Back

Robert Ting-Yiu Chung (Director of Public Opinion Programme, the University of Hong Kong)	Joseph Man Chan (Professor at the School of Journalism and Communication, the Chinese University of Hong Kong)
Translated by Calvin Chun-Kit Chan (Research Executive, Public Opinion Programme, the University of Hong Kong)
Note: This article represents the view of the authors and not their respective universities.
Half a million of people took to the streets on the 6th anniversary of the handover, of whom nearly 60% came from middle class families, 40% were professionals or semi-professionals, we have discussed these before. According to the findings obtained in the on-site survey conducted in the Victoria Park and along the marching route, less than one in twenty respondents went to the rally with the social organizations that they belonged to, while only 34% regarded the mobilizing power of social organizations as influential. Therefore, July 1 Demonstration was self-mobilized rather than association-mobilized.
On July 1, apart from conducting a questionnaire survey enroute, another online survey was also conducted as a pilot study, with the aims of collecting more public opinion and studying the power of the Internet in mobilizing the mass. This article focuses on the findings of the online survey, and the political efficacy of the computer network.
Online survey received warm public responses
On the day of the demonstration, our research team of 20 odd people conducted questionnaire interviews at different times and locations in Victoria Park and along the route. Besides, almost 20,000 leaflets were distributed to the demonstrators, who were invited to visit a specified website to complete an online survey. However, people who were already interviewed received no such invitation.
In order to distinguish the opinions of those with invitations from other volunteered respondents, all leaflets have their unique codes for identification purpose. The leaflets did not invite other demonstrators to join the survey, but we did not forbid it either. At the end of the study, 17,432 submissions were received in three days, 26,330 in 10 days, while a total of 26,658 completed questionnaires were collected in two weeks. The general public's eagerness to express their dissatisfaction with the government was clearly borne out by their passionate responses.
Using the daily number of returns as an indicator of the public mood (see Chart), most demonstrators filled in the online survey between July 3 and 4, with 6,705 and 5,841 submissions received respectively. The number of submissions then dropped geometrically. However, another peak appeared on July 7, obviously due to the fact that the Chief Executive has, after losing the support from Liberal Party Chairman James Tien, announced the deferral of the legislation of Article 23 in the early hours of that day. We have indeed taken into consideration whether other factors, like the effect of post-holiday working day, as well as the release of our "July 1 Demonstration On-site Survey" findings on July 6, had pushed up people's eagerness to respond. After examining the trend of response rate between July 1 and 3, we found that although July 2 was another post-holiday working day, the response rate did not slow down accordingly. Besides, the release of survey findings might as well have cooled down some people's eagerness to respond. Our conclusion was that more demonstrators expressed their views online that day mainly because they were encouraged by the deferral of the legislation, which was a key demand of the demonstrators. As an extension of this conclusion, any positive outcome of a mass protest would not only alleviate people's grievances and make society more stable, it would also encourage more people to express themselves, thereby facilitating the development of a civic society.

Chart

Several shortcomings are commonly found in online surveys: (1) their sampling frames include only computer users; and (2) lack of effective control on their sampling and data-collection processes, so that they would not be manipulated. We tried to overcome these problems by adopting the following measures:

First of all, we first analyzed the data obtained from the on-site survey conducted in Victoria Park and along the marching route on July 1, so as to identify the demographic structure of the demonstrators and set a benchmark for the online survey. Findings of the on-site survey suggested that 14% of the respondents never went online. In other words, the web-based survey would not be able to represent these people's opinions.

Second, information encoded in the almost 27,000 on-line submissions, such as their source, dates, IP addresses, leaflet codes and key contents, were examined again and again, in order to classify them into different groups according to their trustworthiness. The identification procedures are listed as follows:

Step 1: Submissions after the midnight of July 11 were not processed, 26,122 submissions were kept as preliminary effective cases.

Step 2: Another 939 cases without personal particulars were dropped, and the remaining 25,183 cases entered the next stage.

Step 3: For the 110 cases with leaflet codes repeated for more than twice, only the last submissions were kept.

Step 4: The 12,190 cases with repetitive IP addresses were examined in terms of their timestamp, nature of their outgoing website, and content of their submissions, in order to identify and eliminate skeptical and duplicate cases.

After repeated examinations, the submissions were classified into different groups and then compared with the findings of the on-site surveys, at various levels. Because of the complexity of the analysis and the sheer number of classifications, only two relatively more important types of submissions are discussed here. The first type consisted of the 2,431 online samples with correct leaflet codes, received in the first 10 days. They were considered as credible invited samples. The second type consisted of 22,599 submissions also received in the first 10 days, but without leaflet codes. They were considered credible uninvited samples. According to our records, exactly 19,350 invitation leaflets were distributed. According to our on-site survey findings, 86% of these invitations should have been given to demonstrators who would go online, the response rate for the first type of respondents was therefore 2,431 / 16,641, which yielded 15%. Whether the sample was representative of all demonstrators who would go online has yet to be investigated.

Nonetheless, the power of Internet mobilization could very much be illustrated by the fact that 90% of our online sample were uninvited samples. The leaflets distributed during the rally did not ask the demonstrators to invite other people to participate in the online survey, yet, they volunteered to promote the survey, resulting in a 10-time increase on the number of submissions, which reflected people's desire to express their views. This kind of "people's power", if utilized in a proper way, would be rich resource for our society.

Foundation of July 1 Demonstration laid by internet mobilization

Analyzing the demographic profile of the various samples, it was found that the proportion of professionals and semi-professionals in the uninvited sample was 46%, which was 6 percentage points higher than that of the onsite and invited samples. Respondents who had tertiary education or above were as high as 79% in the uninvited sample, which was 23 and 15 percentage points higher than the onsite and invited samples respectively. Regarding their age, the proportion of those who aged between 20 and 39 was 79%, which was 17 and 14 percentage points above the other two samples respectively. Gender-wise, females occupied 47% of the uninvited sample, which was 7 and 9 percentage points higher than that of the other two samples. Hence, participants of our online survey who were mobilized by some kind of networks were mainly those aged between 20 and 30, with higher education attainment. According to our understanding, personal email communication containing our survey website was the major means of mobilization. This mode of self-mobilization should be equally applicable to political mobilizations.

Regarding the demonstrators' demands, the general inclination of the online sample were more or less the same as that of the on-site sample, but the strength of such views has been somewhat weakened, probably because they have been eroded by time and many subsequent events. As shown in Table 1, CE Tung Chee-hwa's rating, on a daily basis, was only 9.9 marks on July 1, it slowly climbed to 14.9 marks on July 5, and then, after the announcement that he would press on with the legislation, it dropped to 13.4 marks. It was until he announced to defer the legislation that his rating rebounded, to 16.2 marks on July 10. Tung's overall rating in the online sample was 12.7 marks, President Hu Jintao scored 58.8 marks, while Premier Wen Jiabao scored 63.2 marks. On the other hand, ratings of the three leaders registered in the on-site survey were 13.8, 57.7 and 62.0 marks correspondingly. Both surveys proved that the demonstrators did not target at the Chinese leaders. Any proposition that people took to the streets to rock the rule of the Central Government was not supported by evidence.

Table 1：Daily ratings of CE Tung Chee-hwa (online full sample)

	Number of raters	Support rating*	Standard error
July 1	484	9.9	0.68
July 2	4,225	10.7	0.24
July 3	6,430	12.2	0.20
July 4	5,511	12.6	0.22
July 5	2,733	14.9	0.38
July 6	1,565	13.4	0.46
July 7	2,064	13.3	0.37
July 8	1,016	14.8	0.55
July 9	564	15.1	0.69
July10	320	16.2	0.93

* 0 means absolutely no support, 100 means absolute support, 50 means half-half

With respect to the online activities participated by the respondents, as seen from Table 2, the online samples were clearly more active than the onsite sample. The proportion of respondents who exchanged public affairs information, news about the demonstration, opinions on public affairs, or even messages satirizing against the CE and the SAR Government, were over 80% in the uninvited sample, around 20% higher than that of the onsite sample, and about 10% higher than that of the invited sample. Such difference were, by and large, also found in the two items "getting in touch with family and friends" and "general online discussion with friends". The findings suggested that online discussions and the exchange of information had opened up a new mine of public opinion, it also helped to lay the foundation for the self-mobilized July 1 demonstration.

Table 2: Online behaviour of the respondents

	On-site sample	Online invited sample	Online uninvited sample
Sample base	587	2,354	21,745
Online experience	86%	97%	98%
Chatroom discussion*	45%	45%	46%
General discussion*	57%	68%	74%
Getting in touch with family and friends*	69%	85%	91%
Getting in touch with unknown people*	46%	49%	51%
Exchanging public affairs information	68%	76%	87%
Exchanging opinions on public affairs*	67%	73%	84%
Exchanging messages satirizing against the CE or the SAR Government*	64%	72%	81%
Exchanging news about the demonstration*	64%	71%	84%

* Respondents who answered "very often", "quite often", "sometimes" and "not quite often" in the long version of the on-site questionnaire survey. Because these questions
were not included in the short version of the on-site questionnaire, the sample base was only 587.

Traditional gossips replaced by online discussions

In traditional agricultural societies, gossiping in the market place was the primary channel for people to share information and exchange views. At the advent of media in the modern era, such channel of receiving information has changed. Nevertheless, most people still need to seek reference from other people's opinions, including opinion leaders and members in their social network, before actually taking their own stands on social events. The very popular "spiral of silence" theory in opinion researches laid much emphasis on the effect of other opinions and the perception of the public mood in the formation of mainstream opinion.

With the arrival of the IT era, the Internet has gradually become the most convenient, speedy, multi-directional and highly personalized information channel. The old-fashioned market gossips have now mutated into cyber chats spreading around the Internet. Right before July 1, we kept receiving "sarcastically hilarious" emails which satirized the CE and his wife. These emails were full of innovative ideas, and rich in text and pictures. They demonstrated the creativity and wisdoms of the public. For those who are silently gnashing their teeth with burning hatred, an email joke which brings a smile to their faces is far more important than those a multi-million dollar news advertisements.

Street gossips of the traditional societies could only spread slowly through the interpersonal network of gregarious communities. Even when phones are used in modern days, the transmission mode is still one-to-one, and the speed in passing out the words is still rather limited. However, with the emergence of emails, the mode of transmission has become one-to-many, which allows multi-directional sharing of the message across time and geographical constraints. Provided that the information flow remained unimpeded, with the emails of inferior quality being kept under control, more useful functions of the computer network are yet to be seen. Many professionals are too busy to answer the phone, yet, personalized emails can be stored in their mailbox and read when they are free. This is one popular feature of emails, which helps to mobilize these professionals.

The relation between July 1 Demonstration and Internet mobilization could be further explained by the figures in Table 3. As seen from the figures, when the demonstrators were asked to evaluate the importance of the calls from various sources towards their decisions to participate in the march, the calls from friends via interpersonal network were considered to be of crucial importance among all samples. As for the onsite and online invited samples, the relative importance of Internet mobilization ranked 7th and 8th respectively, which was fairly similar. But for the online uninvited sample, because the importance of many other factors have dropped, Internet mobilization has moved up to the 3rd rank, meaning that these uninvited respondents were relatively more responsive to the calls from the Internet.

Table 3: Differential effects of various mobilization means (respondents who answered "important" / "very important")

	On-site sample	Online invited sample	Online uninvited sample
Sample base	1,154	2,431	22,599
Calls from interpersonal network such as friends, colleagues or fellow students	68%	62%	67%
Calls resulting from news reports on newspapers	66%	54%	53%
Calls from the Internet (including emails among friends)	54%	45%	52%
Calls from radio phone-in programmes	65%	57%	52%
Calls resulting from TV news	61%	52%	52%
Calls from newspaper editorials	64%	49%	48%
Calls from family members	51%	48%	45%
Calls from newspaper columns	60%	46%	43%
Calls from religious leaders and followers	44%	36%	37%
Calls from an organization I belong to	34%	29%	31%
Calls from political parties	44%	35%	29%

Combined with other figures of the survey, respondents of the online survey, especially young professionals with higher education attainment in the uninvited sample, it became obvious that this group of people were more active or responsive when it came to Internet mobilization. To them, the Internet functions more or less as the print and electronic media. As for media usage, 65% of the onsite sample read the newspapers on a daily basis, 62% of the online invited sample did the same, but only 59% of the online uninvited sample did so. With respect to online experience, merely 31% of the onsite sample spent more than 2 hours a day online, while the corresponding figures for online invited and uninvited samples were 41% and 47% respectively.

It is, however, noteworthy that only 25% of the onsite sample belonged to some social associations. However, it was 29% for online invited sample and 36% for the online uninvited sample. This shows that the young professionals who tended to initiate or respond to internet mobilization were actually those who have more contacts with social organizations.

This brings us to the last issue of this article: exactly which mode of mobilization fits better to the development of our local democracy? Exchanging messages by street gossips has already become history. In more traditional Chinese communities, political mobilization is normally intertwined with returning favours, face-saving and making social connections. Horizontal mobilization more often than not involves intricate and secretive exchanges of vested interests, while top-down mobilization often stresses coercion and discipline. The communist mobilization system, on the other hand, stresses greatly on ideological correctness and the exercise of social pressure. These only do harm to the development of rational democracy, they are not suitable for knowledge-based civic societies.

An integration of personal and computer networks

July 1 Demonstration was not organized by some powerful associations, instead, it was a horizontal communicated and self-mobilized social movement. Most participants were highly educated, and they had remained rational, peaceful and patient. It was a self-improving movement which Hong Kong people should feel proud of. According to our survey findings, the mobilizing power of the Internet was different for different people, but its influence on those professionals aged between 20 and 30, who were highly educated and often went online, was very significant.

The characteristics of internet communications include its multi-directional feature and non-restricted across time and space. It allows instant responses from the recipients, but they can also choose to react in another time or method most convenient to them. What's more, the recipient's identity can be kept with strictest confidentiality, or according to different levels of confidentiality. Whether mobilizing others or being mobilized, there is a bigger room for thought and responses. Certainly, there are irrational and irresponsible behaviours existing in the Internet, yet, from the perspective of political mobilization, eliminating face-to-face social pressure by internet mobilization is certainly more healthy and rational.

The popularity of internet usage would certainly become more extensive and osmotic, and messages broadcast by the media could be re-transmitted more widely through personalized interpersonal networks. The power exerted by one single person could be expanded far beyond one's own social network, if used in an appropriate way. In view of this, an integration of interpersonal and computer networks would become a brand new force for mobilization, spreading outward from the young and more educated professionals. This significance of this new force cannot be underestimated.