HKU POP SITE releases headcount figures of the June 4 vigilBack

The Public Opinion Programme (POP) at the University of Hong Kong started to study mass gatherings and rallies in mid-2003. Over the years, POP team members, together with different experts and scholars, academic institutes, student organization and volunteers, have conducted many onsite observation, headcounts and opinion surveys in different gatherings and rallies. Meanwhile, POP also opened a feature page on 「studies of mass rallies and gatherings」 in the POP Site to record related activities and commentaries. On June 4, 2009, which marks the 20th anniversary of the June 4 Incident, POP again deployed a team of researchers in Victoria Park, the venue of the June 4 vigil, to count the number of participants. POP will also conduct a headcount in this year's July 1 rally, for the 7th consecutive year.

POP started the headcount on June 4 this year starting from 7:30pm, it ended at 10:30pm. The research team comprising 18 members was divided into 2 groups. One group was responsible for sketching and recording the situation at different points in time, while the other group was responsible for counting the number of people within specific areas by manual tally counters.

Specifically, the second group was to measure the total amount of people gathered in each small penalty boxes of the soccer pitches at different times, from which the average density was calculated. With reference to the data collected from the first group, the total amount of people assembled could be obtained by multiplying the area of each location by its respective density.

When the headcount began at 7:30pm, most of the participants were mainly assembled at the two soccer pitches closest to the stage. About half of the third soccer pitch was unoccupied.

By 8:00pm, all 6 soccer pitches were saturated. Participants then began to gather at the Central Lawn until it was also saturated at 8:30pm. After that, people gradually gathered alongside the soccer pitches, the border of the Central Lawn, the Pavilion Area, the basketball courts and the pathways around the soccer pitches.

1. People occupied all 6 soccer pitches, their adjacent pathways and border areas. Excluding the area of the stage, the total area was around 28,300 sq. meters.
2. The total area of the Central Lawn covered by people was around 10,800 sq. meters.
3. The total area covered by people in the Pavilion Area was around 1,110 sq. meters.
4. The total area covered by people in the basketball courts was around 1,050 sq. meters.
5. The total area of the pathway around the soccer pitches was around 4,160 sq. meters, but they were not fully occupied, hence different density figures need to be used.

The sum of areas (1) to (4) is 41,260 sq. meters. The sum of areas (1) to (5) is 45,420 sq. meters.

Because of different densities in different areas, one should have to multiply different areas by their corresponding density in order to obtain accurate figures. However, because there were too many people assembled, and the researchers could hardly move inside the soccer pitches, nor see the sideline clearly. The research team finally had to use the average density obtained from measuring small penalty boxes to represent the average density of all soccer pitches, the Central Lawn, the Pavilion Area and the basketball courts. The figure was 2.67 people per sq. meters, obtained by repeated measurements taken at different time and different points. It was a bit lower than the figure collected in 2004 and 2005, which was 2.82 people per sq. meters, by a more sophisticated method.

According to the above information, POP obtained the following figures:

The total area of all locations except the pathways around the soccer pitches is about 41,260 sq. meters, multiplied by 2.67 people per sq. meter, the total number of people was 110,000. Adding the number of people in those pathways calculated by different density figures, and using +/- 10% as the operational error, the final number should fall in between 108,000 and 132,000.

Note: The following commentary was written by Director of POP Robert Chung.

Whether it is June 4, July 1 or any other mass gatherings or demonstrations, the headcount figures announced by the organizers and government units often show a huge difference. Director of POP Robert Chung have the following impressions after conducting totally 16 headcounts over the past 7 years:

• All organizers tend to exaggerate their headcounts, while government units tend to underestimate the turnout. Using June 4 vigils as an example, the organizer's headcounts are usually 2 to 3 times that of the police figures.
• Since both parties do not reveal their methodologies and detailed figures, there is no way people can monitor the process or check the results. The degree of scientificity and amount of exaggeration can hardly be estimated.
• Interestingly, when measuring the same kind of activities over the years, although the headcount figures are so different from both parties, their trend of change is very consistent. We can therefore infer that the figures announced by organizers must have included some political and psychological factors which blow up their true values, while the government figures must have included some factors which compress the true figures. If there is basically no change in the ways the figures are stretched or compressed, then the direction of change across different years may be true, while the headcount figures themselves are not.
• At certain times for certain activities, the differences between organizers' figures and government figures are reduced, probably because both parties are expecting some headcount figures compiled by third parties. 「Third parties」 here means various scholars and experts who occasionally conducting headcounts. However, because these headcounts are irregular, and not meant to be official audits, they have set very little pressure on the relevant parties.

Before our society takes scientific headcount figures seriously, when reporting these figures, it would be better for our media to add expressions like 「according to organizers' claim」, 「police estimates」, and 「method unknown」 when quoting them.