques­tions—like "Do you think you are Chinese or Taiwanese?"—could be asked over the telephone without arousing antipathy in those polled.

This represents a real sea change. Tim T.Y. Ting (丁庭宇) is now chief consultant to the Taiwan Gallup Organization, Inc., which began operating in Taiwan in 1989. He has long been a controversial trail­blazer of public opinion surveys. In 1988, when Taiwan's democratic movement had just started to ferment, he was working for the newly established, non-profit Public Opinion Poll Foundation. Ting found him­self in the unenviable position of having to do a "politically incorrect" survey that put his personal safety at risk. One question raised in that poll was: "Do you think President Chiang Ching-kuo is in good health?" Half of the people interviewed answered "No." The day after the result of the survey was published in the press, the government announced that the president was in perfect health. A week later, Chiang died of a combination of chronic illnesses. "In those days, you really had to have guts to conduct such a survey and, what's more, to publish the results," Ting laughs. "You could be killed because of that, you know.">

In Taiwan, political parties naturally also rely on public opinion polls to find out what people want from the government. Based on the results, they try to frame their platforms in language that will win voter support. This explains how the New Party chose the four planks in its campaign plat­ form—no tax increases, no war with main­ land China, a blitz on corruption, and a five-day workweek—for last December's legislative elections. "Based on the sur­veys we did, that was what people wanted," says Fei Hung-tai (費鴻泰), a Taipei city New Party councilman in charge of the party's polling apparatus. "We simply articulated their wishes."

Surveys are even used by political parties to defend themselves. A few weeks before the December election, the DPP announced the results of a survey claiming to show that many Taiwan residents con­sidered the KMT to be a corrupt party with links to gangsterdom. Shortly afterwards, however, the KMT published a survey report of its own, indicating that over 60 percent of the people polled supported the KMT-ruled government's work of cracking down on gangster organizations, and wanted to see new, stricter laws for that purpose. "People out there keep on saying the KMT is tied to gangster society," the KMT's Lin Yu-siang says excitedly. "Whereas this survey tells you that the KMT is doing something to crack down on gang­ster organizations. When you're under attack, you need to fight back. This is what we fight with.

While public opinion polls obviously serve a purpose in the run-up to elections, many experts strongly question the reli­ability of surveys conducted by the three main political parties. "Political parties often put professionalism and ethics on one side when conducting surveys," says Lin Tung-tai of the ROC Public Opinion Survey Association. "Nothing is more important than that party's political inter­ests. Whether it's the KMT, DPP, or the NP, they're all the same. So I'm afraid that the public opinion poll has already become just another electoral tool of the main parties, stunting its growth in Taiwan before the general public has had a chance to discover the beauty of it, its essential qualities."

Lin justifies his belief by reference to the 1994 Taipei mayoral election. Before the contest, at least two parties announced survey results predicting that their candi­date would win. "How could that be," he laughs, "when there can only be one win­ner?" He thinks it was a stratagem to win over those lukewarm voters who might decide how to vote only after reading a survey result, despite yet another recently published survey showing that most peo­ple do not think their voting intentions would be affected in that way. "There's a saying that people like to be on the winning side," Lin says. "Public opinion polls have this kind of effect."

The Taipei mayoral election was also the scene of one strange phenomenon. The DPP kept announcing in the media that, according to its own surveys, their candi­date, Chen Shui-bian (陳水扁), was least likely to win. The poll was conducted impeccably, with one exception—the samples used in the survey were drawn exclusively from a traditionally pro-KMT district where the DPP enjoyed least sup­port. The public was never told. Why did the DPP act that way? "Taiwan voters are strange," Lin says. "Often in an election they have a tendency to vote for the underdog out of sympathy, and then the underdog turns out to be the big winner." Which is exactly what happened in Chen Shui­-bian's case: he went on to become mayor of Taipei city.

The problem with the parties' own polls is that everything—the formation of the initial survey idea, question-design, sample selection, analysis, interpretation, and subsequent publication—is in the hands of partisans, rather than some inde­pendent body. One common motive is a desire to economize. The DPP now has thirty-five telephone lines dedicated to surveys set up in its headquarters, and You Ying-lung points out that, "By doing it ourselves, we save a lot of money." The New Party also has telephone lines for polling hooked up in its headquarters and in local chapter offices. "In the past," says New Party councilman Fei Hung-tai, "Gallup Taiwan did public opinion polls for us. Gallup's a profit-making organi­zation. We don't have a lot of money, so we decided to do it ourselves." Even the KMT, which occasionally commissions outside bodies to do surveys for the party, on the whole prefers to do its own surveys. "It's better that we do them our­selves," Lin Yu-siang says. "First of all, we save money. Most important, only our people know what we want to get from a survey."

Tim Ting of Gallup Taiwan thinks that when political parties entrust surveys to their own people, political interests weigh far more with the organizers than profes­sional ethics. For example, he ridicules the KMT's use of pre-recorded telephone messages as a means of conducting a poll—instead of somebody personally explaining to interviewees the purpose and meaning of the survey questions, the per­son polled is expected to answer questions which have been recorded beforehand and integrated into a computer-controlled acoustic system. "It's a big joke!" Ting laughs. "They don't know the first thing about public opinion polls. Random sam­pling means that everyone has an equal opportunity to be interviewed, but that won't work if you use a machine to talk to your interviewees."

Lin Yu-siang of the KMT disagrees, pointing out that answers given by people under the voting age of twenty will not be integrated into the party's computerized system. He actually believes that such a system is superior. "Some people would rather deal with a computerized acoustic system than a human caller," he says. "At least the person being interviewed can feel safe talking to an acoustic system. What's more, we don't have to arrange for twenty­ odd people to make phone calls all night. They'd go crazy. It's much easier to leave it all to the computer." But Lin Tung-tai of the ROC Public Opinion Survey Associa­tion maintains that, "Chinese people feel uncomfortable dealing with lifeless objects." He thinks, however, that use of an acoustic system need not necessarily affect a survey's reliability, and he acknowledges that it also saves money.

According to Gallup's Tim Ting, the real damper on the devel­opment of polling in Taiwan is the way political parties publicize the results of their surveys, espe­cially during the three-month period prior to voting. "They use a survey as if it was just another kind of publicity to boost their image in the voters' minds," he says. "They don't really try to understand what people want. What's worse, the publicity is often selective." Political parties rarely make public all the questions included in the survey-something any unbiased organization would do as a matter of course. "You know that the answer to the first question will affect the answer to the sec­ond question, and so forth," Ting says. But when dealing with their own surveys, po­litical parties "often announce just the re­sults of five questions, when in fact the entire survey contained twenty-five. How do you think those five questions were chosen? No party would ever announce anything detrimental to itself."

Naturally, the political parties deny these accusations. The DPP thinks it unfair to denigrate all surveys conducted by political parties; You Ying-lung empha­sizes that whether surveys conducted and publicized by political parties are reliable is up to the electorate to decide. "We believe that voters have the wisdom to make comparisons between parties' surveys," he says. "If a party consistently announces absurd survey results, people are eventu­ally going to stop taking it seriously."

Although most of the DPP's surveys concerning candidates' popularity are done for internal party reference only, You Ying-lung also insists that political parties have an obligation to publicize issue-oriented surveys, whether during electoral campaigns or otherwise. He stresses that this is done not just as a means of gaining publicity but, more impor­tantly, "for the public good." He believes the goal of surveys is to make the results known to the public, which in turn leads to dialogues in the mass media and, it is to be hoped, will have an effect on govern­ment policy-making. "That is the real power of public opinion polls," he asserts. "They help reveal what people really want, and make the government respect their wishes."

Since the New Party began to conduct its own surveys last year, it has been low­ key about publicizing the results-at least, compared with the other parties. "Most of the surveys we do are not intended for public consumption," Fei Hung-tai says. "We don't have to brag. We're well aware that we can only win a lim­ited number of seats in the Legislative Yuan."

Lin Yu-siang of the KMT thinks that what makes political parties' surveys different is diversity of interpretation. A party is bound to have a unique political standpoint. He says that if any given survey were to be carried out again by another body, the resulting fig­ures would be pretty much the same, always supposing that both polls had been carried out professionally. The difference lies in how the figures are interpreted. "Just to show you one example: Based on surveys by the New Party and the DPP, President Lee Teng-hui (李登輝) doesn't seem to have a clear policy for the coun­try's future," Lin says. "But if we did our own survey, it would show that the other two parties' wish to dichotomize the issue is really what makes Lee look as though he doesn't have a clear policy. Some people say he's pro-unification, others pro-independence. But in fact Lee is doing just what most people want: main­taining Taiwan the way it is now. Is there anything wrong with that?"

Gallup's Tim Ting agrees with Lin that a political party should have its own distinct stance. In fact, he thinks every person has an opinion about politics, and the real question is whether that person can put professional ethics above partisan interests once he or she becomes involved with a political organization. Ting does not think that is possible, and so is strongly against the relatively common practice of scholars getting involed with party surveys. The DPP's You Ying-lung, for example, is an associate professor of political science at Soochow University. The New Party's Fei Hung-tai, also an associate professor, teaches statistics at National Chunghsing University. "When scholars are hired by the parties to conduct public opinion surveys, they become 'court scholars, "'Ting claims. "They have only the party's best interests in mind.

Ting thinks that there is still a long way to go before Taiwan uses polls properly. "Where there is democracy, there are public opinion polls, "he says. "But just as true democracy cannot be achieved overnight, neither can objective and fair polling. If the main parties continue to trumpet their surveys, people will soon come to see them as just another political strategy and, what's more, one based on highly selective materials. Gradually the mass media will lose interest in covering them. Have you noticed that coverage of party surveys is already starting to shrink?"

Lin Tung-tai suggests that the main political parties should commission outside organizations to conduct public opinion polls for them. He also encourages the mass media to play a role, thus counterbalancing the political parties'own surveys. But Fei Hung-tai of the New Party thinks otherwise. He believe the media should be very careful when dealing with surveys concerning social and political issues. "They should try to avoid generating conflicts among different social groupings, or tensions across the Taiwan Straits," he says. Meanwhile, the DPP's You Ying-lung has announced that he plans to publish an annual report setting out the results of all the issue-oriented surveys he does for the party. "In the future," he says, "these annuals should prove very useful to anyone who's interested in the opinions of the Taiwan public."

Taiwan's first direct presidential election is just around the corner, and when people cast their votes on March 23 it will mark a milestone on Taiwan's progress toward democratic consolidation. Every election is preceded by a warm-up, however-it is not entirely inappropriate that public opinion polls conducted by the major parties seem destined to generate more heat than light.