Back

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

The government has finally unveiled its Fifth Report of the Constitutional Development Task Force. Shortly before its release, Chief Secretary for Administration Rafael Hui briefed the Legislative Council on its contents, and claimed that "the proposed package has the support of the majority of the public", quoting an opinion poll commissioned by the Central Policy Unit (CPU). The report of the poll was appended to the main report of the Task Force.

In the author's view, while the government has released the details of the opinion poll after quoting it in a timely manner, this is a kind of advancement, as well as an important step towards the development of a rational civil society.

Is the concerned poll in line with the scientific principles? Has the poll misled the public? Probably these should not be answered by someone who always conducts opinion surveys, so as to avoid conflicts of interest. However, since Hong Kong does not have a completely independent supervising mechanism or body in this industry, seemingly this responsibility has still fallen on individual researchers.

As a note of caution, the author first has to declare that, the Public Opinion Programme of the University of Hong Kong, due to unsuitable criteria, has never submitted tenders to the surveys commissioned by the CPU, and therefore, there should not be any conflicts of interest. POP does not dismiss the possibility of co-operating with other bodies to conduct constitutional development poll in the future. Yet, the most important criteria is that we are allowed to design the research instrument independently, and we bear the full responsibility of the survey results. This is our service policy all along.

Regarding the CPU's opinion poll quoted by the Chief Secretary for Administration, after reading the whole survey report, the author viewed that the whole operation of the survey, together with the presentation and analyses of figures, are basically in line with the professional standard of the opinion polls. However, every survey has its limitations. The author would like to name some here, so as to enhance people's understanding towards the constitutional development poll.

The CPU's constitutional development poll has used 13 questions to test the respondents' support level towards specific concepts or proposals. Due to the limitation of space, the author will only discuss the first 4 questions in the questionnaire as examples:

[v5] At present, the Chief Executive is elected by an Election Committee composed of 800 members. For the next Chief Executive election in 2007, do you agree that the number of Election Committee members be increased?

[v6] (For those who agree that the number of Election Committee members be increased) There are views in the community that the number of Election Committee members be increased to 1600. Do you agree to this proposal? (For those who "disagree" that the number be increased to 1600, please prompt: "In your view, how many Election Committee members should it be increased to?")

[v7] If the number of Election Committee members is to be increased, do you agree that more District Council members be included?

[v8] (For those who agree that more District Council members be included) At present, in the District Councils, there are over four hundred and twenty members returned by election and around one hundred by appointment. Do you agree that all District Council members be included in the Election Committee?

Apparently, the above questions are aiming at testing the respondents' agreement level towards certain concepts and proposals, such as increasing the number of Election Committee members, expanding the Election Committee membership to 1,600, including District Council members into the Election Committee, and including the appointed District Council members into the Election Committee. Survey shows that, all the above suggestions were able to secure a support rate of around 60%. While Rafael Hui claimed that "the proposed package has the support of the majority of the public", this is technically correct. However, the limitations of the poll lie in its inability to test the critical points of pubic support over specific proposals, and its lack of data for comparing alternative proposals.

For instance, if an open-ended question were added before V6: "(For those who agree that the number of Election Committee members be increased) How many members should the Election Committee be increased to?" The figure of 5.7% in "universal suffrage" should not be found in the results then (See the figures of Table 2, read together with Table 1 of the survey report).

Take for another example, if a question, similar to V6, were asked after V8: "There are views in the community that the Election Committee should not include appointed members. Do you agree to this proposal?" The analyses derived will be more comprehensive.

The author is not saying that the government is making up figures and findings, but is just pointing out the limitations of the opinion poll. We can say that, the report proposed by the government, before having heated public discussions in the community, seemingly is accepted by the public. Yet, other possible options, such as a proposal with faster democratic pace, may even be more well-received. How to draw the lines probably is simply not a scientific question.

If politics and science can really be separated, the author then needs to point out that the government committed a mistake two days ago, either intentionally or unintentionally. Two days ago, when Chief Secretary for Administration Rafael Hui quoted the government poll figures to show that all proposals have obtained the support of the majority of the public, he also said, "there are all along different views within the community on setting a timetable for universal suffrage", and "it would be quite difficult to reach a consensus in the near future." In this way, the government seems to have used different reference frames in different issues. If the opinion poll can be used, it comes first. If the written submission can be used, then it comes first. If both are inapplicable, then minutes from meetings will be used to support oneself. This is a kind of unscientific method which should be avoided.

At last, the author needs to point out that, in Paragraph 1.3 of the opinion survey report, it is stated that "the Task Force commissioned, through the Central Policy Unit, the Poly U Technology & Consultancy Company Limited of the Hong Kong Polytechnic University (PolyU) to conduct the poll. The Centre for Social Policy Studies (CSPS) of the Department of Applied Social Science of PolyU was tasked to execute the poll, including sample selection, fieldwork monitoring, data quality assurance and data analysis." There is also a footnote explaining that "PolyU Technology & Consultancy Company Limited is founded by PolyU to coordinate consultancy and research projects carried out by the research units of various departments, including the CSPS, which are commissioned by outsides parties."

Putting aside the complex political and business relations between these bodies and units, the author only wishes to seek clarifications from the government and the relevant bodies: who is responsible for the questionnaire design and deliberation of the concerned survey? This question is asked not because there are problems with the questionnaire, instead, the author hopes to let the public understand that this is an important part in the scientific research, as well as one of the professional codes of conduct in the opinion poll.