1. Background

On July 29th Bersih 2.0 announced that a rally entitlted ‘Bersih 4’ would be held on the streets of Kuala Lumpur, Kuching and Kota Kinabalu from August 29th, 2pm to August 30th. The demands of the rally are for Prime Minister Najib Razak to step down and the following institutional reforms to be implemented:

  1. Clean Elections
  2. Clean Governments
  3. Saving Malaysia’s Economy
  4. Right to Dissent
  5. Strengthening Parliamentary Democracy (added on August 14th)

On August 14th Bersih released a statement adding a demand for a transitional government to be formed after Najib’s resignation. This government would need to implement 10 institutional reforms within the next 18 months to ensure the next General Election would be conducted in a clean, free and fair manner:

  1. Reform of electoral system and process
  2. Reform of the Election Commission (EC)
  3. Separation of Prime Minister and Finance Minister
  4. Parliamentary Reform
  5. Separation of the functions of Attorney General and Director of Public Prosecution
  6. Reform of the MACC
  7. Freedom of Information laws
  8. Asset declaration by Ministers and senior state officials
  9. Abolishment of/Amendment to draconian laws
  10. Establishment of the Independent Police Complaints and Misconduct Commission (IPCMC)

On August 27th Opposition MPs from PKR, DAP and GHB released a joint statement declaring they would work with BN MPs to form a new government provided that Anwar Ibrahim and other prisoners of conscience be released; and political reforms be the core agenda of the new government.

2. Twitter Statistics

All statistics referring to tweets include retweets unless otherwise stated.

2.1 Basic Stats

111,879 users made 583,338 tweets about Bersih from July 28th – August 30th 2015.

96,890 users made 446,967 tweets about Bersih during the rally period (August 29th – August 30th 2015). In other words, 86.6% of the total users and 76.6% of total tweets were made during the rally.

The chart below shows a comparison in the number of users tweeting about the Bersih 2 (July 2011), Bersih 3 (April 2012) and Bersih 4 (August 2015) rallies in the days leading up to the event.


The number of users for each rally on Day -1 (the day before the rally), Day 0 (the rally day) and Day 1 are listed below:

Rally Day -1 Day 0 Day 1
Bersih 2 7,083 19,190 7,492
Bersih 3 9,974 26,497 11,936
Bersih 4 13,432 70,527 52,552


2.2 Spammers

1,964 users spammed 45,924 tweets about Bersih during the rally period. This represents 2% of total users and 10% of total tweets during the rally.

2.3 Geo-tagged Maps

The map below shows the location of geo-tagged tweets mentioning Bersih. Yellow markers are tweets mentioning Bersih, red markers are tweets with images.


This map shows a close-up of Malaysia:


This map shows a close-up of Peninsular Malaysia:


This map shows a close-up of Kuala Lumpur where the Bersih rally was held. Blue markers indicate geo-tagged political tweets:


2.4 Interest by State

Both Facebook and Twitter are mediums that are used more by people in KL and Selangor with 49.44% of Facebook users and 48.6% of Twitter users residing in both territories. On Facebook 52.46% of interest in Bersih is from users in KL and Selangor.

Unlike previous Bersih rallies we do not have to depend on the limited sample provided by geo-tagged tweets.

Based on a listing of 40,244 users based in Malaysia from the total 96,890 users, we can measure the level of interest by users in each state by matching non geo-tagged tweets about Bersih to their user IDs.

The results are listed below. The ‘Diff from Msia’ column shows the difference in percentage points between Bersih’s state distribution and Malaysia’s state distribution (Msia %). This helps illustrate how balanced the distribution is.

State % of Total Msia % Diff from Msia
Johor 8.66 10.36 -1.69
Kedah 4.18 4.5 -0.32
Kelantan 2.51 2.7 -0.19
KL 17.18 17.25 -0.07
Labuan 0.16 0.23 -0.07
Melaka 4.41 4.34 +0.07
Negeri Sembilan 3.82 3.72 +0.10
Pahang 4.55 4.63 -0.08
Penang 3.67 4.56 -0.89
Perak 5.52 5.35 +0.17
Perlis 1.27 1.04 +0.23
Putrajaya 0.76 0.82 -0.06
Sabah 1.76 3.01 -1.25
Sarawak 2.03 3.42 -1.39
Selangor 36.92 31.35 +5.57
Terengganu 2.59 2.71 -0.12

54.1% of Twitter users based in Malaysia tweeting about the rally were from KL and Selangor. This was due to the increased interest of +5.57% from users in Selangor.

Other states that showed increased interest were Melaka, Negeri Sembilan, Perak and Perlis. Users Johor, Sabah, Sarawak and Labuan showed less interest.

2.5 The scale of conversation

This network graph shows how Twitter users tweeting about Bersih were connected. This includes all users on Twitter limited to those who retweeted or had conversations with each other. This helps visualise the scale of the conversation.


Each user is represented by a node (circle) that is coloured based on the number of their tweets that were retweeted and the number of tweets sent to them. The more attention they receive, the larger the node. Any node that retweets another node or tweets to another node is connected.

Nodes are positioned based on their connections to other nodes – strong connections pull them closer. Large nodes are considered influential within the network. We have coloured the nodes based on a scale of blue (least influential) to green; yellow; orange; red; and purple (most influential).

Due to the scale of the graph, we can only show names from the top 412 most influential users as seen below:


There are 86,733 users with 274,711 connections within the unfiltered graph. The most popular users were:

  1. 501awani
  2. adrianlimcheeen
  3. hafizrayyan
  4. mkini_bm
  5. malaysiakini
  6. bersih2
  7. thesundaily
  8. rafiziramli
  9. themmailonline
  10. malaysla
  11. abdmalekhussin
  12. santaidansampah
  13. hafizhamidun
  14. annabellybutton
  15. klubbkiddkl
  16. cheanchung
  17. netrakl
  18. a_marea
  19. mhmmdsyhmeee
  20. newsbfm
  21. faris_shafii
  22. melgoh
  23. hilalazmi
  24. gengbebel
  25. jeremyjwwt
  26. staronline
  27. jarniblakkarly
  28. fadiahnadwa
  29. bongkersz
  30. hairienorli
  31. milosuam
  32. iapplemustache
  33. haritskarev
  34. adrianncf
  35. hohohoelaine
  36. timothytiah
  37. hannahyeoh
  38. chelle_yesudas
  39. rachelyappppp
  40. syedsaddiq
  41. imokman
  42. saysdotcom
  43. jaspreetaulakh_
  44. bharianmy
  45. justintwj
  46. kama7esh
  47. abelljefrry
  48. twt_malaysia
  49. baitialjannati
  50. sampahremajakl
  51. stankidrauhl
  52. ladymissazira
  53. _jamesjemz_
  54. lawaktahapdewa
  55. jaerocks
  56. tm_insider
  57. bedlamfury
  58. ezrazaid
  59. zurairi
  60. najibrazak
  61. nichocheng
  62. putrareformasi
  63. channelnewsasia
  64. tianchua
  65. sumishanaidu
  66. ericpaulsen101
  67. twt_gagalis
  68. asrulmm
  69. keowweeloong
  70. peteteo
  71. pisaukarat
  72. sinaronline
  73. fazallahpit
  74. bncybertrooper
  75. kinggopoh
  76. bernamadotcom
  77. firdausdarus
  78. ciakhalid
  79. gman_h
  80. stcom
  81. saysseismik
  82. _amranfanz
  83. haaziq_mz
  84. ikhwanredzuan_
  85. norman__g
  86. khairilanuar
  87. nizambakeri
  88. my_mgag
  89. radeonmbps
  90. bernamaradio24
  91. sitifyruz
  92. khalids
  93. koksenwai
  94. globalbersih
  95. juntothehan
  96. eida0109
  97. faktabukanauta
  98. reaproy
  99. satsrvp
  100. theasyrafzabani

3. Racial Sentiment

During the first day there were expressions online about the ethnic composition of the rally. By our own estimate, over half of the protesters appeared to be Chinese. This prompted some users to remark on ‘how Chinese’ the rally was, and the lack of Malays in the crowd.

During the Bersih 3 rally in 2012 images were circulated online showing people of different ethnicities helping each other especially after police action took place.

During Bersih 4 images highlighting kind gestures by Chinese towards others (primarily Malay Muslims) were popularly circulated. Examples below:


The same images were used together with different messages:

There were also users circulating photos highlighting people of different colour:

To get an idea of how significant race or religion-related remarks were in relation to Bersih, we can check for correlation in tweets using English and Bahasa Malaysia terms mentioning race or religion.

Out of the 446,967 tweets made by 96,890 users about Bersih during the rally:

  • 7,030 users made 11,087 tweets mentioning Malays
  • 5,384 users made 8,471 tweets mentioning Chinese
  • 898 users made 957 tweets mentioning Indians
  • 7,709 users made 8,619 tweets mentioning Muslims
  • 7,357 users made 7,999 tweets mentioning non-Muslims (specifically ‘bukan Islam’ and similar. This does not include Buddhist/Hindu/Christian/etc.)
  • 15,277 users made 25,823 tweets mentioning the above terms and other racial remarks (e.g. rasis)
  • 10,295 users made 16,649 tweets mentioning ‘Malaysian’
  • 22,158 users made 41,711 tweets mentioning ‘Malaysian’ in addition to the racial terms. That means 6,881 users mentioned Malaysian without mentioning other races.

16% of the total users were talking about or sharing content with keywords mentioning race/religion and Bersih.

Whether the intention was good or bad it seems that racial/religious content did not get much traction within the Bersih tweets. 7% of total users mentioned Malaysian without mentioning race.

However when we look at tweets mentioning Malays, Chinese and Muslims (without Bersih keywords) there was some racial sentiment being expressed over the weekend in response to Bersih. This will require further analysis to determine whether the numbers are significant.

4. Misinformation

We have not done a detailed study on misinformation however the following misinformation were popularly shared across multiple tweets and retweets.

Doctored images:

The following photos are from Bersih 3 (2012). The 1st photo was reposted multiple times.

5. Crowd Measurement

The total attendance for Bersih 4 in Kuala Lumpur is estimated to have been between 79,919 – 108,125 people. It peaked on August 29th between 45,892 – 62,089 people.

It is not physically possible to fit 80,000 – 200,000 people within the occupied space shown in photographs online.

This number is not expected to be revised, unless some new evidence surfaces to suggest a different crowd size. This figure does not include Global Bersih events or events held in other cities in Malaysia.

5.1 Methodology

Photos and videos that showed the size of the crowd were evaluated. These were used to determine the extents of the area covered by the crowd and the crowd density. The area was then refined by removing obstacles and estimating the percentage of space used. The space was measured in square feet based on satellite images.

A gridded map was used to help estimate crowd density. An example showing aerial photos by Malaysiakini (link) is shown below. Each square is 25 square metres, which can fit up to 75 people (standing tightly packed). It is possible to fit up to 100 people but at that point density would be too high as people would be unable to turn/move:


Another example based on photos from Says.my (link):


5.2 Core Area

The shaded area in the image below shows the rough extents of the core area covered by Bersih 4 protesters. Smaller parts of the city where protesters were present are not shown. Additional extent where the crowd density was sparse is also not shown. This visual is just to give an idea of where the bulk of the crowd was at different times on both days.


The size of the area shown is 367,487 square feet. Based on photos that we have found, there was no point in time was this area completely filled with protesters.

By using this area as a starting point and adjusting for sparse areas; other areas nearby; buildings occupied by protesters; large obstacles and crowd density we can estimate the size of the crowd within 4 periods:

  • August 29th (Day 1)
  • August 29th (Night 1)
  • August 30th (Day 2)
  • August 30th (Night 2)

The results are shown in the table below:

Period Crowd Size Average
Day 1 45,892 – 62,089 53,991
Night 1 9,837 – 13,308 11,573
Day 2 14,782 – 20,000 17,391
Night 2 23,511 – 33,797 28,654

A percentage of protesters in one period likely attended the next period. Turnover is an issue as well, though there we found no mention of large numbers of people coming and going at the same time. The change in ethnic composition of the protesters from a Chinese-majority (Day 1) to a more mixed composition (Day 2) indicates more new-comers on the 2nd day.

There is no easy way to calculate the total protesters who attended Bersih 4. Due to this overlap, we present 2 methods to calculate this that may accommodate the overlapping crowd factor within a margin of error.

Method 1

Total the minimum crowd sizes for each period and adjust within a margin of error.

Total protesters = 45,892 + 9,837 + 14,782 + 23,511
= 94,022

Adjusted to +/- 15% margin of error, the total is 79,919 – 108,125 protesters.

Method 2

Day 1 saw a large number of protesters that reduced over time once it hit its peak in the afternoon. Day 2 saw a small number of protesters from the night before that grew over time.

Because of this, we can take the peak average values from both days and consider that to be the total protesters for each day.

Total protesters = Day 1 Average + Night 2 Average
= 53,991 + 28,654
= 82,645

Adjusted to +/- 15% margin of error, the total is 70,248 – 95,042 protesters.

For our records we will use the results of Method 1. The higher range should accommodate the turnover factor; the change in ethnic composition; and the likelihood that most protesters wouldn’t have attended all 4 periods.

6. Summary

With 79,919 – 108,125 people attending the rally in KL, Bersih 4 has made a historical achievement.

Based on Facebook statistics in our previous analysis (link) and tweets during the rally, Bersih attracts more interest from users in KL and Selangor compared to other states. Interest from users in East Malaysia remains low.

When it comes to race, 16% of the total users were talking about or sharing content with keywords mentioning race/religion. 7% of total users mentioned Malaysian without mentioning race, compared to 5.5% of total users that mentioned Chinese (including those who were highlighting Chinese together with other Malaysians).

This could be an indicator that the ethnic composition of those attending the rally was not a big issue on Twitter. However we will need to do a further study to determine that.

Published On: September 2nd, 2015 / Categories: Analyses, Statistics / Tags: , , , , , /