Back

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

This article concentrates on our exit poll results of Kowloon East. Our data was collected on the election day outside 16 polling stations in the geographical constituency. It comprised 1,393 successful cases and 40 partially completed cases, with a response rate of 70.6%. We will not go into the research design here, all information of the survey will be published in our HKU POP Site by the end of next week.

Kowloon East attracted huge public attention because new-comer Alan Leong parachuted into this constituency and Albert Cheng suddenly ran in the election in a high-profile way, which complicated the internal coordination of the democratic camp. How to fill in the vacancy of Szeto Wah and proceed with obtaining one or two more seats irritated the camp. On the other side, Chan Yuen-Han was the "Queen of Votes" in the last Legislative Council election. This year she parted company with Chan Kam-Lam and ran in the election with Lam Man-Fai. Apparently this was the strategy of obtaining a second safe seat with the hope of getting a third seat. There were 5 seats in Kowloon East. The pro-Beijing camp hoped to obtain 2 - 3 seats while the democratic camp hoped to obtain 3 - 4 seats. It was clearly a competition for the last seat. At the later stage of the election, the Democratic Party's list apparently was unable to fight for 2 seats after the Alex Ho's Incident. Besides, the support rating of Alan Leong and Albert Cheng's list had become relatively stable. Therefore, the focus of the election fell onto the vote allocation between Chan Yuen-Han and Chan Kam-Lam.

According to the analysis of the actual vote shares of the 2 camps, if the pro-China camp could not allocate votes appropriately between Chan Yuen-Han's and Chan Kam-Lam's lists, whereas the democratic camp could make good use of the remaining votes obtained by the 3 lists, then the democratic camp could get 4 seats under some combinations. However, at the final stage, since the pro-China camp shifted its target to rescue Chan Kam-Lam and gave up helping Lam Man-Fai to enter the Council at the risk of losing 2 seats, the source of votes between Chan Yuen-Han and Chan Kam-Lam had experienced great changes. One of the key point of the exit poll analysis is to examine the flow of votes of the concerned lists.

Let us first analyze the flow of votes among camps between the 2 elections. Table 1 selects the lists which have won in the last election and compares them with the voters' choices in this election. Figures reveal that, only 5 % of voters in support of the Democratic Party last time shifted to the pro-China camp. Yet, 27% of votes obtained by the pro-China camp last time flowed to the democratic camp. This explains why the proportion of the democratic camp's to the pro-China camp's strength had been altered from 5:5 last time to 6:4 this time. Regarding the vote shares obtained by the pro-China camp last time, 15% were shifted to Albert Cheng's list, probably because Albert Cheng had absorbed a certain amount of vote shares from the grassroot level.

As regards the intra-party flow of votes, what is more interesting is that, Chan Yuen-Han, who topped the candidate list last time, whose vote shares reserved for herself (34%) was even less than Chan Kam-Lam who came second last time (40%), accounting for a large difference of 6 percentage points. This was apparently the result of tactical voting within the camp. On the side of the democratic camp, the Democratic Party could only keep 35% for Li Wah-Ming's list, while the remaining 35% and 24% had flowed to Albert Cheng's list and Alan Leong respectively. For the voters who had not voted in this constituency last time, or had registered for less than 4 years, their proportion in support of the democratic camp was higher than the general voters. Yet it did not affect much.

In consideration of the voters' background, Table 2 reveals that, except for Chan Yuen-Han's list, the proportion of male to female in each candidate list was around 6:4. Together with the proportion of full-time housewife, no rumoured "lady-killer" candidate could be found. Looking at the other demographic factors, the source of votes of Chan Kam-Lam's list was unique. Among his supporters, 56% were aged 50 or above and with relatively low education standard. They were quite different from Chan Yuen-Han's within the same camp, and were believed as the targets of the pro-China camp in the final stage of mobilization. Figures also reveal that, Alan Leong's supporters clearly possessed higher education level, with a higher proportion of executives and professionals. Apparently they were the middle-class people. In the democratic camp, Albert Cheng's list was closer to the worker class and similar to the 2 pro-China camp's lists.

Although there was a clear line of demarcation between camps, the party sense of supporters of each list varied. Table 3 reveals that, the party sense of the Democratic Party's supporters was the strongest, as 45% of them said political party factor was their main consideration. While they still voted for the Democratic Party under the influence of Alex Ho's Incident, this group of people was clearly supporters with stronger party sense. For supporters of other lists from the democratic camp, their main considerations were the candidate factor. On the other side, supporters of Chan Kam-Lam's list also had a strong party sense, at 39%, which was totally different from the 17% of Chan Yuen-Han's list. This demonstrated clearly the tactical voting strategy of the pro-China camp. However, in another question, very few Chan Kam-Lam's supporters said they based their choice on vote allocation. This probably reveals that they themselves were not too clear about the tactical voting strategy, but just followed the appeals and instructions of the political party. Alan Leong seemed to have demonstrated his personal charm best, followed by Chan Yuen-Han and Albert Cheng. The style and source of votes of Chan Yuen-Han and Chan Kam-Lam were completely different, which was an important development of the political ecology in the constituency.

When one wants to analyze the differences between Chan Yuen-Han and Chan Kam-Lam, we can also look at their supporters' political orientations. Table 4 reveals that, 43% of Chan Kam-Lam's supporters were satisfied with the performance of the HKSAR Government whereas the respective percentage among Chan Yuen-Han's supporters was only 21%. Around 60% - 70% of voters in support of the democratic camp expressed dissatisfaction with the HKSAR Government, while the 70% of Albert Cheng's list, together with the New Territories East's Leung Kwok-Hung, came first among all lists in Hong Kong.

If one uses the participation in the July 1 Demonstration as the political orientation's indicator, supporters of Chan Kam-Lam's list apparently lay in an extreme. Its 3% of participation rate was the lowest among all pro-China camp's lists, which was much lower than the 10% of Chan Yuen-Han's list. When one further looks at the political orientation of supporters towards the universal suffrage, only 11% and less than 30% of supporters of Chan Kam-Lam's list supported direct elections in 2007/08 and before or in 2012 respectively. The respective percentages of supporters of Chan Yuen-Han's list were over 20% and around 50%, which were closer to the democratic camp's 45% - 60% and 65% - 75%.

Using the above figures to form the political spectrum, Chan Kam-Lam's list was clearly on one extreme, whereas Albert Cheng's list, the Democratic Party's list and Alan Leong clustered together on the other extreme. Compared with Chan Kam Lam's list, Chan Yuen-Han's was closer to the middle of the spectrum. The pro-China camp could allocate votes in 2 lists with similar shares in the light of the situation in Kowloon East, as this was clearly the result of mobilization within the camp and propaganda outside the camp. Supporters with stronger party sense and emphasizing on political orientation voted for Chan Kam-Lam while Chan Yuen-Han, with her personal charisma, absorbed those long-term supporters with weaker party sense and general voters emphasizing on economic and livelihood issues. The pro-China camp gave up the risky strategy of fighting for a third seat, so as to save all efforts to rescue Chan Kam-Lam. The tactical voting strategy of the pro-China camp should be considered as very successful.

After the pro-China camp had altered its strategy, it was in fact not difficult for the democratic camp to obtain 3 seats. However, looking at the supporters of the 3 lists from the democratic camp, except for Alan Leong who tended to absorb the middle-aged middle-class people more easily, their source of vote seemed to be seriously overlapping while the allocation of votes by district was not apparent. This resulted in contradictions within the party and was unfavorable to the overall deployment, thus minimizing the chance of winning.

Followed by the increasing number of directly-elected seats, unless there are drastic changes in the electoral system or boundary of constituencies, the election engineering and tactical voting strategy will become more and more important. This article series has already examined the voters' behaviour on Hong Kong Island and Kowloon East, both revealing the importance of tactical voting. How to grab the characteristics and wishes of different voters? When do we need to let the voters choose freely? When is a high-profile tactical voting needed? These will be the key to the success and failure in future direct elections.

Table 1: Flow of votes in 2 elections

	Y.H. Chan's list	Szeto Wah's list	Have not voted last time
W.M. Li's list	4%	35%	15%
Albert Cheng's list	15%	35%	28%
Alan Leong	8%	24%	25%
K.L. Chan's list	40%	2%	11%
Y.H. Chan's list	34%	3%	21%

Table 2: Analysis of voters' background - Proportion of:

	W.M. Li's list	Albert Cheng's list	Alan Leong	Y.H. Chan's list	K.L. Chan's list
Male	58%	60%	56%	52%	56%
Female	42%	40%	44%	48%	44%
18-29 years old	24%	28%	30%	25%	10%
30-49 years old	46%	55%	60%	52%	34%
50 years old or above	30%	17%	10%	23%	56%
Voters possessing primary education level or below	17%	4%	10%	15%	38%
Voters possessing tertiary education level or above	31%	27%	40%	28%	16%
Executives and professionals	24%	26%	41%	26%	16%
Clerical and service workers	27%	30%	24%	24%	8%
Manual workers	13%	15%	11%	15%	19%
Full-time housewives	10%	7%	6%	11%	14%
First-time voters	14%	20%	27%	20%	18%

Table 3: Main considerations for choosing candidates

	W.M. Li's list	Albert Cheng's list	Alan Leong	Y.H. Chan's list	K.L. Chan's list
Candidate factor	36%	69%	83%	75%	44%
Political party factor	45%	7%	4%	17%	39%
Combination of lists	5%	11%	2%	4%	3%
Vote allocation concern	10%	9%	5%	0%	2%
Hard to say	4%	4%	7%	4%	13%

Table 4: Voters' political orientation - Proportion of voters who:

	W.M. Li's list	Albert Cheng's list	Alan Leong	Y.H. Chan's list	K.L. Chan's list
Were satisfied with HKSARG	4%	6%	3%	21%	43%
Were "half" satisfied with HKSARG	32%	24%	39%	59%	41%
Were dissatisfied with HKSARG	59%	70%	57%	18%	13%
Took part in July 1 Demonstration	37%	36%	38%	10%	3%
Supported direct election of CE by 2012	65%	67%	74%	51%	29%
Supported direct election of all LC members by 2012	66%	65%	76%	45%	26%
Chose basing on candidates' political attitude	44%	39%	49%	8%	18%
Chose basing on candidates' economic and livelihood policies	41%	49%	39%	89%	55%