POLITICAL POLLING IN TAIWAN: FEELING THE PULSE OF

C O N F I D E N T I A L AIT TAIPEI 002748 SIPDIS SIPDIS E.O. 12958: DECL: 08/14/2026 TAGS: PGOV SCUL TW SUBJECT: POLITICAL POLLING IN TAIWAN: FEELING THE PULSE OF PUBLIC OPINION REF: 2005 TAIPEI 02118 AND PRECEDING CLASSIFIED BY: AIT DIRECTOR STEPHEN M. YOUNG, REASONS: 1.4 (B/D) ¶1. (SBU) SUMMARY: MANY RECENT POLLS HERE HAVE CAPTURED THE CRISIS OF LEGITIMACY FACED BY PRESIDENT CHEN SHUI-BIAN OVER CORRUPTION SCANDALS. PUBLIC OPINION POLLING PLAYS AN IMPORTANT POLITICAL ROLE IN TAIWAN AS A TOOL FOR MEASURING, AND IN SOME CASES INFLUENCING, PUBLIC OPINION. NUMEROUS ORGANIZATIONS IN TAIWAN CONDUCT OR CONTRACT OUT POLLING ON HOT-BUTTON ISSUES, INCLUDING CROSS-STRAIT RELATIONS, LEADERSHIP STANDINGS, AND REACTIONS TO CORRUPTION SCANDALS, OR ON LONG-TERM TRENDS SUCH AS "TAIWANESE IDENTITY" OR SUPPORT FOR TAIWAN INDEPENDENCE. THE OVERALL RELIABILITY OF TAIWAN POLLING HAS IMPROVED IN RECENT YEARS, BUT LIKE THE PERIPATETIC MEDIA HERE, BIASES AND SUBJECTIVENESS REMAIN A PROBLEM. END SUMMARY. AN INTRODUCTION TO POLLING IN TAIWAN -------------- ¶2. (SBU) PUBLIC OPINION POLLING IS HIGHLY DEVELOPED IN TAIWAN AND IS USED EXTENSIVELY TO GAUGE SUPPORT FOR POLITICIANS, PARTIES, AND POLICES, AS WELL AS TO MEASURE TRENDS ON KEY ISSUES SUCH AS NATIONAL IDENTITY AND CROSS-STRAIT RELATIONS. THE QUALITY OF POLLING -- WHICH IS CONDUCTED BY MEDIA ORGANIZATIONS, THINK TANKS AND UNIVERSITIES, POLITICAL PARTIES, AND GOVERNMENT AGENCIES -- VARIES WIDELY BUT APPEARS TO HAVE IMPROVED SOMEWHAT IN RECENT YEARS, THOUGH SOME PROBLEMS PERSIST. MEDIA POLLS -------------- ¶3. (SBU) SEVERAL MEDIA ORGANIZATIONS THAT PUBLISH SURVEY RESULTS HAVE IN-HOUSE POLLING CENTERS. POLLS CONDUCTED BY THESE CENTERS ARE COMMONLY USED TO REINFORCE HEADLINES ON HOT-BUTTON ISSUES, INCLUDING ELECTION CAMPAIGNS, IN ORDER TO INCREASE AUDIENCE SHARE. MOST OF THESE POLLS ARE QUICK 10-15 MINUTE TELEPHONE SURVEYS OF APPROXIMATELY 1,000 RESPONDENTS CONDUCTED OVER THE COURSE OF ONE TO TWO EVENINGS. SOOCHOW UNIVERSITY PROFESSOR EMILE SHENG, WHO WORKS EXTENSIVELY WITH MEDIA POLLS, EXPLAINED THAT EVEN THOUGH THE MEDIA IN TAIWAN IS OFTEN POLITICALLY-BIASED, MOST MEDIA POLL CENTERS TRY TO BE OBJECTIVE AND METHODICAL IN THEIR POLLING. FOR EXAMPLE, MANY POLL CENTERS WEIGHT RESULTS BASED ON DEMOGRAPHIC INFORMATION IN ORDER TO GET A MORE ACCURATE SNAPSHOT OF PUBLIC OPINION. SOME, PARTICULARLY THE TVBS NEWS STATION, PUBLISH CROSS-TABULATIONS BREAKING THEIR RESPONSES INTO RELEVANT CATEGORIES, SUCH AS PARTY AFFILIATION, GENDER, AND AGE, TO FACILITATE MORE DETAILED ANALYSIS OF POLL RESULTS. SHENG ADVISED AIT TO CHECK WHETHER THE VARIOUS MEDIA POLLS ALL SHOWED SIMILAR RESULTS TO JUDGE WHETHER THE POLLING WAS ACCURATE. ¶4. (C) DESPITE SUCH EFFORTS AT OBJECTIVITY, HOWEVER, EVEN THE MOST METICULOUS MEDIA POLLS ARE SUBJECT TO POLITICAL BIASES THAT REDUCE THEIR RELIABILITY, CHENGCHI UNIVERSITY ELECTION STUDY CENTER (ESC) DIRECTOR YU CHING-HSIN TOLD AIT. BOTH HE AND EMILE SHENG SAID THAT PUBLIC PERCEPTION OF THE POLITICAL POSITIONS OF INDIVIDUAL MEDIA OUTLETS AFFECTS BOTH THE RESPONSE RATE AND INDIVIDUAL RESPONSES RECEIVED, AS, FOR EXAMPLE, SOME GREEN SUPPORTERS WILL EITHER REFUSE TO RESPOND OR RESPOND DECEPTIVELY TO QUESTIONS FROM A MEDIA OUTLET THAT IS VIEWED AS SUPPORTING THE BLUE CAMP. SHENG SUGGESTED THAT SUCH ASSOCIATIONAL BIASES ALONE TYPICALLY SKEW POLLING RESULTS BY 3-5 PERCENT. ¶5. (C) DIRECTOR YU NOTED THAT MEDIA POLL CENTERS OFTEN FAIL TO CONTROL FOR A WIDE RANGE OF METHODOLOGICAL DEFECTS ROOTED IN PROCESSES SUCH AS HOW RESPONDENTS ARE SELECTED, HOW QUESTIONS ARE WORDED AND ORDERED, HOW LONG THE SURVEY PERIOD IS, AND WHEN SURVEYS ARE CONDUCTED. TOGETHER, THESE AND OTHER DEFECTS LEAD TO CONSIDERABLE VARIATION IN SURVEY FINDINGS. FOR EXAMPLE, POLLS CONDUCTED IN MID-JUNE SHOWED A WIDE RANGE IN PRESIDENT CHEN SHUI-BIAN'S APPROVAL RATINGS, VARYING FROM 18 PERCENT (TVBS) TO 28 PERCENT (CHINA TIMES). EVEN SO, DIRECTOR YU NOTED THAT POLLS CONDUCTED BY A PARTICULAR ORGANIZATION USING THE SAME QUESTIONS OVER A PERIOD OF TIME ARE STILL USEFUL, BECAUSE THEY GIVE A FAIRLY ACCURATE IDEA OF TRENDS OVER TIME. ACADEMIC POLLS -------------- ¶6. (C) POLLS CONDUCTED BY ACADEMIC INSTITUTIONS, PARTICULARLY CHENGCHI UNIVERSITY'S ESC, SHIH-HSIN COLLEGE, AND ACADEMIA SINICA, GENERALLY YIELD THE MOST RELIABLE RESULTS BECAUSE ACADEMIC INSTITUTIONS APPLY GREATER SCIENTIFIC RIGOR IN THEIR POLLING AND ARE ALSO PERCEIVED TO BE APOLITICAL. THEY DO, HOWEVER, TEND TO INVESTIGATE ISSUES THAT ARE LESS TIME-SENSITIVE IN NATURE. TO ILLUSTRATE THE GREAT EMPHASIS ACADEMIC INSTITUTIONS PLACE ON ACCURATE AND OBJECTIVE POLLING, DIRECTOR YU TOLD AIT THAT THE ESC SOMETIMES DECLINES TO CONDUCT POLLS REQUESTED BY GOVERNMENT AGENCIES SUCH AS THE MAINLAND AFFAIRS COUNCIL (MAC) UNLESS THE ESC IS GIVEN CONTROL OVER QUESTION CONTENT AND OVERALL POLL DESIGN. ¶7. (U) IN ADDITION TO POLLS CONDUCTED VIA PHONE, ACADEMIC INSTITUTIONS ALSO CONDUCT FACE-TO-FACE POLLS, WHICH YIELD FAR MORE DETAILED AND ACCURATE RESULTS THAN TELEPHONE POLLS. THIS IS BECAUSE IN FACE-TO-FACE POLLS, SURVEY ORGANIZATIONS ARE ABLE TO BETTER TARGET DEMOGRAPHIC GROUPS BY USING HOUSEHOLD REGISTRATION DATA, HAVE MORE TIME TO CONDUCT INDIVIDUAL INTERVIEWS, AND HAVE GREATER CONTROL OVER THE SURVEY PROCESS. HOWEVER, GIVEN THE HIGH COST AND LOGISTICAL REQUIREMENTS, FACE-TO-FACE POLLS ARE ONLY CONDUCTED SPORADICALLY. POLITICAL PARTY POLLS -------------- ¶8. (C) POLITICAL PARTIES CONDUCT OR CONTRACT THEIR OWN INTERNAL POLLS IN ORDER TO PINPOINT PUBLIC OPINION ON POTENTIAL CAMPAIGN THEMES AND OTHER POLITICAL ISSUES, AND TO MONITOR SUPPORT FOR CANDIDATES IN ELECTIONS. MOST INTERLOCUTORS TELL AIT THAT THESE POLLS ARE GENERALLY VERY ACCURATE. TO AVOID BIASED RESPONSES, POLLS CONDUCTED BY A POLITICAL PARTY ARE OFTEN CONDUCTED IN THE NAMES OF COOPERATING PROFESSORS RATHER THAN IN THE NAME OF THE PARTY ITSELF. UNLIKE THE DPP, THE KMT DOES NOT HAVE ITS OWN INTERNAL POLL CENTER AND THEREFORE COMMISSIONS MEDIA POLLING CENTERS LIKE TVBS, ERA, AND ETTV TO CONDUCT ITS INTERNAL POLLS, ACCORDING TO TVBS POLL CENTER DIRECTOR WANG YEH-DING. ¶9. (C) POLITICAL PARTIES AND THEIR ASSOCIATED THINK TANKS ALSO CONDUCT AND SELECTIVELY PUBLISH PUBLIC OPINION POLLS TO MEET POLITICAL NEEDS. FOR EXAMPLE, THE DEEP-GREEN TAIWAN THINKTANK PUBLISHED A POLL TWO DAYS BEFORE THE JULY 27-28 SUSTAINABLE ECONOMIC DEVELOPMENT CONFERENCE PURPORTING TO SHOW THAT 70 PERCENT OF THE PEOPLE IN TAIWAN BELIEVE REGULAR DIRECT CROSS-STRAIT FLIGHTS WOULD ENDANGER THE SECURITY OF TAIWAN. MOST POLLING EXPERTS CONSULTED BY AIT AGREED THAT PUBLISHED PARTISAN OPINION POLL RESULTS ARE SUSPECT AND NOT RELIABLE, BECAUSE THEY SELECTIVELY PUBLISH RESULTS TO INFLUENCE PUBLIC OPINION. RECENT PUBLIC OPINION POLLS: POLITICAL LEADERS -------------- - ¶10. (U) RECENT MEDIA POLLING HAS FOCUSED LARGELY ON PUBLIC SUPPORT FOR POLITICAL LEADERS AND PARTIES AND ON REACTIONS TO SCANDALS SURROUNDING PRESIDENT CHEN, INCLUDING WHETHER THE PRESIDENT SHOULD STEP DOWN OR BE RECALLED AND WHETHER THERE SHOULD BE A NO CONFIDENCE VOTE AGAINST PREMIER SU. PRESIDENT CHEN'S APPROVAL RATING, WHICH WAS 40 PERCENT AT THE TIME OF THE MARCH 2004 PRESIDENTIAL ELECTION, HAS HOVERED AROUND 20 PERCENT SINCE SCANDALS FIRST BECAME AN ISSUE IN THE RUN UP TO THE DECEMBER 2005 LOCAL ELECTIONS. ALTHOUGH POLLING HAS SHOWN SUBSTANTIAL SUPPORT FOR CHEN STEPPING DOWN OR BEING RECALLED, LESS THAN ONE THIRD OF THE PUBLIC SUPPORTS HOLDING A NO CONFIDENCE VOTE AGAINST PREMIER SU, WHICH IS WIDELY VIEWED AS A PURELY POLITICAL MOVE. THE LOW PUBLIC SUPPORT FOR THIS MOTION IS ONE REASON WHY THE OPPOSITION IS UNLIKELY TO CARRY THROUGH ITS EARLIER THREATS TO CALL A NO CONFIDENCE VOTE. ¶11. (C) REACTING TO PRESSURE FROM THE DEEP BLUE, KMT CHAIRMAN MA YING-JEOU SUPPORTED A VOTE IN THE LEGISLATURE IN JUNE TO RECALL PRESIDENT CHEN DESPITE LACKING THE VOTES NEEDED TO PASS THE MOTION. THE FAILED RECALL VOTE DAMAGED MA, PULLING HIS APPROVAL RATING DOWN FROM A HIGH OF 67 PERCENT AT THE TIME OF HIS MARCH 2006 U.S. VISIT TO A NEW LOW OF 50 PERCENT. ACADEMIA SINICA RESEARCH FELLOW HSU YUNG-MING SUGGESTED THAT THIS DECLINE IN APPROVAL WAS DUE BOTH TO DEEP BLUE DISSATISFACTION WITH MA'S WEAK LEADERSHIP AND TO ALARM AMONG MODERATES AT HIS SHIFT TOWARD A HARDER LINE DURING THE RECALL EFFORT. SOOCHOW UNIVERSITY'S EMILE SHENG SUGGESTED THAT TO BOOST HIS PUBLIC APPROVAL RATING, THE EASY-GOING MA NEEDS TO EXERT FIRM LEADERSHIP TO OVERHAUL THE KMT IN ORDER TO HELP THE PARTY SHED ITS CORRUPT IMAGE. PUBLIC OPINION POLLS AND LONG-TERM TRENDS -------------- ¶12. (C) POLLING ON IDENTITY ISSUES IS A REGULAR SUBJECT OF PUBLIC OPINION POLLING IN TAIWAN. SURVEYS CONDUCTED BY THE ESC OVER THE PAST 14 YEARS REVEAL A STEADY INCREASE IN THE PERCENTAGE OF PEOPLE PERCEIVING THEMSELVES AS "TAIWANESE." DESPITE THIS STEADILY GROWING TAIWANESE CONSCIOUSNESS, HOWEVER, MOST PEOPLE CONTINUE TO SUPPORT MAINTAINING THE CROSS-STRAIT STATUS QUO RATHER THAN EITHER TAIWAN INDEPENDENCE OR UNIFICATION WITH THE PRC. NATIONAL TAIWAN UNIVERSITY PROFESSOR CHU YUN-HAN TOLD AIT THAT THIS APPARENT PARADOX CAN BE EXPLAINED BY A GRADUAL SHIFT AMONG YOUNG PEOPLE IN TAIWAN AWAY FROM ALIGNMENT WITH EITHER OF THE TWO TRADITIONAL IDEOLOGICAL CAMPS OF INDEPENDENCE ("GREEN") OR UNIFICATION ("BLUE") TOWARDS A MORE OPEN-MINDED PRAGMATISM. ¶13. (C) ESC SURVEYS ALSO SHOW THAT SUPPORT FOR THE RULING DPP HAS STEADILY DECREASED OVER THE PAST TWO YEARS, WHILE THE KMT HAS SEEN AN UPSWING IN SUPPORT, PARTICULARLY AFTER THE JULY 2005 ELECTION OF MA YING-JEOU TO THE KMT CHAIRMANSHIP. ACCORDING TO A POLL CONDUCTED BY THE PRO-BLUE UNITED DAILY NEWS IMMEDIATELY AFTER THE FAILED JUNE 2006 PRESIDENTIAL RECALL VOTE, DISSATISFACTION WITH THE DPP HAD RISEN TO A NEW HIGH OF 68 PERCENT DUE LARGELY TO THE RECENT SPATE OF SCANDALS. COMMENT -------------- ¶14. (SBU) THE READY AVAILABILITY OF PUBLIC OPINION POLLS IN TAIWAN, PARTICULARLY TIME-SENSITIVE AND POLITICALLY-RELEVANT POLLS CONDUCTED BY THE MEDIA, IS A MIXED BLESSING. METHODOLOGICAL SHORTCOMINGS, POLITICAL BIASES AND THE FACT THAT SOME POLLS MAY BE INTENDED TO INFLUENCE OPINION COMPLICATES ASSESSMENT OF POLLING ACCURACY. IF LIMITATIONS ARE TAKEN INTO CONSIDERATION, HOWEVER, POLLS CAN PROVIDE A WEALTH OF INFORMATION ON BOTH SHORT-TERM AND LONG-TERM POLITICAL TRENDS IN TAIWAN. ¶15. (U) THIS CABLE WAS DRAFTED BY POL INTERN EUGENE CHEN. YOUNG