July 1 Afternoon March Estimates
Of course, this is my specialty (see The Hong Kong 7/1 March: Crowd Size
Estimates for last year). Here is the summary of the situation
- At 6:45pm, according to Ming Pao (via Yahoo
[translation] The organizers (Civil Human Rights Front) said that
45,000 persons participated in the march, while the police counted only
11,000 persons at Victoria Park where the march began.
- At 7:30pm, according to Ming Pao (via Yahoo
[translation] The Civil Human Rights Front had previously estimated
that the afternoon march had 45,000 participants. But later, it issued
a revision and the number of marchers should only be 21,000 persons.
But the police said that 11,000 persons started from Victoria Park, and
there were 17,000 persons over the whole march.
In time, we will undoubtedly hear about how the
45,000 number precipitously fell down to 21,000. The organizers ran their
own count, but they used the methodology of a Hong Kong University team. The actual counting was done by the
Hong Kong Professional Teachers Union, and there was an independent count done by HKU
Perhaps the official designated counters told the CHRF in no uncertain terms that they counted
21,000 and that they will go public with that figure and it is likely that HKU
POP will come up with the same numbers.
The organizers had insisted that there were
530,000 people out there last year. So 21,000 is about 4% of last year's
number. But as horrible as this was, the
organizers really had no choice but to issue the revision. In the meantime, let me
count the damage for you on all the mainstream media in the western world that
bought into the initial 45,000 figure. We will wait with abated breath to
see if any will retract the original figure. I expect that nothing will be
done. Maybe that is why the 45,000
figure was thrown out there, because there is the universal recognition that
revision as respect to the truth is
not a habit in the western media world. Here are the misquotes:
Presse-Agentur via Monsters and Critics) Organizers say around 30,000 people gathered in Victoria Park before
beginning the march to the central government offices at around 3 p.m.
A pro-Beijing parade in the morning drew 30,000 people, according to organizers, while organizers of the pro-democracy march in the afternoon put their event's turnout at 45,000. The police said 20,000 participants attended the pro-Beijing event and 17,000 people arrived at the end of the pro-democracy march, a local radio station reported.
Times Online) Up to 45,000 join muted pro-democracy march in Hong Kong.
Organizers say around 45,000 people gathered for the march from Victoria Park to the territory's central government offices, but police estimated the turnout at only 17,000.
Press, Helen Luk, July 1)
Police said they counted 11,000 at the march's start, while the organizers said 45,000 joined the event.
[Note: Later copies were corrected to 21,000]
Press, Helen Luk, July 3)
Police said they counted 11,000 at the march's starting point, while the organizers said a total of 21,000 turned out for the
Also Friday, another large crowd gathered at Hong Kong Stadium for a patriotic parade with traditional lion dancers, marching bands, people dressed in Chinese opera costumes and marchers of all ages carrying small Chinese flags made of paper. Organizers said about 20,000 turned out, but police had no estimates.
[Note: Now this is wrong because the organizers said
about 30,000 turned out whereas the police said 20,000 left from Hong Kong
by Tan Ee Lyn, July 1)
Muted protest as Hong Kong marks China handover.
The protest organizer, Civil Human Rights Front, estimated the crowd at 45,000, a mere fraction of the demonstrations in 2003 and 2004 which drew turnouts of 500,000. Local authorities gave no crowd estimate.
At around 4pm, the Civil Human Rights Front spokesperson admitted publicly:
the number of participants was less than last year, at only 45,000 persons.
Star, Martin Regg Cohn)
Police counted only 11,000 people, while organizers said 45,000 were on the streets carrying banners saying "We want democracy."
- (United Press International via Washington
Organizers said 45,000 people took part in the march, but police and observers put the figure closer to 15,000.
- (AFP, Mark McCord, July 3; via China
Post) Hong Kong activists were struggling to put a brave face on an annual march for democracy in the Chinese territory Friday after it drew far fewer protesters than previous years.
They denied the city's once formidable campaign for universal suffrage had lost its bite despite the turnout being estimated by organizers at 21,000 people, way down on the half a million that marched for reform the past two years. Police put the figure at 11,000.
[Note: Hurrah, someone got it right!]
News) One media report said 10,000 people took part in the rally, compared to hundreds of thousands last year.
Century Net) According to BBC, the police said that 11,000
persons demonstrated, but the organizers said the number was 45,000.
Free Asia) The Civil Human Rights Front announced after the
marchers left Victoria Park that 45,000 people marched. But at 730pm,
they revised the number down to 21,000. [Note: They
got it right too!]
Century Net) According to Radio Free Asia, the march
organizers said that about 45,000 persons participated in the activities on
News, July 5) The number of protesters for the now traditional July 1 march for democracy in Hong Kong declined from approximately 250,000 last year and over 500,000 on July 1, 2003 to about 40,000 this year.
This is how things were at midnight. This
post will undoubtedly be updated when the details of the counting processes are
revealed. Meanwhile, let us track carefully who will revise and who won't
Update on July 3 midnight: Only AFP
and the revised AP report got the right 21,000 number. Overall, the
western coverage of the event was extremely sparse compared to 2003 and 2004
(nothing in the New York Times, Washington Post, The Guardian, etc). When
people march for democracy, the western media hype it up. When people
don't show, it is suddenly not newsworthy. By inference, one has to
conclude that the western media is agenda-driven and, as well, not especially
interested in the truth as the above showed.
Here are the reports on the numbers the next
day, and the various explanations and reactions in Hong Kong media as to why the 45,000 had to be
revised down to 21,000.
- (SCMP) Organiser the Civil Human Rights Front announced at 6pm that 45,000 people had taken part. But Front convenor Chong Yiu-kwong said an hour later that the figure was 21,000, blaming the discrepancy on errors in counting.
The University of Hong Kong's Public Opinion Programme put the turnout at 18,000 to 21,000 while police said it was 17,000.
Pro-democracy lawmaker Lee Cheuk-yan estimated 40,000 people had taken part.
Albert Cheng King-hon, another pro-democracy legislator, said the Front might have revised the turnout as it was afraid that the original estimate might be challenged. The Front claimed 530,000 last year but police put the crowd at 200,000.
of America) The headline in the English-language South China
Morning Post was: "21,000 march for democracy, 30,000 parade for patriotism,
200,000 attended the dinosaur exhibit at Tai Koo Shing" was the most
Sun) The Civil Human Rights Front announced after half an hour
after the march began that the estimated number of marchers was
45,000. In the evening after the march ended, the number was revised
down to 21,000, or 24,000 fewer.
Civil Human Rights Front convenor Chong Yiu-kwong explained the method of counting. The workers were posted near the Wanchai and Admiralty MTR
stations. Every five minutes, they counted the number of marchers
during one minute and multiply that number by five for the total flow.
During the initial half hour, the crowd was relatively dense and therefore
the estimated number of people was higher. In the middle and end of
the march, the density was significantly smaller. After the march was
over, they took the average of the total number of people counted at the two
locations and the answer was 21,000.
Daily) The Civil Human Rights Front originally predicted
50,000 to show up. At one time, they announced 45,000 participants,
but the final number was significantly less. Chong Yiu-kwong explained
that the Civil Human Rights Front had four teams at Wanchai and
Admiralty and they counted the number of people going past a fixed position
every minute out of five minutes and multiplying the number by 5. The
error occurred because they used the count for the first 30 minutes to come
up with the estimate.
Wei Po) Immediately after the march ended, the Civil Human
Rights Front told the media that 45,000 attended. But then several
Civil Human Rights Front leaders got together and had a tense discussion of
30 minutes. Then spokesperson Chong Yiu-kwong announced that the
number was revised to 21,000. He explained that the big difference
emerged because the volunteers initially used the density at the front of
the march and applied it to the whole march. But the density was much
less in the back of the march, thus causing the error.
Pao) Hong Kong University Public Opinion Programme had a team
counting in Wanchai, and estimated that 15,000 people passed under the
pedestrian overpass at the intersection of Hennessey Road and Arsenal Road. Adjusting for those who left before or after the point, the
estimated number of marchers was between 18,000 and 21,000.
If the Civil Human Rights Front insists that 530,000 people showed up last
year, then this represents a 96% drop.
Tao) At 530pm, the Civil Human Rights Front said that more
than 45,000 persons have gone past the Wanchai MTR station, but they later
revised the number of people past that point to be 21,000. Civil Human
Rights Front spokesperson Chong Yiu-kwong explained that the 45,000 was
based only up on the first half hour and then extrapolated from the number
of people past Wanchai. But the back of the march was less denser and
therefore the extrapolation was imprecise. At Wanchai and Admiralty,
the Civil Human Rights counted one minute out of five minutes and the final
count was 21,000. Although last year's number was subjected to doubt,
Chong Yiu-kwong insisted that there were 530,000 persons. In other
words, the number dropped by 96% this year, but Chong insists that this
year's methodology was rigorous.
Lee Cheuk-yan and Albert Cheng both expressed doubts, because they believe
that 20,000 is an 'underclaim.' Lee Cheuk-yan said that based upon
what he saw, there should be 40,000 people. Albert Cheng believed that
the Civil Human Rights Front typically 'underclaims': "When there was
one million, they said 500,000. They are worried about being
But the police, HKU's Robert Chung and HKU Department of Statistics and
Actuarial Science Senior Lecturer Yip Sui-fei all had obtained counts
similar to the Civil Human Rights Front. Yip Sui-fei said that he
accepts the Civil Human Rights Front estimate this year because it is
relatively close to reality. By contrast, when some democrats claim
that there may be 40,000 or 50,000, that is lacking objectivity as well as
unscientific. Yip said: "Perhaps because they 'inflated' the
number last year, and it becomes difficult to accept the reality this
The July 1 march organizer Civil Human Rights Front announced at one time
that the number of marchers was 45,000 and then suddenly revised the number
downwards to 21,000. This caused many march participants to complain
because they believe that the Civil Human Rights Front understated the
number. Some people even called our newspaper to express their
dissatisfaction, and said that unless the Civil Human Rights Front change
the persons in charge, they will not participate in the July 1 march again.
The deputy convenor Yip Bo-lam for the Civil Human Rights Front explained
that although she felt that there were more than 21,000 participants, but
this was the number that the organization that had been given the assignment came
up with. They had no choice but to report as is, and she hopes that
the citizens will understand that.
Yip Bo-lam explained that the initial figure of 45,000 was based upon an
extrapolation from the first half hour count provided by the counting
organization. After the march ended, they received the final count
from Wanchai as well as Admiralty.
Legislator Lee Cheuk-yan also thought that the number of marchers was
understated. After the march, he was on a radio program and after the
organizers announced the final count, many listeners called the radio
program to criticize the inaccurate count. Lee asked citizens to stay
calm and not let the number affect their intent to march.
A Mrs. Ng who marched yesterday called our newspaper. She was very
emotional about the number announced by the Civil Human Rights Front.
She said that her family of six came back especially from the United States
to attend the July 1 march. They also did that at the first July 1
march. Based upon what she saw at the scene, there should be at least
40,000 people. When she heard that the Civil Human Rights Front
counted only 21,000 people, she said, "I was so emotional that I wanted
Legislator Leung Kwok-hung is a member of the Civil Human Rights Front and
he is very angry about the Civil Human Rights Front underestimating the
number of marchers. He intends to hold a press conference on Sunday and
he will demand a re-count based upon the films taken during the march.
POP) The student researchers obtained these counts at the
pedestrian overpass at the intersection of Hennessey Road and Arsenal Road:
Time: 430pm-459pm: Lane 1: 1,832; Lane 2: 1,775; Lane 3: 1,830. Total:
Time: 500pm-529pm: Lane 1: 2,078; Lane 2: 2,763; Lane 3: 2,452. Total:
Time: 530pm-559pm: Lane 1: 575; Lane 2: 1,114; Lane 3: 857. Total:
Total: Lane 1: 4,485; Lane 2: 5,652; Lane 3: 5,129. Total 15,265.
Last year, HKU POP conducted telephone interviews with those who claimed to
have participated and found that 77.4% had passed through the Arsenal Road
counting station with a 95% confidence interval of plus or minus 5.5%.
Consequently, the adjustment factor between 1.20 and 1.39 applied to the
raw count. Multiplying 15,265 by 1.20 and 1.39 would yield an estimate
of the total number of marchers between 18,000 and 21,000. HKU POP
will be running the telephone interviews again this year.
Daily, July 4, 2005)
The number of July 1 marchers continued to be disputed. According to
information, the Civil Human Rights Front initially announced the number of
marchers to be 45,000 due to an incorrect interpretation of the data from
the Professional Teachers Union. The Professional Teachers Union used the first
person to go past the Wanchai counting station as the starting point, and
counted that about 10,000 had passed by after the first half hour (technical
note: this is different from the HKU POP data where the first half hour
started at 430pm, even though nobody had appeared yet). But the Civil
Human Rights Front assumed that the entire march would take about two hours
to go from Victoria Park in Causeway Bay to reach the Government
Headquarters in Central, and so they multiplied the number by a factor of
four and estimated that the number of marchers is 45,000. And they
announced it to the world.
The problem was that the entire march took only one hour and ten minutes for
everyone to file past the Wanchai counting station. In other words,
the accurate method would have been to mutiply 10,000 (which was the count
for the first half hour) by 2.1 to obtain 21,000. The Civil Human
Rights Front later explained to the Professional Teachers Union about the reason
for reporting the wrong number.
Cheung Man-kwong said that the counting method was accepted by both HKU POP
director Robert Chung and Civil Human Rights Front deputy convenor Yip
Bo-lum. The counting was done by teachers and university students, and
it was impossible for them not to report what they actually got. As
for some participants and Civil Human Rights Front members not willing to
accept the results, Cheung pointed out that 20,000 is quite a lot
already. "In reviewing the history of self-organized
demonstrations (excluding June 4 and those organized by the leftists), very
few of them have more than 20,000 people. If you understand this
point, then you will realized that the Civil Human Rights Front has actually
done quite well."
(Ming Pao via Yahoo!
[translation] The Civil Human Rights Front is holding an
informal meeting to discuss the third annual July 1 march held last
Friday. Apart from letting the core members express their thoughts, the
inevitable question would be how to deal with the number of marchers.
So far, the the Civil Human Rights Front has
held three July 1 marches. The first one in 2003 drew 500,000 people out
on the streets and successfully stopped the Article 23 legislation, thus
making the Front heroes to the people. Last year, the number dropped as
the academic community said that the number of marchers was around 200,000 but
the core members of the Civil Human Rights Front insisted that there were
530,000, 'breaking' the preceding year's record. As a result, the front
was heavily criticized for "exaggeration" and
This year, the Civil Human Rights Front is
facing a calmer social atmosphere. Under these circumstances, the number
of marchers dropped drastically. Actually, everybody realized that, so
the numbers do not mean too much. The Front asked the Professional Teachers Union
to conduct the count, and the final number was 21,000, which was
close to the numbers from Hong Kong University and the police
department. One would think that the Civil Human Rights Front has turned
over a new leaf and become more realistic and pragmatic.
Unfortunately, perhaps because the Civil
Human Rights Front lost out to the pro-government parade in terms of the
attendance counts, this pragmatic approach towards counting has upset certain
extreme elements in the march. In their opinion, the Civil Human Rights
Front is too conservative and do not understand the key in political movement.
Leung Kwok-hung of the April 5 Movement, a
member of the Civil Human Rights Front, fired the first salvo and accused the
Front of "under-reporting". He believes that the July 1 march
had at least 30,000. The April 5 Movement collected 80,000 dollars in
contribution in Wanchai. He said that Robert Chung's HKU student team
only counted people at Wanchai, and they should have counted also at Causeway
Bay, Wanchai and Central in order to accurately come up with the number of
people. He said that he iis collecting on how various countries around
the world count the number of people and he will disclose that information to
the public soon.
Since public opinion seems to sway one way
and then other, the core members of the Civil Human Rights Front appear to be
somewhat lost. Should they act honestly, or should they suitably
"over-state" the number of marchers in order to please the
participants? How are they supposed to run a political movement?
the lesson from 2004 had not sunk in yet -- claiming a stratospheric number
without any basis allowed the discussion to be hijacked. You want to
talk about freedom and democracy; but your audience and the media preferred to
ask you once again about that ridiculous number instead. If the CHRF had stuck to
the 45,000 number, it will be death by a thousand nicks once again.
of course it is eye-opening for people to state publicly that the politically
mature approach is to claim a stratospheric number in order the please the
participants! It is acceptable, even essential, to lie this way because your
core listeners will like it! Remember the names of those politicians and
never ever trust anything that they ever say again!]
[SCMP] July 1 march organiser sticks with 21,000 tally.
By Ambrose Leung. July 5, 2005.
The organiser of the July 1 march sees no need to revise its estimated turnout of 21,000, despite discontent from marchers who said it was inaccurate.
Civil Human Rights Front convenor Chong Yiu-kwong admitted the way the figure was announced had been unsatisfactory but stood firm on the figure itself.
"Perhaps we should have waited for the Professional Teachers Union to finish their headcount before we announced the final figure. If we didn't do it well this time, we will do better next time," he said.
Members of the union had been asked to count the number of people marching.
Because of a misunderstanding, the front announced before the march was over that 45,000 had taken part. It later revised the figure to 21,000.
(Ming Pao via Yahoo!
News) The quality of the democracy based upon how the number of
marchers was handled. Yip Sui-fei. July 14, 2005.
[The author is a senior lecturer in the Hong
Kong University Department of Statistical and Actuarial Sciences.
[translation] The counting of the
number of marchers is not simply a problem of "counting the number of
bodies." It involves the selection of a model that included
population movement, and that is a specialized subject in the biological
sciences. This author and a group of graduate and undergraduate students
have conducted in-depth research studies of the number of marchers on July 1
the last three years. Our goal was to implement the theory for society
to see, and to address the public's concerns and misunderstandings about the
number of marchers.
Over the past three years, our group set up
two counting stations, one near the Causeway Bay Electronics City (near the
Victoria Park starting point) and the other at the pedestrian overpass near
Tai Koo Plaza in Admiralty (and near the Government House). At each
counting station, we had 3 workers who counted the number of marchers
independently. Furthermore, for the purpose of estimating the number of
people who joined or left the march midway, we conducted a sample survey at
the second counting station by asking "Did you start from Victoria Park
for the march?" This formed the basis for estimating the number of
If the response to the survey question was
100% "Yes", then this means that everyone who passed through the
second counting station passed through the first counting station as
well. If the response is 0% "Yes", then this means that
everyone who passed through the first counting station left before the second
station, and so the total number of marchers is the sum of the counts from
At this year's July 1 march, the group
counted 16,800 and 16,900 respectively at the Causeway Bay and Admiralty
counting stations. We obtained more than 100 respondents for the
intercept survey study at Admiralty, of which 94% said that they started at
Victoria Park. In other words, 6% of the people joined the march
somewhere between the first and second counting station, and this is smaller
than the 23% at last year's march. The total number of marchers is
therefore estimated as the sum of the number of marchers at Causeway Bay plus
6% of the number of marchers at Admiralty. That is, 16,800 + 16,900 x
0.06 = 17,800 approximately.
This calculation does not include:
1. The people who departed from Victoria Park, but left before Causeway Bay.
2. The pepole who joined after the first counting station and left before the
second counting station.
3. The people who joined after the Admiralty counting station.
From the personal observations of the group
members, the march was orderly and smooth. We believe that the number of
people in these three groups could not be very high. We estimate the
sampling error around our estimate to be around plus or minus 800.
Overall, our estimate is very close to the police estimate.
This year, the organizers estimated 21,000
people marched. The police estimated 17,000. Our estimate was
18,000 and quite close. If everybody now considers that during the past
two years, the numbers between the organizers and the police were off by a
factor of two to three times, then this deserves some attention. The
police did not obviously change their methodology of counting over the past 3
years, but the organizers voluntarily changed their methodology of counting
this year and then the difference in the results was obvious.
All along, everybody has been skeptical
about the estimates from the organizers and the police. People often
think that the former overstates while the latter understates, and some people
even think that the average of the two is closer to the truth. In truth,
the fact that the organizers decided to choose a pragmatic approach this year
was a vast improvement as they counted the number of marchers in a scientific
and independent manner. They were able to rebut the criticisms about
under-claiming which have no basis, and this was a responsible way to
act. By comparison, the attitude of some people who 'estimated' the
number of marchers based upon their personal intuition and then criticized the
organizers in unreasonable ways because they feel that must always fight or
that they possessed the truth was quite disappointing.
For the 2003 and 2004 July 1 marches, the
police estimates were quite close to the estimates obtained by our
group. This year, the organizers had an estimate close to the estimates
from the police and academic researchers. This fact will allow certain
people to forget about the uncalled-for conspiracy theory and consider the
police methodology in a more open and considered manner that is consistent
with democratic principles.
The road for the democratic movement in Hong
Kong is long. The number of marchers should not be regarded as the sole
indicator for the desire of the people of Hong Kong for democracy, and this is
something that the organizers, the Hong Kong government and the Chinese
central government should be keenly aware of. Yet, from this matter, we
can see that the different ways in which the number of marchers were handled
have shown everybody that the quality of democracy has risen up this
year. So we hope that we can all continue to work together to improve
the quality of our democracy and create a bright future for Hong Kong.
(SCMP) Mock referendum biased: academics.
By Klaudia Lee. July 24, 2005.
The results of a mock referendum on universal suffrage in 2007 and 2008 are biased, academics monitoring it said yesterday.
However, they said the exercise had achieved its educational aims. Both the academics and the referendum organiser, the Civil Human Rights Front, said it could pave the way for future referendums conducted by the government or an impartial non-governmental body.
The mock referendum, conducted by the front, which also organised the July 1 rally, found more than 98 per cent of 7,725 people who cast votes supported universal suffrage to elect the chief executive in 2007 and the legislature in 2008.
But the independent "observers' report" released yesterday found that the lack of resources and proper venues "had severely handicapped the operation of the mock referendum" as only a limited portion of the general public could vote.
"Moreover, since the mock referendum was organised by a political group which also organised the July 1 rally, and in the same venue as that of the rally, it naturally attracted people who supported the motions. These factors caused biases," the report said.
The report, complied by Kenneth Chan Ka-lok, associate professor of the Hong Kong Baptist University's department of government and international studies, and Robert Chung Ting-yiu, director of the University of Hong Kong's Public Opinion Programme, outlined 11 shortcomings and strengths of the exercise.
The report found the "yes" vote was 24 to 30 per cent over-represented, and the "no" vote under-represented by eight to 12 per cent.
Despite this, Dr Chan and Dr Chung said a POP survey conducted after the referendum found that the results of the mock referendum were "in line with the patterns of public opinions towards democratic reforms."
The front issued a report in which they denied the exercise was a "small-circle election", saying they welcomed anybody to cast their votes no matter which view they held. It also showed they spent $55,483 - $10,467 less than $65,950 they initially collected - on the exercise.