According to the Straits Times’ election website,
The turnout for WP’s Thursday rally, its first for this election, was estimated by observers to be over 20,000 people.
— Straits Times GE2011 website, 29 April 2011, Shanmugam says WP rally crowd ‘not unusual’, Link.
This would be the rally on Hougang field on 28 April 2011, for which there are plenty of pictures now circulating on the web. I wasn’t there but “over 20,000” is very far off what the number looks to be from the photos.
Before the Straits Times does another dampener on another Workers’ Party rally, this one held at Serangoon Stadium on Friday 29 April, let’s do a simple exercise to estimate numbers ourselves.
(The picture above of the Serangoon rally is stitched from several photos; click on it for a larger version)
To determine the number of people, we need to know only two starting numbers:
1. The area in square metres,
2. The density of people.
Even so, it is difficult if density varies from one part to another, but in the confined space of Serangoon Stadium, density was relatively even throughout. People squeezed in until the place could hardly accommodate any more, inserting themselves into any available space.
To determine the area, Google maps is extremely useful since it has a scale over the satellite image.
The image above is that of Serangoon Stadium. I drew in a white rectangle to show where the stage was, and marked out the dimensions from one edge of the crowd to the other.
The area is thus 130m x 160 m = 20,800 square metres, though we need to subtract a bit away because of the rounded corners at the northern end. Let’s say, the area that people filled is about 19,000 square metres.
The density on the bleaches is quite easy to tell just by looking. People were sitting shoulder to shoulder and about 1 metre behind each other. The density is therefore about 2 persons per square metre. You can see a bit of the stands towards the end of this 33-second video.
However, the bleaches (under the roof in the satellite photo) make up less than 10 percent of the total area. The great majority of the people in the stadium were standing on the field. What was the density there?
Here’s a telephoto picture of a spot roughly in Location D:
And here’s another picture, this time of Location E:
As you can see, the back of the stadium, furthest from the stage, was just as crowded as the middle, if not even more so, because it was closer to an entrance and people were wiggling their way in. As for the front, nearest the stage, it was too far for me to take a density check picture. But we can safely assume it was packed crazy out there.
What do you think the density is in the two pictures above? It looks to me to be at least 2 persons per square metre, possibly even 2.5 persons. It may even be 3 persons per square metre in the lower of the two pictures.
Multiply the area by density, and you’ll have the total number in the stadium. If you use 2 persons per square metre as average density, the total crowd would be 38,000. If you use 2.5 persons per square metre, the total would be 48,000. If you use 3 persons per square metre, the total would be 57,000 — but this, in my view is on the high side.
* * * * *
Virtually all of them in the stadium were opposition sympathisers. With such a crowd, does it indicate a likely victory for the Workers’ Party in Aljunied constituency?
Bear in mind, many of the people at the rally came from other parts of Singapore. If the crowd was drawn randomly from among all Singaporean voters, then only 6.1 percent of the number in the stadium would have been actual Aljunied voters. This is because the 143,148 voters in Aljunied make up 6.1 percent of the total 2,350,873 voters in Singapore for this election.
But let’s say Aljunied voters were over-represented at the rally. At the low end, we can assume that they made up twice their normal weightage, i.e. 12.2 percent; at the high end, we’ll assume they made up four times their normal weightage. Then, if we use the middle figure of 48,000 people in the stadium, only 5,800 to 11,700 of those at the rally were really Aljunied voters.
That’s not a big chunk of Aljunied voters. That’s only 4 to 8 percent of the 143,148 who will cast their ballots in that constituency next week. In other words, however enthusiastic the people were at the rally, it really says very little about how the vote will turn out in that constituency.