Back

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

People in Hong Kong seem to have become used to expressing their opinions through rallies recently. Apart from the theme of a rally, the media and society have become particularly concerned of its scale and the number of participants. Past evidences show that the number of participants estimated by the rally organizers is usually very different from that estimated by the police. The organizers tend to overestimate the number while the police tend to be too prudent in their estimation. People tend to believe what they think rather than scientific facts. This has led to unscientific arguments which bring no avail. Therefore, with the help of the Public Opinion Programme, the University of Hong Kong Research Team tried to estimate the number of participants in rallies in a scientific way in order to provide Hong Kong society with scientific estimates from a neutral corner.

The research team has estimated the number of participants of the two rallies held on 1st January, 2005. They were held by The Defend-Hong Kong's Stability Alliance and the Hong Kong Confederation Trade Unions respectively. We will now explain the methodology used.

Since it is impractical to count every participant in a huge demonstration, we use sampling. In the demonstration organized by the Defend-Hong Kong's Stability Alliance, we set up a counter on a crossover in Queen's Road Central near Ice House Street. We counted the total number of participants passing the crossover in one-minute intervals. That means we started counting heads for one minute when the march arrived the crossover. Then, we stopped counting for the following minute. The process of counting for one minute and then skipping the subsequent minute was repeated until the march ended.

Number of participants in Defend-Hong Kong's Stability Alliance

Time at which counting was made	Lane 1	Lane 2	Total
11:44-11:59	2,082	1,660	3,742
12:00-12:15	1,602	1,200	2,802
12:16-12:32	1,393	1,022	2,415
Number of participants in the whole journey	5,077	3,882	8,959*

* People joining and leaving in the middle of the route were not included here

Participants of the rally mainly occupied two lanes. We arranged altogether four members with two responsible for each lane. On each lane, one member counted by facing the people flow while the other counted by facing the back of the flow. We chose one of the two results in every count. Then we took the average to calculate the total number of participants. We also took down the number in writing and videotaped the whole process so that we could verify our estimation later. There were nearly 9,000 participants in this rally. According to an online survey done by the Public Opinion Programme, about 3% to 8% of the participants joined or left the march along the way. Since those people were not included in our counting process, the number of participants should range from 9,200 to 9,700.

On the other hand, participants in the rally organized by the Hong Kong Confederation Trade Unions that afternoon occupied one lane only and their speed was slow, so we counted every participant. We set up a counter at a crossover at the intersection of Hennessy Road and Arsenal Street in Wan Chai. We arranged five members altogether, three facing the march while two facing the back of the march. There was also one member responsible for videotaping the whole process. We obtained five results: 329, 329, 324, 325 and 308. The average was 323. Moreover, since no participant joined or left the march along the way, the number of participants was estimated to be around 320 and 330.

The research team gained precious experience through estimation in the two demonstrations. During the counting process, we learned how to cope with critical situations and respond to changes. We need to check the route of the march beforehand and choose an appropriate location for our counters. We discovered that the crossover was a good location for counters. Videotaping was also vital because this made verification much more convenient. In addition, we would jot down special events such as pausing and interceptions of the march. However, as there were many changes in the arrangements of each demonstration, such as arrangements of roadblocks, traffic lights, the final arrangement of the march was not finalized until the last moment. Therefore, very often, our division of labor could only be confirmed after the rally had started.

The University of Hong Kong Research Team will continue to estimate the number of participants of rallies in future in order to provide Hong Kong citizens with an objective estimation. We hope to aid people to grasp a clearer picture of the instant situation so that everyone can make a better choice. Meanwhile, we want to prevent the estimated number of participants from being distorted deliberately or even being used as means by the press. Seeking truth is one of the foundations of Hong Kong's success and stretching a point will only harm us.