Public opinion polling in its modern sense, however, did not emerge until the 1950s after the ROC central government moved to Taiwan. The first opinion poll was conducted by the Hsin Sheng Daily in 1952, regarding the Sino-Japanese Treaty then under negotiation. Some 80,000 out of 280,000 questionnaires were returned. In spite of the scale of this poll, it consisted mostly of questions on the contents of the treaty and can hardly count as an opinion poll.
Two years later, the United Daily News followed with a poll on a controversial issue: whether or not simplified Chinese characters should be adopted. In 1956, the Hsin Sheng Daily set up an opinion polling department. Under the auspices of this department, which was dissolved in 1962, about a hundred polls were taken.
The 1950s also witnessed the rise of one of the leading pollsters, the Society of Public Opinion Polls. Organized in 1948 by Wu Wang-chi, a legislator, it had in fact been the only active force in the polling business for almost twenty years after the dissolution of the polling section of the Hsin Sheng Daily. As of September 1987, the Society has conducted about 500 polls. In conjunction with its polling efforts, the organization in 1977 started the journal Public Opinion Monthly to provide readers with research reports on polling theory and methodology as well as poll results.
Telephone surveys have become the most frequently used approach to public opinion polling.
Public opinion polling up to the 1980s in general smacks of amateurism, and as a result received little more than nominal recognition from the public. This may have been due as much to the homogeneous climate of opinion during that period as to the lack of knowledge concerning the role of polls as an important institution in a democracy. Until recent years, there was a tendency to equate public opinion with the opinion of the mass media and legislators; no felt need thus existed for other channels of opinion. The majority of the polls dealt with essentially non-controversial issues such as attitudes toward health insurance, railroad services, and disco dancing. Some of the polls taken by the Society have in fact been nothing more than marketing surveys.
The 1980s, especially during the last three years, mark a sharp upturn in polling activities. Several factors may account for the accelerated growth. First, two of Taiwan's leading newspapers, the United Daily News and the China Times should be given credit for taking the lead in giving a boost to polling. In fact, two thirds of the polls conducted in the past three years have been sponsored by the two dailies. The newspapers, which have traditionally conceived of themselves as organs of the people, have seen in polls a new avenue for performing services to democratic development. Also, the two newspapers, which have scrambled for the same share of the reader market, plunged into polling for competitive reasons. After the United Daily News started pre-election polls in 1983, the China Times joined the race, feeling that it could not afford to lose the news war to its competitor.
Second, the newly-found attentiveness to polls in a way echoes the increasingly pluralistic mood of society. The 1980s has seen sweeping changes affecting all sectors of life in Taiwan. In the span of only one year, the Emergency Decree that activated martial law has been lifted, the ban on visiting relatives on the Chinese mainland has been essentially removed, and new political parties have been allowed to form.
Perhaps even more significantly, voices from all walks of life are now eager to be heard, sometimes even in street demonstrations. Government agencies and other institutions have become increasingly sensitive about accounting for their actions, and have also found it desirable to cultivate more channels for suggestions and recommendations about policy decisions. Polls have a function in facilitating this communication process.
About 180 polls have been taken since 1985, compared to an annual average of about 20 during the 1970s. Newspapers have been the prime movers. Media- sponsored polls, as in other countries, originated as pre-election forecasts. Starting 1983, the United Daily News began questioning voters on their favorite candidates. From then on, the newspaper, which set up its own polling division, has conducted polls on a regular basis.
The China Times joined the field during the 1985 local mayoral election. Encouraged by the favorable reception of these pre-election polls, the Times escalated its polling operations during the 1986 election for the Legislative Yuan under the supervision of a team of social scientists who introduced sophisticated statistical methods. They took a total of twenty-three polls using telephone interviews, focusing on pre-election forecasts and related issues. The number of respondents rose to an average of 1,500 persons for each poll.
In addition to the two newspapers and the Society of Public Opinion Polls, the government has also joined the pollsters. Special mention should be made of seven large-scale opinion surveys sponsored by the Research, Development, and Evaluation Commission (RDEC) of the Executive Yuan. Since 1975 the RDEC has taken nation-wide opinion polls with samples of between 4,000 and 6,000 persons. These surveys covered a wide variety of topics such as evaluation of policies, subjective perceptions of personal well-being, and voting intentions. The results have been used for policy-making purposes.
Earlier this year, Chao Shao-kang, a newly elected legislator, set up another polling institute with the assistance of fifteen social scientists and statisticians. Called the Public Opinion Research Foundation, it has taken 8 polls as of October 1987.
Political topics now account for almost two thirds of the issues covered by polls, which is a reverse in the trend found during the period prior to 1980. This in a sense mirrors the rising interest in political issues confronting contemporary society. Many polls document the attitudes toward current controversies and sensitive issues such as allowing visits to relatives on the Chinese mainland, the new Civic Organization Law, and changes in the structure of the National Assembly.
Especially noteworthy is the increasing number of polls on environmental issues, which reflects a greater public awareness of new social problems. In contrast to the polls taken before the 1980s, recent polls show more diversity in their coverage and exhibit a willingness to venture into controversial topics. Generally speaking, however, media-sponsored polls have mainly dealt with newsworthy issues, with pre-election polls being the main thrust. But polls on more light-hearted matters, such as whether the ROC should host the Miss Universe Contest, have been taken as well.
Public opinion as revealed by recent polls reflects the dynamics of social transformation underway on the island. Significantly, the middle class has become the backbone of society, and its attitudes prevail in the polls. According to the surveys of the RDEC, more than 50 percent of the electorate label themselves as belonging to the middle class. The RDEC surveys show that the dominant attitude tends to converge toward the middle tilting toward pro-government positions.
Considering the short history of polls in the ROC, however, public polling has yet to establish itself as a prestigious institution in Taiwan. Technical problems abound. Pollsters feel uneasy about their methodology, which has been borrowed primarily from their counterparts in Western societies, and they are groping for a Chinese way of asking questions during polling. In addition, the lack of an adequate sampling frame may introduce some bias into the sample selection procedure.
More importantly, Chinese have a traditional suspicion of strangers knocking at the door and asking questions. A recent issue of Public Opinion Monthly published the results from a poll that asked about attitudes toward answering sensitive questions, which many of the respondents equated with political and personal matters. To the astonishment of pollsters, the poll found that a mere 15 percent of the respondents said that when confronted with sensitive questions they would tell the truth! Experience shows that two or three days before an election, 60 to 70 percent of the respondents shy away from revealing their voting intentions.
Polling is one of the long-standing institutions in democracies, and the history of polls in the ROC in a sense mirrors the process of its political and social development. It is therefore by no means accidental that polls, along with the recent strides towards more democratization, are coming to the fore. —(Dr. Chung Wei-wen is an associate professor of journalism at National Chengchi University, Taipei).