Back

Note: This article represents the view of the author and not the University of Hong Kong.

With the official start of the election campaign period, coordination among candidate lists has finally come to an end. From the perspective of independent public opinion research work, the end of the nomination period means the beginning of another stage of public opinion surveys in the elections. The author would like to summarize the pre-election public opinion survey work here, in order to provide some materials for the civic education.

Since the development of the political party is still in an early stage, various camps are having coordination among their lists either openly or secretly when the election is approaching. In societies where party politics are more developed, the candidate lists will generally be determined by the democratic mechanism within the party. Since there are no major parties in Hong Kong, coordination among camps forms the basis of integration among lists.

The term "pan-democracy camp" appears in Hong Kong recently. The concept should originate from the concept of "pan-blue" and "pan-green" in Taiwan. "Pan-blue" camp appears in Taiwan because the two parties, Kuomintang and People First Party, are having equal status and cannot be merged. (Terms like "pan-green" and "pan-purple" are in fact corresponding to the development of the concept of "pan-blue", instead of meeting one's own needs.) Therefore, the concept of "pan-democracy camp" itself consists of the elements of political immaturity.

In view of the forthcoming Legislative Council Election, both the pro-China camp and the pan-democracy camp have admitted in public that they have conducted several public opinion surveys, as references for the candidate listing and combination. The citizens most probably will have no access to them since they are for internal reference only. As the public opinion survey researcher, the author welcomes more surveys from the political parties, including election surveys, so as to master the public opinion and the survey techniques in a better way. This is of course the political parties' right not to release these internal surveys to the general public. Yet, if the political parties openly quote these survey results to demonstrate to the public their decisions are correct, they should explain their survey methods and relevant figures clearly according to the general professional code of conduct widely accepted in the international world.

In the author's view, different political parties may have adopted different survey methods so as to study the possibilities to win under different combinations. The author would like to introduce a simple, practical and economical way of questions' phrasing, which is applicable before the nomination period.

All candidates who might take part in the election in a certain constituency, irrespective of their combinations, would first be listed out. Second, the respondents would be asked to choose one candidate from the list. Third, the respondents would be asked to choose another candidate among the remaining candidates if their first choice dropped out from elections. After that, the respondents would then be asked to choose again another candidate among the remaining candidates if both the respondents' first and second choices dropped out from elections. The questions could be roughly phrased as below, assuming we would ask 3 questions:

[Q1] If the Legislative Council elections were to be held tomorrow, and you had the right to vote, which candidate or list would you vote for?
[Q2] If the above candidate dropped out from elections, which one of the remaining would you vote for?
[Q3] If the above candidate also dropped out from elections, which one of the remaining would you vote for?

This method can in fact be simplified by asking the respondents to list out their first, second, third… choices among the given candidates or lists, which means to put the candidates or lists in a ranked order. However, judging from the flow of interviews, it will be more convenient and closer to real-life situations to ask the respondents question by question. The advantage of this method is that we need not list out all permutations, as otherwise, the whole survey will become invalidate if some preset candidates drops out from elections, unless those who do not participate are the least influential candidates.

Based on the answers obtained from the above questions, we can then deduce the strength and change of source of votes in each candidate list under different combinations and scenarios. Take the survey on candidates' strength in Hong Kong Island constituency conducted by the author on 24 May as an example, we have listed out 10 candidates who might take part in the election in that constituency, including Martin Lee, Yeung Sum, Jasper Tsang, Ma Lik, Choy So Yuk, Rita Fan, Audrey Eu, Alan Leong, Cyd Ho and Tsang Kin Shing (Table 1). At the end, both Jasper Tsang and Alan Leong did not run in that constituency. However, since the questionnaires have recorded the respondents' second and third choices, all the votes obtained by Jasper Tsang and Alan Leong could be allocated to the remaining 8 candidates.

Besides, if we want to estimate the result of a party in a certain constituency with separate or combined lists, we can just add up the support ratings obtained by each candidate and use the Largest Remainder Formula to deduce the result under different combinations. Of course, a more accurate method is to simultaneously interview those respondents who will change their preferences when the ranking of the candidate they support changes. Yet, survey reveals that this is applicable only to a minority of respondents.

The author would like to use the examples of Kowloon East and Kowloon West to illustrate the above survey method for two reasons: (1) Kowloon East and Kowloon West have the smallest number of seats, thus illustration can be easier. (2) Completely different development can be seen in these two constituencies after the nomination has closed, evolving to two extremes.

First, let us look at Kowloon West. Table 2 reveals that, according to the survey conducted by the author from 24 to 27 June, the differences of candidates' strength were apparent after excluding the votes obtained by Cyd Ho and Selina Chow. Since there have been clear-cut distribution of votes and Jasper Tsang has a stable source of votes, thus, no matter the pan-democracy camp has a combined or separate list, it seems it is impossible for them to occupy 4 seats. Yet, combining the list may help to build up an image of solidarity.

Let us now come to Kowloon East. Table 3 reveals that, according to the survey conducted by the author from 9 to 11 July, basing on the results obtained, if Chan Yuen Han, Lam Man Fai and Chan Kam Lam teamed up in a list, the 45% of vote shares obtained can only occupy 2 seats, with 5% remaining. However, if Chan Yuen Han and Lam Man Fai teamed up in a list, and Chan Kam Lam ran in the election in another list, the 36% obtained by the former could occupy 2 seats, while Chan Kam Lam could occupy the third seat with his 9% of vote shares. The possibility to win was one time higher than the first case. If Chan Yuen Han teamed up with Chan Kam Lam, the 42% of vote shares obtained by them could occupy 2 safe seats, while Lam Man Fai would definitely fail with his 3% of vote shares. Thus, viewing from the perspectives of DAB and FTU, while Chan Yuen Han and Lam Man Fai team up in a list, and Chan Kam Lam participates in another list will be the wisest choice. Yet, Albert Cheng runs in Kowloon East constituency out of everyone's expectation. If part of his source of votes overlaps with that of DAB and FTU, it will render the above deduction method invalidate.

Of course, even though the above deduction method is completely valid, this can only reflect the candidates' strength during the survey period. When the candidate lists have been confirmed and the election campaign officially begins, strategies such as consolidating "iron" votes, opening source of votes, fighting for the support of stray voters, as well as instructing voters to vote tactically, will directly affect the election result. The survey method introduced by the author here can only be used as a reference for the pre-election period, which may possibly help the political parties to search for a common method as the basis of coordination.

The above examples show that, if purely viewing from the perspective of electoral strategy, running in the election in either separate or combined lists have their respective merits, as this depends on the candidates' strength and opponents' deployment. Yet, in the author's view, instructing voters to allocate votes and running elections in separate lists mechanically are not in line with the original intent of the democratic system. While the number of seats will keep on increasing and the threshold will lower accordingly, the Largest Remainder Formula will encourage mechanical vote allocation strategy which needs to be further evaluated. If there are no changes in the system, then the author would like the political parties to put more emphasis on the district work so as to introduce district vote allocation mechanism. Or some political parties can consolidate their bases among certain social classes, so as to achieve the same vote allocation effect.

In any case, after this year's Legislative Council Election, the direct election will have a history of 13 years in the Legislative Council. Excluding the 12 months after the handover when there was no Legislative Council, it just happens to have a history of 6 years both before and after the handover. It should be time to have an evaluation on the proportional representation system and the Largest Remainder Formula.

Table 1: Comparison between the surveys' preset lists and the real-life lists

	Candidates within the scope of survey and running in elections	Candidates within the scope of survey but not running in elections	Candidates beyond the scope of survey but running in elections
Hong Kong Island	Ma Lik, Choy So Yuk, Rita Fan, Tsang Kin Shing, Audrey Eu, Cyd Ho, Yeung Sum, Martin Lee	Jasper Tsang, Alan Leong	Christopher Chung, Yeung Wai Foon, Lee Yuen Kwong, Cheung Kwok Kwan, Chung Chung Fai, Tang Chui Chung, Lai Chi Keong, Wong Kam Fai
Kowloon East	Fred Li, Wu Chi Wai, Ho Wai To, Alan Leong, Chan Yuen Han, Lam Man Fai, Andrew To, Chan Kam Lam	Kenneth Ting, Ma Fung Kwok	Tang Ka Piu, Albert Cheng, Choi Chun Wa, Chan Tak Ming
Kowloon West	Jasper Tsang, Chung Kong Mo, Frederick Fung, Liu Sing Lee, Lau Chin Shek, James To	Selina Chow, Cyd Ho, Eric Wong, John Wong	Starry Lee, Chan Ka Wai, Lam Ho Yeung, Ma Kee, Samuel Iu, Lau Yuk Shing, Leung Suet Fong, Lau Po Kwan
New Territories East	Andrew Wong, Andrew Cheng, Emily Lau, Tong Ka Wah, Wong Sing Chi, Richard Tsoi, Lau Kong Wah, James Tien, Leung Kwok Hung	David Chu, Choy Kan Pui, Gary Fan	Shirley Ho, Ricky Or, Li Kwok Ying, Mok Kam Kwai, Chan Kwok Kai, So Sai Chi, Wong Pik Kiu, Chan Hak Kan, Tso Wung Wai
New Territories West	Lee Wing Tat, Ho Chun Yan, Cheung Yin Tung, Albert Chan, Tam Yiu Chung, Cheung Hok Ming, Selina Chow, Lee Cheuk Yan, Leung Yiu Chung	Tang Siu Tong	Chan Yuen Sum, Yim Tin Sang, Kong Fung Yi, Tai Yin Chiu, Kwun Tung Wing, Leung Che Cheung, Au Yeung Po Chun, Tsui Fan, Chan Han Pan, Andy Lo, Ng King Wah, Kenneth Ting, Ng Tak Leung, Stephen Char, Ip Ngok Fung, Lui Hau Tuen, Siu Shing Choi, Chan Choi Hi, Andrew Wan, Carl Ching, Chow Ping Tim

Table 2: Candidates' strength in Kowloon West (seats=4, threshold=25%)

	Vote shares	Basic seats	Remainder	Seats from remaining votes	Total no. of seats
James To	33%	1	8%	0	1
Lau Chin Shek	20%	0	20%	1	1
Frederick Fung	23%	0	23%	1	1
Liu Sing Lee	2%	0	2%	0	0
Jasper Tsang	18%	0	18%	1	1
Chung Kong Mo	1%	0	1%	0	0

* Survey date: 24 to 27 June, excluding vote shares obtained by Cyd Ho and Selina Chow, while vote shares obtained by Eric Wong, John Wong, and "any list from the pro-democracy camp" have little impact and thus have not been listed.

Table 3: Candidates' strength in Kowloon East (seats=5, threshold=20%)

	Vote shares	Basic seats	Remainder	Seats from remaining votes	Total no. of seats
Chan Yuen Han	33%	1	16%	0	1
Lam Man Fai	3%	0		1	1
Chan Kam Lam	9%	0	9%	1	1
Fred Li	20%	1	5%	0	1
Wu Chi Wai	4%	0		0	0
Ho Wai To	1%	0		0	0
Alan Leong	18%	0	18%	1	1
Andrew To	3%	0	3%	0	0
Any list from pro-democracy camp	7%	－	－	－	－
Any list from pro-China camp	2%	－	－	－	－

* Survey date: 9 to 11 July, excluding vote shares obtained by Kenneth Ting and Ma Fung Kwok, while Albert Cheng was beyond the scope of the interview at that time, and the allocation of vote shares obtained by "any list from the pro-democracy camp" and "any list from the pro-China camp" may also affect the result.