HKU POP SITE and Civic Exchange releases the 2nd Legislative Council election rolling survey resultBack

Preamble

This year, the Public Opinion Programme (POP) at the University of Hong Kong and the Civic Exchange are jointly conducting a series of public opinion surveys on the 2004 Legislative Council direct elections, including the rolling surveys starting from August 4. This press release mainly highlights the key findings in the rolling surveys conducted from August 5 to 9. The detailed figures are available in the websites of POP and Civic Exchange at http://hkupop.pori.hk and http://www.civic-exchange.org respectively.

The focuses of the rolling survey are to understand the voters' latest propensity to vote, their level of knowledge and support towards the candidate lists in each constituency, as well as their general opinion on elections. The POP Team started the rolling survey mechanism right after the nomination closed on August 4, and will frequently release the survey result via the "HKU POP SITE" from August 9 for public consumption. The work will continue on a daily basis until the day before the election.

Methodology

Since 1995, the POP Team has conducted rolling surveys before the Legislative Council elections. A particular number of interviews will be completed each day, while figures obtained will be consolidated a few days later. Figures will be added and dropped on a "plus-one-minus-one day" basis, and then released consecutively.

334 interviews with registered voters are expected to be completed every day in this year's rolling survey. In other words, more than 1,000 successful cases will be accumulated in 3 days and form a complete survey. Figures will then be released on a rolling aggregate basis. However, since the sample size in individual constituencies is relatively smaller, the relevant figures will roll every 5 days, with 150-250 registered voters to be covered each time, depending on the size of the geographical constituency. The closer the election, the larger the sample size will become in order to minimize the errors.

Same as the other telephone interviews conducted by the POP Team, all interviews in this rolling survey were conducted by telephone interviewers under close supervision. The target population was Cantonese-speaking population of Hong Kong aged 18 or above. Telephone numbers were first drawn randomly from the residential telephone directories as "seed numbers", from which another set of numbers was generated using the "plus/minus one/two" method, in order to capture the unlisted numbers. Numbers were then mixed in random order to produce the final telephone sample. When telephone contact was successfully established with a target household, one of the registered voters available was selected using the "next birthday rule", which selected the person who had his/her birthday next from all those present. Besides, all the raw data obtained in the rolling survey had been adjusted according to the age and gender distributions of voters in 5 geographical constituencies reported by the government. All data obtained in the rolling survey will be available in the "HKU POP SITE".

Feature questions

The latest territory-wide rolling surveys (i.e. August 7 to 9) revealed that, if the voters were to vote, 56% said their main consideration would be the quality of individual candidates, 20% said it would be their support for a specific political party, while 5% claimed the combination of candidates within a list would be their main consideration. Besides, as high as 80% of the voters claimed their propensity to vote would not decrease because the candidate they supported ranked rear on the list, while 73% expressed that their propensity to vote would not increase because the candidate they supported ranked front on the list. The figures obtained in the 4 recent rolling surveys are summarized as follows:

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.

Since a large number of figures have been accumulated in 6 days, we can make some more in-depth analysis concerning the voters' demographics, with 2 particular points highlighted below:

First, by defining new voters as those registered after 2000, results revealed that, although 50% of new voters would consider mainly the quality of individual candidates, a considerable proportion of new voters also chose "support for a political party" and "combination of candidates within a list" as their main considerations, accounting for 32% in total. Regarding the old voters, the figures were 56% and 24% respectively. Apparently, the old voters would put more emphasis on the quality of candidates. The relevant figures are summarized as follows:

Results also revealed that, the propensity to vote among the highly educated voters was more easily affected by the order of candidates within a list. Figures showed that, the propensity of 24% voters with tertiary education or above would increase because the candidate they supported ranked in the front on the list, while 14% with decreased propensity because the candidate they supported ranked in a rear position. On the other hand, among voters with primary education or below, only 12% said their propensity would increase because the candidate they supported ranked front on the list, while 5% with decreased propensity because the candidate they supported ranked rear. The figures are summarized as follows:

Knowledge of the election

Regarding the knowledge level of voters towards the candidates' list, the latest territory-wide rolling surveys (i.e. August 7 to 9) revealed that, 52% had no idea on any list in their geographical constituency, while 48% knew at least one list. The rolling surveys in individual constituencies from August 5 to 9 showed that, the ignorance level towards the candidates' list was the highest in Kowloon East, with 59% claiming they had no idea at all, while that of New Territories East was the lowest, at 48%. The relevant figures are summarized as follows:

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.
^ Since overlapping exists among these 3 answers, the total would add up to over 100%.

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.
^ Since overlapping exists among these 3 answers, the total would add up to over 100%.

Propensity to vote

The territory-wide rolling surveys from August 7 to 9 revealed that, the registered voters' propensity to vote was 85%. Nevertheless, this percentage should never be taken as a projection of the actual turnout rate because many people who claimed they would vote at this stage would eventually not vote. These figures are summarized as follows:

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.

The rolling surveys in individual constituencies from August 5 to 9 showed that, the propensity to vote among voters in 5 constituencies varied within a narrow range of 82% to 85%. The relevant figures are summarized as follows:

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.

Stray voters

Since many voters still had not decided how to cast their vote, while stray voters would always have significant influence on the election result, the POP Team therefore further look into the distribution of these votes in 5 constituencies. Firm voters are defined as those who said they would definitely and most likely vote, and have also decided to vote for which list, while stray voters are defined as those who said they would definitely and most likely vote, but have not decided to vote for which list. The rolling surveys in individual constituencies from August 5 to 9 showed that, there were around 23% - 32% of stray votes in each of the 5 constituencies, while the largest proportion of stray votes was found in Kowloon East (32%) and the smallest in New Territories East (23%). The figures are summarized as follows:

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.

Candidate strength

Concerning the rolling surveys in individual constituencies from August 5 to 9, the support ratings of various lists are summarized as follows:

* "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified.

The support ratings shown above have not excluded those who had not decided and claimed not to vote. Readers may exclude this group of uncertain voters, reallocate the vote share according to the new base, with the assumption that these uncertain people would not vote. The number of seats obtained by each list can then be deduced based on the Largest Remainder Formula. Yet, one must take note of the sampling errors.

The latest rolling survey revealed that, in Hong Kong Island constituency, Eu Yuet Mee Audrey and Ho Sau Lan Cyd of pro-democracy camp became the most popular list for the time being, obtaining 25% of vote shares. The 2nd place fell to independent Hsu Lai Tai Rita, obtaining 19% of vote shares. Besides, 16% of the voters would vote for Yeung Sum, Lee Chu Ming Martin, Lai Chi Keong of DP, while 6% would vote for Ma Lik, Choy So Yuk, Chung Shu Kun Christopher, Yeung Wai Foon, Lee Yuen Kwong, Cheung Kwok Kwan of DAB.

Regarding the candidates' strength in Kowloon East, 23% of the voters would vote for Chan Yuen Han, Lam Man Fai, Tang Ka Piu of FTU and DAB while 15% would vote for Li Wah Ming, Wu Chi Wai, Ho Wai To of DP. Meanwhile, independent Cheng Albert Jing Han and To Kwan Hang Andrew obtained 8% of vote shares, while independent Leong Kah Kit Alan obtained 7% of vote shares. Finally, 5% would vote for Chan Kam Lam, Choi Chun Wa, Chan Tak Ming of DAB and FTU.

As regards the candidates' strength in Kowloon West, To Kun Sun James, Chan Ka Wai, Lam Ho Yeung, Ma Kee of DP attained 20% of vote shares. Tsang Yok Sing Jasper, Chung Kong Mo, Lee Wai King Starry of DAB obtained 17% of vote shares. Besides, 11% of voters would vote for Fung Kin Kee Frederick of ADPL and Independent Lau Chin Shek respectively, while Liu Sing Lee of ADPL obtained 2% of vote shares.

Concerning the candidates' strength in New Territories East, Cheng Kar Foo Andrew, Lau Wai Hing Emily, Tong Ka Wah, Wong Sing Chi, Tsoi Yiu Cheong, etc. of pro-democracy camp obtained 34% of vote shares. Meanwhile, Tien Pei Chun James of LP obtained 12% of vote shares, while Lau Kong Wah, Li Kwok Ying, Mok Kam Kwai, Chan Kwok Kai, So Sai Chi, etc. of DAB obtained 11%. Besides, another 6% would vote for Independent Wong Wang Fat Andrew, whereas Tso Wung Wai of HKPA and Leung Kwok Hung of AFA obtained 3% and 2% of vote shares respectively.

New Territories West has the largest number of lists. The findings revealed that, the vote shares obtained by Tam Yiu Chung, Cheung Hok Ming, Leung Che Cheung, etc. of DAB, Ho Chun Yan, Cheung Yin Tung of DP, Leung Yiu Chung, Wan Siu Kin Andrew of NWS, and Chow Liang Shuk Yee Selina, Ting Woo Shou Kenneth of LP were highly comparable, each obtaining 13%, 10%, 10% and 9% respectively. Besides, 6% of voters would vote for Lee Wing Tat and Chan Yuen Sum of DP, Independent Chan Wai Yip Albert, Lee Cheuk Yan and Ip Ngok Fung of HKCTU respectively. Yim Tin Sang, Kong Fung Yi, Tai Yin Chiu, etc. of ADPL and independent Chow Ping Tim obtained 3% and 1% of vote share. Less than 1% of voters would vote for Independent Char Shik Ngor Stephen.

It must be stated that since many voters have not yet made their choice, while most candidates have not begun their full-scale election campaign, the influence of stray voters on votes obtained by candidates is yet to be observed.

For the results of the other rolling surveys, please refer to the websites of POP and Civic Exchange at http://hkupop.pori.hk and http://www.civic-exchange.org respectively, in order to save time and space.

Commentary

Regarding the rolling poll results, Robert Ting-Yiu Chung, Director of Public Opinion Programme, observed: "According to the result of our rolling polls almost spanning over the first week of the campaign period, there is a slow growth of in voters' knowledge of the election, as reflected in their knowledge of the candidate lists. There was an increase of about 2 percentage points each day in the proportion of voters who knew at least one list in their constituency, the situation more or less the same with other indicators. The propensity to vote has also increased a bit. The latest proportion of voters who said they would definitely vote was 58%, while 85% said they would definitely or probably vote. Concerning the candidates' strength in individual geographical constituencies, since there are only two rolling samples at this moment, changes are not yet apparent. However, an upward trend in propensity to vote was observed among voters in Hong Kong Island and New Territories East. Besides, figures over these 6 days revealed that the propensity to vote among voters with higher education level was more easily affected by the order of candidates within a list, while new voters would put more emphases on the combination of candidates within a list and the candidates' political background, when compared to the old voters."

Other points to note

POP will release the latest results of the rolling surveys every day at 2 pm via our POP Site. Shall anyone have any question regarding the research design of the surveys published in the POP Site, members of the POP Team will be happy to answer them, but we will not further comment on the findings. Shall any person or journalist have any other questions, please email them to us at <[email protected]>. Please note that everything carried in the POP Site does not represent the stand of the University of Hong Kong. Dr Robert Ting-Yiu Chung, Director of Public Opinion Programme, is responsible for everything posted herewith, except for column articles which represent the stand of their authors.

**This is a translation of the Chinese press release; should there be any discrepancy or ambiguity, please refer to the Chinese version.