2010 Legislative Council By-election Survey Third Summary ReportBack
|Press Release on May 14, 2010|
To fulfill its social responsibility, the Public Opinion Programme (POP) at the University of Hong Kong has completed its 2010 Legislative Council by-election survey on May 12. Because of resources constraint, the survey is conducted like an irregular quasi rolling poll with small samples size per day but covering a long fieldwork period. When sufficient data is accumulated, the findings will be released in the form of summary reports. All findings have now been uploaded onto the "HKU POP SITE" (http://hkupop.hku.hk) for public consumption. On April 26, POP released the first summary report which covered the survey period of April 8 to 25. Then on May 6, POP released the second summary report which covered the survey period of April 26 to May 4. The third summary report released today covers the survey period of May 5 to 12.
Like other telephone surveys conducted by POP, these LC by-election surveys are conducted by real interviewers under close supervision. Target population of these surveys are general population of all the geographical constituencies aged 18 or above. Telephone numbers are first selected randomly from telephone directories as seed numbers. Another set of numbers is then generated by using the plus/minus one/two method. Duplicate numbers are then screened out, and the remaining numbers mixed in a random order to become the final sample. Upon successful contact being made with a target household, one member of the household is selected among those present using the "next birthday" rule.
As a general practice, all figures have been weighted according to provisional figures obtained from the Census and Statistics Department regarding the gender-age distribution of the Hong Kong population in 2009 year-end. All the result findings are based on the weighted figures.
POP interviewed a total of 1,523 Hong Kong citizens including 1,243 registered voters about their views on 2010 Legislative Council by-election in six separate days between May 5-7 and May 10-12 by means of random telephone interviews conducted by real interviewers. The sampling error of figures using the overall sample is below +/-3% at 95% confidence level, and the response rate of the survey is 72%. Contact information of the survey is as follows:
Propensity to vote
The survey of the third stage reveals that 43% of registered voters throughout Hong Kong would likely vote, which is similar to that in both first and second stages. However, the percentage in Kowloon West constituency has dropped, while that in Kowloon East constituency has increased (please see "candidate strength in each constituency" table for details). It should be noted that, 43% likely voters would not give a 43% turnout rate, because many people who claim they would vote at this stage would eventually not vote.
Candidate strength in each constituency
In this analysis, "firm votes" is defined as likely voters who have already decided how to vote, while "stray votes" is defined as likely voters who have not decided how they would vote, or those who would not support any candidates and might cast blank votes. Statistics for the five constituencies are as follows, using "firm votes" as basis of all "raw" support rates, and all registered voters as the sample base:
 Chan Yuet-tung and To Sum were candidates of HK Island and Kowloon East constituencies respectively in the first place. Since they have been verified as invalid by Election Affairs Office, they are discarded from this table.
 Such changes have gone beyond the sampling errors at the 95% confidence level, meaning that they are statistically significant prima facie. However, whether numerical differences are statistically significant or not is not the same as whether they are practically useful or meaningful.
 These analyses are based on "firm votes" using all registered voters as sample base, while "stray votes" have been excluded. It is slightly different from those presented in the first summary report. If necessary, readers can re-base their analyses after excluding "unlikely voters", or even "stray voters". The percentages will be different, but candidate strength will remain unchanged.
Using likely voters as the base, meaning that unlikely voters are excluded, the "effective support rates" of the candidates in each constituency are as follows:
Note: The following commentary was written by Director of POP Robert Chung.
Fulfilling its social responsibility, POP has completed its pre-election survey for the forthcoming Legislative Council by-election. The survey takes the form of an irregular quasi rolling poll divided into three stages. Our first stage covers the first half of the campaign period. Our second stage covers the mid-late stage, while our third stage covers the final stage. The sample size for each stage is roughly the same, with over 1,500 valid samples each, comprising over 1,200 registered voters.
Our survey shows that voters' propensity to vote has been very stable over the entire campaign period, at not more than 45%. Depreciate that by half or one-third, we have the final turnout rate in between 22% and 29%, which is much lower than that of 2008. Broken down by constituencies, the fluctuation is relatively bigger, probably due to smaller sample size, but they remain close to the overall figure. In our latest survey, Hong Kong Island and Kowloon East voters show the highest propensity to vote, but differences across constituencies remain within sampling errors.
Regarding candidate strengths, we split the sample into "firm votes", "stray votes" and "unlikely voters" within each constituency according to register voters' propensity to vote and choice of candidates. "Firm votes" is defined as likely voters who have already decided how to vote; "stray votes" is defined as likely voters who have not decided how to vote, plus those who would not support any candidate and might cast blank votes. In our latest survey, the percentages of "unlikely voters" across different constituencies stand high at 52% to 59%, quite similar to those registered in the early stage of the campaign.
From our analysis, although our sample size is small and sampling errors are high, because voting propensity is low, candidate strengths are quite lopsided across all five constituencies. The situation in Kowloon West was getting tight last time, but it has sunk back to normal. Nevertheless, the difference between Wong Yuk-man and Peck Wan-kam remains closest across all constituencies.
Looking at all three stages of survey together, one could say that there is not much change during the campaign period, and voters' propensity to vote remains low. When voters finally decide to vote, or not to vote, exactly how they would interpret their political action, and how that would change the political culture of Hong Kong, should lay the direction of our future research.