Politweet.Org

Observing Malaysian Social Media

An Analysis of Opinions on Tun Mahathir, Trends in Political Interest on Facebook and Political Support by Malay Youth in Malaysia

1. Introduction

This report is divided into the following sections:

  1. Introduction
  2. List of Acronyms
  3. An Analysis of Opinions on Tun Mahathir Working With the Opposition Following the Citizens Declaration in March 2016
  4. An Analysis of Opinions on Tun Mahathir as the PM candidate for Pakatan Harapan by Malay Youth in Peninsular Malaysia
  5. Facebook Trends for Interest in Tun Mahathir and PM Najib
  6. Malaysia’s Twitter Demographics Overview
  7. Interest in Tun Mahathir on Twitter
  8. Twitter Sentiment Analysis for PM Najib and Tun Mahathir
  9. Conclusion

Each section will explore indicators from Twitter and Facebook to determine the level of support for Tun Mahathir, PM Najib and political parties.

This study was originally published on our Facebook Page on May 9th, 2018.

A PDF copy of this report can be downloaded from https://drive.google.com/open?id=13EIGTHlY8eClRslaUNhgg30j9ywAW9Fj

Some important notes to remember when interpreting Facebook figures:

  1. Total population refers to Facebook users aged 13 years and above.
  2. Potential voters refer to Facebook users aged 21 years and above.
  3. Youth refers to Facebook users aged 13 – 20 years.
  4. Gender breakdown figures do not always add up to the total. This may be due to Facebook users not sharing their gender, and also due to rounding errors in statistics provided by Facebook. State breakdown figures also do not add up to the total due to the same rounding errors.
  5. Detailed statistics on Putrajaya are not available due to the small number of users in the territory.
  6. Figures provided by Facebook are estimates. Some inaccuracies are to be expected, e.g. the sum of state totals not being equal to the national total.
  7. Facebook users residing in Malaysia are not necessarily Malaysian citizens.
  8. Interest in a topic is equal to the number of users expressing interest in a topic.
    1. To measure interest we used a combination of Facebook Interests (a collection of interests, activities, groups, pages, status updates and job history identified by a common term determined by Facebook e.g. ‘United Malays National Organization’) and specific Group and Page names (e.g. Friends of BN).
    2. These are used to collect the number of users interested in a given party/coalition/politician/group. For example, a user mentioning a party name in a status update; sharing a news link related to the party or sharing content from a party-affiliated page would count towards the total interest in that party
    3. Interest in a political party does not indicate support for the party, only awareness
    4. It is currently assumed that interest in PAS includes some interest in AMANAH as PAS leaders and members migrated to AMANAH
  9. Audience refers to the population of active users that express interest in a topic. Unless indicated, the audiences used in this report are composed of potential voters (users in Malaysia aged 21 years and above).
  10. Based on our research to date, Pages that are of type ‘politician’ are not always included under related Facebook Topics. For example, not all ‘Tony Pua’ (MP, PJ Utara, DAP) Page likes are included under interest in ‘DAP’. However, because Facebook does not make Topic details available we cannot easily determine which politicians, if any, were included.
  11. Statistics on the Opposition primarily refer to component parties of the former Pakatan Rakyat – PKR, PAS and DAP. This includes the ‘Pakatan’ brand name.
  12. Unless otherwise specified, constituency statistics are based on the 2017Q4 electoral roll.

 

2. List of Acronyms

The following table shows a list of acronyms used in this document.

Acronym Full name
PR Pakatan Rakyat
PH Pakatan Harapan
BN Barisan Nasional
UMNO United Malays National Organisation
GERAKAN Parti Gerakan Rakyat Malaysia (also known as PGRM)
MCA Malaysian Chinese Association
MIC Malaysian Indian Congress
PBB Parti Pesaka Bumiputra Bersatu Sarawak
PKR Parti Keadilan Rakyat
DAP Democratic Action Party
AMANAH Parti Amanah Negara
PAS Parti Islam Se-Malaysia
PPBM Parti Pribumi Bersatu Malaysia

3. An Analysis of Opinions on Tun Mahathir Working With The Opposition
(March 3rd – 10th 2016)

3.1 Background

On March 4th 2016, following a campaign to pressure PM Najib Razak to step down as Prime Minister due to the 1MDB scandal in 2015, Tun Mahathir together with various political party leaders and NGO leaders held a press conference to launch the ‘Citizens Declaration’. (http://www.straitstimes.com/asia/se-asia/mahathir-and-opposition-sign-declaration-to-oust-najib )

The declaration had 4 demands (https://www.malaysiakini.com/news/332750 ):

  1. a) The removal of Najib as PM of Malaysia through non-violent and legally permissible means.
  2. b) The removal of all those who have acted in concert with him.
  3. c) A repeal of all recent laws and agreements that violate the fundamental rights guaranteed by the federal constitution and undermine policy choices.
  4. d) A restoration of the integrity of the institutions that have been undermined, such as the police, the MACC, Bank Negara and the PAC.

Among those in attendance at the launch were Tun Mahathir, Lim Kit Siang (DAP), Azmin Ali (PKR, Selangor Menteri Besar), Mukhriz Mahathir (UMNO, former Kedah Menteri Besar), Muhyiddin Yassin (UMNO, former Deputy PM), Ling Liong Sik (former MCA President), Mohamad Sabu (AMANAH), Maria Chin (BERSIH chairperson), Ambiga Sreenevasan (former BERSIH chairperson, former Bar Council President), Anthony Loke (DAP), Khairuddin Aman Razali (PAS Dewan Ulama information chief), Mustafa Ali (PAS election director) and Hishammuddin Rais (Activist).

When asked on who will be the next Prime Minister, Tun Mahathir stated that this is a question for a later time (https://www.malaysiakini.com/news/332716).

The Citizen’s Declaration was part of a campaign to ‘Save Malaysia’ (https://www.channelnewsasia.com/news/asiapacific/save-malaysia-campaign-to-oust-najib-brings-political-foes-toget-8160050 ).
We monitored the progress of both the Citizens Declaration and the Save Malaysia movement and found that while interest was strong at the outset, it was not sustained. Interest from local users in Malaysia started strong with 56,472 users for the month of March 2016 but eventually dropped to 2,138 users for the month of June 2016.

3.2 Our Analysis

We performed an analysis that focused on the response of Twitter users to Tun Mahathir working with the Opposition (which includes PKR, DAP and PAS). This was to gauge the support for the political rivals working together to oppose PM Najib.

We performed opinion-based analysis on 500 users based in Malaysia who tweeted about the Citizen’s Declaration, the Save Malaysia movement, Tun Mahathir and related terms from March 3rd – March 10th 2016. The margin of error is +/- 4.38%.

Users were selected based on their tweet content and activity during this period. Sampling was done per-state based on the current estimated user population. The age group was estimated to be between 21 – 30 years old.

Spammers, news agencies and accounts with automated tweets were not included in the sample.

From this dataset we analysed the individual Twitter user timelines to determine their opinion. This took their tweets, retweets and conversations into account. Only users who had an opinion about Tun Mahathir and the Opposition were used in the sample.

In our initial examination of conversations about the Citizen’s Declaration and the Save Malaysia movement, we found there was division of opinion on this issue. While there was a strong base of supporters who wanted PM Najib to step down, many users did not want Tun Mahathir and/or the Opposition to be involved. For example they might want PM Najib to step down and BN to stay in power.

Based on this analysis we categorised users as belonging to one of the following categories:

  1. Support Tun Mahathir Working with the Opposition
  2. Neutral
  3. Oppose Tun Mahathir Working with the Opposition

The results are shown in the following charts.

Summary

wp_mahathirpm_chart1

 

Category Users (%)
Support 213 43
Neutral 56 11
Oppose 231 46

 

70% of the users in the sample (350 users) expressed support for PM Najib to step down. Despite that they were divided on support for Tun Mahathir working with the Opposition.

3.3 Findings

What follows are the findings for each category.

3.3.1 Support Tun Mahathir Working with the Opposition

213 users (43%)

These users supported Tun Mahathir working with the Opposition to oppose PM Najib. The most common reasons stated for this were:

  • Believe Najib should go because he allegedly stole public funds and/or to stop the country from heading down the wrong path.
  • Cost of living, availability of jobs and rising national debt were the main worries that prompted people to support any pact to ‘save Malaysia’.
  • Want the Opposition (Pakatan) to take over the Federal government
  • Want PM Najib to step down and another BN leader to take over
  • 9 users (1.8%) wanted Tun Mahathir to take over and lead the Opposition

3.3.2 Neutral

56 users (11%)

These users were unsure on whether to support Tun Mahathir and the Opposition working together, or had conflicting opinions.

3.3.3 Oppose Tun Mahathir Working with the Opposition

231 users (46%)

These users did not support Tun Mahathir and the Opposition working together. The most common reasons stated for this were:

  • Don’t trust Tun Mahathir given his past record as Prime Minister, suspect he has a hidden agenda e.g. to keep UMNO in power
  • Blame Tun Mahathir for the current system of governance and do not believe he wants to fix it
  • Have not forgiven Tun Mahathir for past mistakes during his time as PM
  • Don’t believe Tun Mahathir is sincere as he has not apologised for past mistakes e.g. ISA detentions; the Memali incident; the 1988 judicial crisis; Anwar’s imprisonment in 1999; Project IC in Sabah; and the culture of cronyism and corruption
  • Support the Opposition but don’t support them working with Tun Mahathir
  • Want a new generation of leaders to take charge instead of old and former political leaders
  • Because they have a low opinion of the Opposition parties. They may have been more supportive of the Save Malaysia movement if the Opposition was not involved. Among the common opinions on the Opposition were:
    • Do not want the Opposition to take over the government, particularly DAP. PKR’s obsession with making Anwar Ibrahim the next Prime Minister was also criticised. Some users disliked Anwar Ibrahim, indicating support for moving on to new leaders.
    • Considered the Opposition leaders and activists to be hypocrites for working with Tun Mahathir
    • Do not want the Opposition to be empowered by working with Tun Mahathir
    • Considered the Opposition to be unstable
    • Some users expressed that they lost respect for Tun Mahathir for working with the Opposition
  • Do not support the Opposition using Tun Mahathir for their private agenda
  • Do not support Tun Mahathir using the Opposition for his private agenda

 

3.4 Results by State

The chart below shows the results by state for selected states, bearing in mind that our sample size is small at 500 users.

wp_mahathirpm_chart2

 

State Support (%) Neutral (%) Oppose (%)
Selangor 44.14 10.34 45.52
Kuala Lumpur 40.45 8.99 50.56
Johor 45.28 11.32 43.40
Penang 25.00 16.67 58.33
Perak 30.77 7.69 61.54
Kedah 54.17 8.33 37.50
Pahang 43.48 8.70 47.83
Melaka 55.00 15.00 30.00

 

What is important to note here is that support for Tun Mahathir and the Opposition working together was lowest in Penang and highest in Kedah and Melaka.

 

3.5 Support for the next Prime Minister

While campaigners of the Save Malaysia movement and the Citizen’s Declaration were not specific on who should replace PM Najib, users began discussing this issue on their own. We noted their responses and charted the result below:

wp_mahathirpm_chart3

 

Category Users (%)
Support BN for PM 54 11
Don’t Support Opposition for PM 136 27
Support Opposition for PM 77 15
No Opinion 233 47

 

These findings show that support for wanting PM Najib to step down does not translate into support for the Opposition to take over. There were more users opposed to the Opposition (primarily PKR and DAP) compared to users expressing support.

Users supporting Tun Mahathir and the Opposition working together only agreed on the issue of PM Najib’s resignation.

If we were to chart the same result just for the 70% of users who wanted PM Najib to step down, there is still less open support for the Opposition compared to BN or the anti-Opposition group.

wp_mahathirpm_chart4

 

Category Users (%)
Support BN for PM          37 10
Don’t Support Opposition for PM 79 23
Support Opposition for PM 77 22
No Opinion 157 45

 

This highlights the failure of the anti-Najib campaign to make the Opposition the ‘obvious choice’ to replace BN for the majority of youth on Twitter in March 2016.

We also observed that many Opposition supporters and Tun Mahathir supporters were not on the same page. Their expressed motivations can be described as follows:

  • Opposition supporters wanted Tun Mahathir’s influence to help the Opposition take over the Federal Government.
  • Mahathir supporters wanted the Opposition’s influence to push for PM Najib to be replaced with a BN leader.

 

3.6 About the Population Sample

The results reflect a young demographic, by our estimates to be between 21 – 30 years old. Users were primarily Bahasa Malaysia speakers:

  • 78% Bahasa Malaysia speakers
  • 15% English speakers
  • 7% Mixed/other language speakers

 

wp_mahathirpm_chart5

4. An Analysis of Opinions on Tun Mahathir as the Prime Minister Candidate by Malay Youth (21-30 yrs) on Twitter in Peninsular Malaysia
(December 3rd 2017 – January 17th 2018)

4.1 Background

On December 4th 2017, Pakatan Harapan formally proposed Tun Dr.Mahathir as its candidate for Prime Minister and Datuk Seri Dr.Wan Azizah as its candidate for Deputy Prime Minister. (https://www.thestar.com.my/news/nation/2017/12/04/pakatan-proposes-dr-m-for-pm-decision-meets-with-opposition-from-some-coalition-members/)

This decision was finalised on January 7th 2018 and Tun Mahathir was confirmed as Pakatan Harapan’s choice for PM along with Wan Azizah as the Deputy PM (https://www.thestar.com.my/news/nation/2018/01/07/mahathir-unanimously-chosen-as-pakatan-harapan-pm-candidate/).

The plan was to have Tun Mahathir as an ‘interim PM’ with Anwar Ibrahim to take over as PM following a royal pardon and presumably a by-election to qualify him as an MP. There was no mention on who would be Anwar Ibrahim’s deputy or when exactly would the transition take place. (https://www.themalaysianinsight.com/s/31286/)

4.2 Our Analysis

We performed an opinion-based analysis that focused on the response of Twitter users to Tun Mahathir being the PM candidate for Pakatan Harapan. This was to gauge the support for Tun Mahathir as the next PM. We also took note of political support expressed by the users, not only to measure how they plan to vote but also the impact of Tun Mahathir’s candidacy on strong PH supporters.

Unlike the previous study on support for Tun Mahathir and the Opposition working together, this study is focused only on ethnic Malays in Peninsular Malaysia instead of all races across the country. This is because we encountered some issues that are explained below:

  • When we initially took a non-racial sample, we found that only 5.7% were non-Malays. After closer examination of the total population of Twitter users tweeting about Tun Mahathir, we determined that the level of interest from non-Malays was too low, particularly Chinese users.
  • This led us to revise the study to only focus on ethnic Malays. Any relevant findings from opinions by non-Malays will be summarised separately.
  • In addition to that there were too few users in East Malaysia showing interest in Tun Mahathir during this period, so our study only focuses on Peninsular Malaysia users.
  • There were also too few BN supporters expressing opinions about Tun Mahathir. Instead they chose to promote BN politicians and pro-BN content. The views of BN supporters are considered to be under-represented in this study.

We performed our analysis on a sample of 602 users based in Peninsular Malaysia who tweeted about Pakatan Harapan, BN, PM Najib, Tun Mahathir and related terms from December 3rd 2017 – January 17th 2018. The margin of error is +/- 4%.

Users were selected based on their tweet content and activity during this period. Sampling was done per-state based on the current estimated user population. The age group was estimated to be between 21 – 30 years old.

Spammers, news agencies and accounts with automated tweets were not included in the sample.

From this dataset we analysed the individual Twitter user timelines to determine their opinion. This took their tweets, retweets and conversations into account.

Based on this analysis we categorised users as belonging to one of the following categories:

  1. Support Tun Mahathir as PM Candidate
  2. Neutral
  3. Oppose Tun Mahathir as PM Candidate

The results are shown in the following charts.

Summary

wp_mahathirpm_chart6

 

Category Users (%)
Support Tun Mahathir as PM Candidate 104 17.28
Neutral 73 12.13
Oppose Tun Mahathir as PM Candidate 425 70.60

 

44.02% of the users (265 users) expressed negative sentiment towards PM Najib.

18.27% of the users (110 users) were categorised as strong PH supporters based on their stated party membership or tweeting history (e.g. reliably expressed support for PH).

 

4.3 Findings

What follows are the findings for each category.

4.3.1 Support Tun Mahathir as PM Candidate

104 users (17.28%)

These users supported the nomination of Tun Mahathir as Pakatan Harapan’s PM candidate. The most common reasons stated for this were:

  • Believe PM Najib should be replaced regardless of who the PM candidate is
  • Believe Tun Mahathir is the best PM Malaysia ever had
  • Believed that quality of life was better under Tun Mahathir when he was PM, and expect him to return Malaysia to those good times. It is important to note that these users were only children or teenagers when Tun Mahathir was PM, so Tun Mahathir appeals to their childhood nostalgia
  • Pity Tun Mahathir due to his old age, and willing to give him a second chance
  • Dislike Tun Mahathir, but believe he is the lesser evil compared to PM Najib
  • Wrongly believed that Tun Mahathir had apologised for past wrongs, and were willing to give him a second chance
  • Oppose perceived corruption by the BN government and want a change, no matter who the PM is
  • Want Pakatan Harapan to take over the Federal government
  • 15 users (2.49%) believed Tun Mahathir was necessary for PH to make headway in rural Malay areas

 

4.3.2 Neutral

73 users (12.13%)

These users were unsure on whether to support Tun Mahathir as the PM candidate, or had conflicting opinions. A number of them declared they would wait until polling day to decide who to vote for.

 

4.3.3 Oppose Tun Mahathir as PM Candidate

425 users (70.60%)

These users did not support Tun Mahathir as the PM candidate. The most common reasons stated for this were:

  • Identical reasons to our previous study on Tun Mahathir and the Opposition working together:
    • Don’t trust Tun Mahathir given his past record as Prime Minister, suspect he has a hidden agenda e.g. to keep UMNO in power
    • Blame Tun Mahathir for the current system of governance and do not believe he wants to fix it
    • Have not forgiven Tun Mahathir for past mistakes during his time as PM
    • Don’t believe Tun Mahathir is sincere as he has not apologised for past mistakes e.g. ISA detentions; the Memali incident; the 1988 judicial crisis; Anwar’s imprisonment in 1999; Project IC in Sabah; and the culture of cronyism and corruption
  • Disappointed by his insincere apology made on December 30th 2017 (http://www.straitstimes.com/asia/se-asia/dr-mahathir-says-sorry-for-past-mistakes-as-he-promises-bold-steps-to-remake-malaysia) when he clarified that apologising did not mean he did anything wrong (http://www.themalaymailonline.com/malaysia/article/dr-m-saying-sorry-doesnt-mean-i-was-wrong). This came following a statement made by Tun Mahathir on November 1st 2017 where he accepted the blame for the Ops Lalang detentions but claimed that it was not his decision, and did not apologise (http://chedet.cc/?p=2657). Tun Mahathir’s refusal to apologise for specific mistakes was an important issue.
  • Do not believe Tun Mahathir will give up his post for Anwar Ibrahim
  • Believed that Tun Mahathir would give the PM or DPM post to his son Mukhriz instead
  • Suspect Tun Mahathir may join BN/UMNO, taking PPBM MPs with him
  • 107 users (17.77%) wanted a new generation of leaders to take charge instead of Tun Mahathir or Anwar Ibrahim
  • Could not understand why Tun Mahathir was nominated as the PM candidate instead of other Pakatan leaders
  • Were strong supporters of the Reformasi movement and/or PKR, and therefore could not accept Tun Mahathir as the PM candidate
  • Disliked the concept of an ‘interim PM’ as it is too similar to the Kajang Move, when a by-election was held to replace the Selangor Menteri Besar.
  • Having Wan Azizah as the possible ‘seat-warmer’ DPM was also viewed negatively
  • Lack of interim plan details (ie. How long Tun Mahathir will be PM; who will step down for Anwar; etc.) increased distrust of Pakatan and Tun Mahathir
  • Believed PM Najib is the lesser evil compared to Tun Mahathir
  • Believed that Tun Mahathir already had his turn as PM for 22 years and it was time for someone else to lead
  • 9 users (1.5%) expressed support for Tun Mahathir to become an advisor to the next PM instead of the PM candidate

 

4.4. Other Opinions

The following views were also shared by the sample of 602 users.

4.4.1 Views on the Interim PM Plan

The majority of users discussing the interim PM plan did not view it favourably. Critics compared it to the Kajang Move (October 2014) when a by-election was held to replace the Selangor Menteri Besar. Most of these users could not understand why Tun Mahathir was chosen instead of another PH leader. Tun Mahathir was also not considered to be trustworthy enough to honour the plan.

Supporters of the plan were more interested in Anwar Ibrahim becoming PM, so any plan to make that happen was acceptable.

From our own observations since the Kajang Move, it was an incident that did long-term damage to PKR’s reputation among the Malay community on Twitter and Facebook. People keep bringing it up on social media when they criticise PKR. On Facebook, interest in PKR (from users aged 21 and above in Malaysia) dropped from 2.2 million users in September 2014 to 1.48 million users by November 2014. It wasn’t until April 2016 when PKR was able to reach 2.2 million users again on Facebook. Since then interest in PKR on Facebook has declined to 1.3 million users as of March 2018.

Having Wan Azizah as the possible ‘seat-warmer’ Deputy PM was also viewed negatively. 11 users (1.83%) supporter her in this position, while 19 users (3.16%) were critical. The remaining 572 users (95%) had no opinion and showed little interest in even talking about her, which shows her reputation is not strong with young Malays on Twitter. Pakatan supporters’ praise for her as ‘Malaysia’s first female Deputy PM’ failed to gain any traction.

Concerns were also raised on who would step down for Anwar Ibrahim to become the PM. As it was not clear on whether it would be Tun Mahathir or Wan Azizah, critics feared that Malaysia would be led by a ‘husband and wife’ team with Anwar Ibrahim as PM and Wan Azizah as DPM.

On February 24th 2018, senior DAP leader Lim Kit Siang was reported as saying that Tun Mahathir will give the PM post to Anwar Ibrahim if PH wins GE14, while Wan Azizah would ‘retreat’ from politics. Therefore the country would not have her as DPM at the same time Anwar Ibrahim is PM (https://www.malaymail.com/s/1584415/kit-siang-dr-m-man-of-word-will-step-down-as-pm-for-anwar ).

However it is not clear whether this is Lim Kit Siang’s personal opinion or if he is speaking on behalf of Wan Azizah. We could find no statement online by Wan Azizah regarding her retirement plans after GE14, so the concerns about who the future PM and DPM would still remain valid.

The lack of a timeline was also an issue – questions were raised on how long Tun Mahathir would be in power as PM. Would it be until July/August 2018, after Anwar Ibrahim is released, pardoned and wins a by-election? What would happen if Anwar Ibrahim does not receive a pardon from the Agong? The lack of specifics increased distrust of Pakatan and Tun Mahathir.

Pakatan has a reputation for being unstable, and introducing uncertainty on who the future PM and DPM will be does not improve on their reputation.

In summary:

  • The plan to have an interim PM was too much like Kajang Move
  • Tun Mahathir was not considered trustworthy enough to honour the plan
  • Wan Azizah is not a popular choice as the DPM candidate
  • There are concerns about having a husband and wife team in charge
  • The lack of a timeline and interim plan details is an issue
  • Supporters were only interested in making Anwar Ibrahim the PM

 

4.4.2 Support for Other Leaders

107 users (17.7%) wanted newer, younger leaders instead of Tun Mahathir to be the PM candidate for PH. Only 4 users in this group expressed support for Anwar Ibrahim, but it is important to remember that Anwar is in jail so he was not considered a PM candidate for the next GE.

We took note of alternative PM and Deputy PM candidates offered by users and summarised the responses below. Please note that users who supported multiple candidates counted to each category, e.g. ‘Azmin or Mukhriz for PM’ counted towards both Azmin and Mukhriz.

PM Candidate Users (% of group) (% of population)
Azmin Ali 49 45.79 8.14
Muhyiddin Yassin 13 12.15 2.16
Mukhriz Mahathir 5 4.67 0.83
Rafizi Ramli 5 4.67 0.83
Nurul Izzah 5 4.67 0.83

 

Deputy PM Candidate Users (% of group) (% of population)
Mukhriz Mahathir 14 13.08 2.33
Rafizi Ramli 5 4.67 0.83

 

The pairing of Azmin Ali as PM and Mukhriz Mahathir as Deputy PM was the most popularly mentioned among the users. This an indicator that some Malay youth are ready to move on from old leaders, and PH has lost some support by nominating Tun Mahathir and Anwar Ibrahim as their future PM candidates.

 

4.5 Support for Tun Mahathir as PM Candidate by State

The chart below shows the results by state for selected states.

wp_mahathirpm_chart7

 

State Support (%) Neutral (%) Oppose (%)
Selangor 14.36 12.23 73.40
Kuala Lumpur 20.51 10.26 69.23
Johor 17.65 10.29 72.06
Penang 6.45 19.35 74.19
Perak 29.41 11.76 58.82
Kedah 32.26 6.45 61.29
Pahang 6.67 13.33 80.00
Melaka 11.54 19.23 69.23
Negeri Sembilan 17.39 13.04 69.57
Terengganu 19.05 19.05 61.90
Kelantan 16.67 11.11 72.22

 

Support is lowest in Penang and Pahang and highest in Perak and Kedah. In our previous study on Tun Mahathir and the Opposition working together, Penang had the lowest level of support (25%) and Kedah had the highest level of support (54%).

So there is some consistency in support expressed for Tun Mahathir from users in these states between March 2016 and December 2017, though the percentage levels are different. Opposition towards Tun Mahathir as the PM candidate is also the lowest in Perak at 59%.

4.6 Political Party Support for GE14

With GE14 expected to be months away, users were more open about which party they were planning to support. We noted their responses and charted the result below:

wp_mahathirpm_chart8

 

Category Users (%)
Vote PH 156 25.91
Don’t Vote BN 15 2.49
Neutral / No Opinion 141 23.42
Don’t Vote BN or PH 55 9.14
Don’t Vote PH 157 26.08
Vote BN 78 12.96

 

An explanation of each category is as follows:

  • Vote PH
    • Users who expressed support for voting for PH
  • Don’t Vote BN
    • Users who expressed support for voting for anyone but BN
  • Neutral / No Opinion
    • Users who had no opinion or were undecided
  • Don’t Vote BN or PH
    • Users who intended not to vote for BN or PH. This includes users who would vote for PAS or PSM. This also includes users who are intending not to vote or to spoil their vote instead.
  • Don’t Vote PH
    • Users who would not vote for PH in the next GE
  • Vote BN
    • Users who expressed support for voting for BN

Based on these results, 39% of Malay youth reject PH, 28% reject BN and 9% reject both options. This leaves PH in a weak starting position at winning the support of the Malay youth.

There has been some improvement since our March 2016 study when PH had the support of 15% from youth of all races in Malaysia but there is little time to sway opinion.

We estimate that at least 50% of users in the ‘Don’t Vote PH’ category would support BN in GE14. This is because these users are not pro-PAS; they are not anti-BN or anti-Najib; and they are either anti-PH or anti-Opposition. The remainder of these users include potential PAS voters and former PH supporters who felt betrayed by PH. It is was not clear how many former PH supporters would turn to BN, PAS or PSM; or not vote at all.

 

4.7 Political Party Support for GE14 from Strong PH Supporters

From the population sample of 602 users, 18.27% (110 users) were categorised as strong PH supporters based on their stated party membership or tweeting history (e.g. reliably expressing support for PH). The chart below shows their stated voting preference after Tun Mahathir was announced as the PM candidate.

wp_mahathirpm_chart9

 

Category Users (%)
Vote PH 76 69.09
Don’t Vote BN 3 2.73
Neutral / No Opinion 9 8.18
Don’t Vote BN or PH 6 5.45
Don’t Vote PH 15 13.64
Vote BN 1 0.91

 

Based on these findings, PH may lose at most 31% of their core supporters among Malay youth. 14% of their base have decided to turn away from PH and may vote for BN. The hypocrisy of Pakatan Harapan leaders for choosing their long-time enemy as their leader was the main reason that drove people away from supporting PH.

19 users (17.27% of strong PH supporters) indicated that they would still vote PH despite not supporting Tun Mahathir. This can be attributed to anti-Najib sentiment and loyalty to their party.

 

4.8 Political Party Support for GE14 by State

The chart below shows the results for selected states:

wp_mahathirpm_chart10

 

State Vote PH Don’t Vote BN Neutral / No Opinion Don’t Vote BN or PH Don’t Vote PH Vote BN Total Anti-PH
Selangor 28.19 3.19 22.34 9.57 28.19 8.51 46.27
Kuala Lumpur 28.21 1.71 23.93 9.40 21.37 15.38 46.15
Johor 30.88 2.94 23.53 7.35 19.12 16.18 42.65
Penang 19.35 0.00 16.13 6.45 38.71 19.35 64.51
Perak 26.47 5.88 23.53 11.76 14.71 17.65 44.12
Kedah 35.48 0.00 19.35 12.90 9.68 22.58 45.16
Pahang 13.33 0.00 30.00 6.67 43.33 6.67 56.67
Melaka 15.38 3.85 34.62 3.85 26.92 15.38 46.15
Negeri Sembilan 17.39 8.70 30.43 8.70 21.74 13.04 43.48
Terengganu 19.05 0.00 23.81 9.52 42.86 4.76 57.14
Kelantan 22.22 0.00 16.67 22.22 27.78 11.11 61.11

 

The last column (Total Anti-PH) is the sum of ‘Don’t Vote BN or PH’, ‘Don’t Vote PH’ and ‘Vote BN’ categories. This is to show the proportion of users who are not supporting PH. Based on this, the states with the highest anti-PH sentiment are Pahang (56.67%), Terengganu (57.14%), Kelantan (61.11%) and Penang (64.51%).

Kelantan and Terengganu have stronger anti-PH sentiment due to PAS’ influence there. This is more obvious in Kelantan where the proportion of ‘Don’t Vote BN or PH’ is 22.22%.

Support for PH is highest in Selangor (28.19%), Kuala Lumpur (28.21%), Johor (30.88%) and Kedah (35.48%). The high percentage for Johor is a notable change in support for PH.

Support for BN is highest in Perak (17.65%), Penang (19.35%) and Kedah (22.58%).

As we stated in Section 4.2, BN supporters are under-represented in our sample. As such we can expect support for BN from Malay youth to be a little higher in every state.

4.9 Reasons Influencing Political Party Support

The following sections list the common rationale expressed by users when stating their voting preference. This serves as a way of understanding why young Malays would vote or not vote a certain way.

Vote PH

  • Believe BN/UMNO is corrupt and needs to go
  • Believe PM Najib allegedly stole funds from 1MDB and needs to be punished
  • Believe PH is a better option for good governance and less corruption in politics compared to BN
  • Uncertain on whether Pakatan Harapan is capable, but willing to give them a try
  • Don’t support Tun Mahathir but believe BN/UMNO/Najib needs to go
  • Concerned about cost of living and willing to vote Pakatan Harapan in the hope of change
  • Support Tun Mahathir

Don’t Vote BN

Same criticism of BN and PM Najib as the ‘Vote PH’ users, but uncertain if PH is the right alternative to BN.

Don’t Vote PH

  • PH has a track record of being an unstable coalition of parties. There is no guarantee that there won’t be infighting or a breakup if they win the election.
  • PH’ nomination of Tun Mahathir proved that they lack capable leaders, and was a betrayal of their long-term supporters
  • PH promises an uncertain future where the PM and DPM candidates’ term is unknown (refer to comments on the interim PM plan)
  • Consider PH leaders and/or component parties to be incompetent or untrustworthy
  • Dislike of DAP leaders’ behaviour, treatment of coalition partners and alleged mistakes made in governing Penang
  • Dislike DAP’s dominance within the PH coalition
  • Dislike PKR’s Kajang Move and allegations of corruption in Selangor
  • Dislike populist/unrealistic policies of PH
  • Dislike practice of ‘hate politics’ by PH supporters e.g. labelling welfare recipients as ‘dedak eaters’ and cursing people who disagree with their political views
  • Do not support ‘win at all costs’ attitude of PH, given that they are willing to sacrifice their principles by supporting Tun Mahathir
  • Do not trust Tun Mahathir

Don’t Vote BN or PH

Same issues as users in the ‘Don’t Vote PH’ or ‘Don’t Vote BN’, but don’t believe PH or BN are the answer. Voting for PAS or PSM, not voting and spoiling their vote were options considered by this group.

Vote BN

Apart from the issues listed by the ‘Don’t Vote PH’ users, these users also had the following views:

  • Support BN’s image as a coalition of parties working together to ensure the rights of all races are represented in East and West Malaysia. The image of racial harmony is very important. Either the Opposition parties’ identities need to be more multi-racial, or their top leadership needs to be more multi-racial in order to increase their appeal. They also need to work on their representation in East Malaysia.
  • Support the image of peace and stability represented by BN
  • Support BN’s policies (particularly welfare)
  • Concern about DAP leaders gaining power
  • Support PM Najib’s leadership

 

 

4.10 Other Observations

Users tweeting about Tun Mahathir also expressed other opinions on related topics. This listing was summarised based on a manual reading of a sample of 2,678 users in Peninsular Malaysia, inclusive of the sample used for the above analysis.

Unless otherwise stated these findings apply to users of all races, not just Malays.

4.10.1 Main Election Concerns

The following issues were the most frequently mentioned by users of all races:

  • The cost of living
  • The availability of jobs
  • Rising national debt, which fed into fears of increased cost of living
  • Perception of corruption as ‘common practice’ by the BN-led Federal government

There was frequent mention of paying higher prices for goods now during ‘zaman Najib’ (the age of Najib).

4.10.2 Views of Chinese and Indian users

The majority of Chinese users tweeting about politics during this period avoided offering any opinion on Tun Mahathir. Instead they preferred to focus their attention on repeating anti-Najib and anti-BN remarks. Some of these users ignored the event of Tun Mahathir’s nomination. At most they retweeted links to media coverage of the announcement. This was also true for DAP and PKR party members.

In conversations we observed between pro-PH Chinese users and other users, pro-PH Chinese users adopted the position of ‘another term under Najib is worse’ without addressing concerns (about Tun Mahathir) raised by other users.

When engaged by other users, pro-PH Chinese users were not able to explain why Tun Mahathir was needed, other than that Tun Mahathir is needed to ‘reach the rural Malays’. Our own study has shown that this talking point did not gain traction among young Malays, and few Chinese users made use of it.

As most Chinese preferred to avoid talking about Tun Mahathir, evaluating their opinions using our method would be misleading. What is clear is that young Chinese users who are pro-PH cannot be relied on to promote Tun Mahathir.

Indian users did not exhibit the same avoidance behaviour when it came to the topic of Tun Mahathir. However we did not evaluate many Indian users before revising our study to focus only on Malays. There were also too few users from other ethnicities for us to notice a pattern in expressed opinions.

4.10.3 Criticism of Negative Tun Mahathir coverage

Many users complained about the amount of coverage of Tun Mahathir’s statements and the negative tone used. It prompted users (even those who are anti-Mahathir) to express pity for Tun Mahathir for ‘being bullied’ by mainstream media.

4.10.4 Criticism of Personal Attacks Against PM Najib

Many Malay users expressed negative remarks aimed at PM Najib’s critics who in their view went too far during his visit to Saudi Arabia. Some users had made fun of PM Najib performing the Umrah in Mecca during an official working visit from January 8th – 12th 2018 (https://www.thestar.com.my/news/nation/2018/01/10/najib-visiting-saudi-arabia-on-invitation-of-king-salman/) (https://www.nst.com.my/news/nation/2018/01/324040/najib-shares-image-him-visiting-kaabah ). These users made personal remarks about PM Najib similar to, ‘he has sins to pay for’ or ‘his prayers would not be accepted’.

Personal attacks aimed at Najib, especially towards his religious beliefs, were considered unacceptable by Malay users. This made the pro-PH supporters and anti-Najib movement look bad.

4.10.5 Issues with DAP

The majority of Twitter users who did not like DAP were Malay. The common complaints were:

Unlike previous years where DAP’s negative reputation focused on its position on Hudud, Malay youth on Twitter are now citing examples that deal with governance. DAP’s track record has become an issue of concern for young Malays across Peninsular Malaysia, and not just Penang.

4.10.6 Issues with PKR

Common complaints about PKR were:

4.10.7 Issues with PH Supporters

PH supporters and party members seemed unable to explain why their coalition chose Tun Mahathir as their PM candidate. Some of their common responses to criticism can be paraphrased as follows:

  • Let’s leave the past (Tun Mahathir’s history) behind and focus on the present
  • PM Najib is worse than Tun Mahathir
  • PM Najib is continuing Tun Mahathir’s legacy and we should vote to end it
  • Let’s pick the lesser evil, which is Tun Mahathir
  • Let us win first then we can decide on policies or issues with our proposals later
  • Give PH a chance
  • Who else can lead us? (no supporter explained why other choices such as Azmin or Muhyiddin were not acceptable)
  • We need Tun Mahathir to reach rural Malays

This indicates that PH party members were not given talking points to explain their support for Tun Mahathir. Top party leaders need to address this issue so their members can work better to win over voters.

Ethnic Chinese PH party members tended to keep quiet and avoid talking about Tun Mahathir. PKR members seemed divided on the issue – some protested; some kept quiet; and few supported the decision. There were more PKR members protesting the decision compared to those openly supporting it.

4.10.8 Freedom of Speech

Many users highlighted the fact that freedom of speech is much better now under PM Najib compared to Tun Mahathir’s time as PM. This was in response to allegations by PH supporters and Tun Mahathir supporters that PM Najib is authoritarian and was leading a repressive regime under BN.

 

5. Facebook Trends for Interest in Tun Mahathir and PM Najib

5.1 Overview of Malaysia’s Facebook Population

As of May 2018, there are 24.7 million Facebook users in Malaysia. 54.17% are men and 45.83% are women. The chart below shows how users are distributed by age group.

wp_mahathirpm_chart11

From this total, 20 million (83.33%) users are aged 21 years and above. 55.00% are men and 45.00% are women. These are potential voters residing in Malaysia.

The chart below shows the distribution by age group for Malay speakers, who make up 62.34% of the Facebook user population.

wp_mahathirpm_chart12

The chart below shows the distribution by age group for Chinese speakers, who make up 18.14% of the Facebook user population.

wp_mahathirpm_chart13

95% of potential voters on Facebook use a combination of English, Malay or Chinese languages. There is significant overlap between users of these 3 languages.

From a previous study we have done on language groups in Malaysia, we have identified the top 4 language groups in Malaysia to focus on. These are summarised in the table below:

Language Group Code Users % of Population (>=21 years) Description
Bilingual Malay + English BME 10 million 50.00 Users who speak Malay and English. May also speak other languages except Chinese.
Bilingual Chinese + English BCE 3 million 15.00 Users who speak both Chinese and English. May also speak other languages except Malay.
English Only / English + Other languages EO 3 million 15.00 Users who speak English but do not speak Malay or Chinese. May also speak other languages.
Malay Only / Malay + Other languages MO 2 million 10.00 Users who speak Malay but do not speak English or Chinese. May also speak other languages.

(‘other languages’ refers to languages other than English, Malay or Chinese)

Other language groups that contribute to the remaining 10% of the population are Trilingual Malay+English+Chinese speakers (TRI); Chinese-Only speakers (CO); Bilingual Malay+Chinese (BMC) speakers and speakers of other languages such as Hindi or Tamil. These groups are small and hard to measure and isolate by interest.

These statistics cover 90% of potential voters in Malaysia, but do include foreigners as well. To get a better sense of the local Malaysian division by language, we can look at the language groups for potential voters who are interested in political parties (both Opposition and BN).

Out of 20 million potential voters on Facebook, 7.8 million are interested in political parties. 56.41% are men and 43.59% are women. 98.72% use a combination of English, Malay or Chinese languages.

These can be divided into the following language groups, covering 92.31% of the 7.8 million users:

Language Group Code Users % of Population Interested in Political Parties
Bilingual Malay + English BME 5.4 million 69.23
Bilingual Chinese + English BCE 1.0 million 12.82
English Only / English + Other languages EO 300 thousand 3.85
Malay Only / Malay + Other languages MO 500 thousand 6.41

 

It is safe to assume that a majority of ethnic Malays can be found in the BME and MO groups, while the majority of ethnic Chinese would be found in the BCE, EO and BME groups. To study trends in Malay and Chinese interest we can look at BME+MO statistics combined for Malay users, and BCE statistics for Chinese users. It won’t be as accurate for the Malay users because many non-Malays will be in the BME group, but that is what we have to work with at the moment.

To reach the Malay majority in Malaysia, political coalitions and party leaders should work to maintain a high level of interest from the BME+MO groups.

 

5.2 Facebook Demographics of Tun Mahathir and PM Najib

5.2.1 Total Interest on Facebook

We collected statistics on the audience size interested in PM Najib and Tun Mahathir in July 2017, December 2017, February 2018, March 2018, April 2018 and May 2018. These statistics only cover active Facebook users in Malaysia aged >=21 years (i.e. potential voters). The number of potential voters interested in Tun Mahathir / PM Najib is referred to as the ‘audience’ for that politician.

As of May 2nd 2018, there are 2.7 million potential voters interested in Tun Mahathir and 5.6 million potential voters interested in PM Najib. The trend in interest is shown in the chart below.

wp_mahathirpm_chart14

Interest in Tun Mahathir dropped from 2.9 million users in July 2017 to 1.6 million users by December 2017. Whereas in the same period, interest in PM Najib rose from 3.4 million users to 4 million users. This was followed by a drop to 3.7 million by February 2018 before rising to 4.4 million by April 2018.

Interest in both Tun Mahathir and PM Najib increased dramatically once the campaign period started, with PM Najib reaching 5.6 million users and Tun Mahathir reaching 2.7 million users.

5.2.2 Interest by Age Group

The charts below show the distribution of users by age group for users interested in PM Najib and Tun Mahathir. The combined total of PM Najib and Tun Mahathir’s audience (aged 13+) is 7.12 million users.

wp_mahathirpm_chart15

 

Age Group PM Najib Tun Mahathir
13 – 20 820,000 310,000
21 – 30 2,600,000 1,100,000
31 – 40 1,600,000 740,000
41 – 50 740,000 420,000
51 – 65+ 630,000 470,000

 

The chart below shows a comparison of the total number of users interested in both politicians by age group.

wp_mahathirpm_chart16

There are more users interested in PM Najib compared to Tun Mahathir in each age group. Tun Mahathir has a stronger concentration of users within the 41 – 65+ year old range compared to PM Najib.

From this we can say that Tun Mahathir has a stronger appeal to the middle-aged and older generation of Facebook users. The gap between PM Najib and Tun Mahathir is also smaller for age groups 41-50 and 51 and above.

5.2.3 Interest by Age Group, Malay Speakers

The charts below show the distribution of users by age group for Malay speakers interested in PM Najib and Tun Mahathir. The combined total of PM Najib and Tun Mahathir’s audience of Malay speakers (aged 13+) is 5.87 million users.

wp_mahathirpm_chart17

wp_mahathirpm_chart18

 

Age Group PM Najib Tun Mahathir
13-20 670,000 240,000
21-30 2,200,000 890,000
31-40 1,300,000 610,000
41-50 620,000 320,000
51 – 65+ 480,000 300,000

The chart below shows a comparison of the total number of Malay speakers interested in both politicians by age group.

wp_mahathirpm_chart19

There are more users interested in PM Najib compared to Tun Mahathir in each age group. Tun Mahathir has a stronger concentration of users within the 51 – 65+ year old range compared to PM Najib. The gap between PM Najib and Tun Mahathir is smaller for users aged 51 and above. When it comes to Malay speakers on Facebook, Tun Mahathir has a stronger appeal to the older generation and a very weak appeal for the youngest generation (13-20 year olds).

5.2.4 Interest by Age Group, Chinese Speakers

The charts below show the distribution of users by age group for Chinese speakers interested in PM Najib and Tun Mahathir. The combined total of PM Najib and Tun Mahathir’s audience of Chinese speakers (aged 13+) is 1.03 million users.

wp_mahathirpm_chart20

 

Age Group PM Najib Tun Mahathir
13-20 86,000 41,000
21-30 260,000 140,000
31-40 190,000 120,000
41-50 130,000 110,000
>50 160,000 170,000

The chart below shows a comparison of the total number of Chinese speakers interested in both politicians by age group.

wp_mahathirpm_chart21

There are more users interested in PM Najib compared to Tun Mahathir in each age group except users aged 51 and above, where Tun Mahathir leads PM Najib by 10 thousand users.

Tun Mahathir has a stronger concentration of users within the 41-50 year old range and 51 – 65+ year old range compared to PM Najib. The gap between PM Najib and Tun Mahathir is also smaller for users aged 31-40 years and 41-50 years. Tun Mahathir performs better with Chinese speakers on Facebook compared to Malay speakers.

5.2.5 Interest by Language Group

The chart below shows the breakdown in audience size by language group for Tun Mahathir from July 2017 – April 2018 for the 2 largest groups – BME+MO and BCE. Both groups combined make up a minimum of 86.94% of Tun Mahathir’s audience each month.

wp_mahathirpm_chart22

The table below lists the number of users in each language group interested in Tun Mahathir.

Language Group Jul-17 Dec-17 Feb-18 Mar-18 Apr-18 May-18
Bilingual Malay+English / Malay Only 2,470,000 1,290,000 1,210,000 1,280,000 1,300,000 1,952,500
Bilingual Chinese+English 115,000 105,000 190,000 210,000 250,000 447,500

 

The percentage of Tun Mahathir’s audience that are BME+MO speakers has been declining, despite the increase in the number of users from February 2018 – May 2018. From July 2017 – Dec 2017, interest from BME+MO speakers fell by 1.18 million users, while interest from BCE speakers fell by 10 thousand users. Interest from BME+MO speakers subsequently increased by 10 thousand users by April 2018, while interest from BCE speakers increased by 145 thousand users.

Between April 2018 and May 2018, the audience size for BME+MO speakers grew by 50% while the audience size for BCE speakers grew by 79%.

This means that Tun Mahathir has been drawing more interest from users that are not in the BME+MO group, namely groups that speak Chinese (BCE, BMC, TRI), English-Only (EO) users and users who speak languages other than Malay, English and Chinese.

The following chart shows the breakdown in audience size by language group for PM Najib for the same period. Both BME+MO and BCE combined make up a minimum of 87.50% of PM Najib’s audience each month.

wp_mahathirpm_chart23

The table below lists the number of users in each language group interested in PM Najib.

Language Group Jul-17 Dec-17 Feb-18 Mar-18 Apr-18 May-18
Bilingual Malay+English / Malay Only 2,750,000 3,230,000 3,490,000 3,090,000 3,500,000 4,380,000
Bilingual Chinese+English 225,000 285,000 405,000 355,000 500,000 510,000

 

Following an increase in audience size by 1 million users from July 2017 – April 2018, PM Najib’s audience of BME+MO speakers only dropped by 0.2 percentage points. He maintained a fairly consistent base of 80.68% – 81.62% for BME+MO speakers.

Once the campaign started in April, the percentage of BME+MO speakers dropped by 2.47 points. This was due to increased interest from users in other language groups, particularly English speakers and Chinese speakers not in the BCE group.

5.3 Trends in Tun Mahathir and PM Najib’s Audience Growth by Age Group

There is an overlap between users who are interested in Tun Mahathir and users who are interested in PM Najib. We can measure this overlap and track it over time to see how interest has shifted by age group and language group.

Interest in Tun Mahathir and PM Najib decreased from users aged 21-30 years old between December 2017 and February 2018.

This can be attributed to Tun Mahathir’s nomination as the PM candidate, causing him to lose 60 thousand users in the 21-30 year old range. Tun Mahathir’s audience aged 41 years and above grew by 90 thousand users, helping to offset the loss of younger users.

wp_mahathirpm_chart24

Statistics for Tun Mahathir:

Age Group December 2017 February 2018 Change (%)
13-20 130,000 130,000 0.00
21-30 650,000 590,000 -9.23
31-40 470,000 480,000 2.13
41-50 240,000 270,000 12.50
>50 240,000 300,000 25.00
Total 1,730,000 1,770,000 2.31

 

Statistics for PM Najib:

Age Group December 2017 February 2018 Change (%)
13-20 490,000 430,000 -12.24
21-30 1,800,000 1,700,000 -5.56
31-40 1,200,000 1,100,000 -8.33
41-50 530,000 500,000 -5.66
>50 420,000 420,000 0.00
Total 4,440,000 4,150,000 -6.53

 

Tun Mahathir’s audience grew by 40 thousand while PM Najib’s audience reduced by 290 thousand.

 

Interest in both politicians started to increase from all age groups between February 2018 and April 2018.

This can be attributed to campaign preparations from both sides. There was a significant increase in interest from PM Najib possibly from college/university students below the voting age. Tun Mahathir’s audience aged 41 years and above continued to grow. Interest from 21-30 year old users increased by 80 thousand users, recovering from the loss during the December 2017 – February 2018 period.

wp_mahathirpm_chart25

Statistics for Tun Mahathir:

Age Group February 2018 April 2018 Change (%)
13-20 130,000 150,000 15.38
21-30 590,000 670,000 13.56
31-40 480,000 530,000 10.42
41-50 270,000 310,000 14.81
>50 300,000 360,000 20.00
Total 1,770,000 2,020,000 14.12

 

Statistics for PM Najib:

Age Group February 2018 April 2018 Change (%)
13-20 430,000 630,000 46.51
21-30 1,700,000 2,000,000 17.65
31-40 1,100,000 1,300,000 18.18
41-50 500,000 590,000 18.00
>50 420,000 490,000 16.67
Total 4,150,000 5,010,000 20.72

 

Tun Mahathir’s audience grew by 250 thousand while PM Najib’s audience grew by 860 thousand.

Tun Mahathir’s audience from ages 13 – 40 increased at a higher growth rate compared to PM Najib from April 2018 to May 2018, but still gained less users compared to PM Najib.

This growth rate can be attributed to pre and post-nomination day campaigning as the GE14 campaign period started at the end of April. However Tun Mahathir still lags far behind PM Najib in terms of their total audience sizes – 6.39 million for PM Najib and 3.04 million for Tun Mahathir, inclusive of 13-20 year olds.

wp_mahathirpm_chart26

Statistics for Tun Mahathir:

Age Group April 2018 May 2018 Change (%)
13-20 150,000 310,000 106.67
21-30 670,000 1,100,000 64.18
31-40 530,000 740,000 39.62
41-50 310,000 420,000 35.48
>50 360,000 470,000 30.56
Total 2,020,000 3,040,000 50.50

 

Statistics for PM Najib:

Age Group April 2018 May 2018 Change (%)
13-20 630,000 820,000 30.16
21-30 2,000,000 2,600,000 30.00
31-40 1,300,000 1,600,000 23.08
41-50 590,000 740,000 25.42
>50 490,000 630,000 28.57
Total 5,010,000 6,390,000 27.54

 

Tun Mahathir’s audience grew by 1.02 million while PM Najib’s audience grew by 1.38 million. 

Tun Mahathir gained more Chinese speakers compared to PM Najib from April 2018 to May 2018.

wp_mahathirpm_chart27

Statistics for Tun Mahathir:

Age Group April 2018 May 2018 Change (%)
13-20 32,000 41,000 28.13
21-30 110,000 140,000 27.27
31-40 99,000 120,000 21.21
41-50 91,000 110,000 20.88
>50 150,000 170,000 13.33
Total 482,000 581,000 20.54

 

Statistics for PM Najib:

Age Group April 2018 May 2018 Change (%)
13-20 70,000 86,000 22.86
21-30 220,000 260,000 18.18
31-40 180,000 190,000 5.56
41-50 120,000 130,000 8.33
>50 150,000 160,000 6.67
Total 740,000 826,000 11.62

 

Tun Mahathir’s audience of Chinese speakers grew by 99 thousand while PM Najib’s audience of Chinese speakers grew by 86 thousand.

5.4 Measuring Tun Mahathir and PM Najib’s Reach by Age Group

Based on the data collected, we can measure the divide in interest between Tun Mahathir and PM Najib by age group. This will tell us how much of their combined audience is reachable by either politician. As the politician with the smaller audience, Tun Mahathir’s goal would be to continue expanding his reach particularly among Malay speakers.

5.4.1 Reach by Age Group (Total Audience)

The charts below show the divide in interest between the combined audiences of PM Najib and Tun Mahathir by age group for February 2018 and May 2018. This shows the proportion of users in each age group that is reachable by PM Najib and Tun Mahathir. The last 2 columns on the right show how to calculate the reach for each politician by age group, which is the sum of their exclusive reach (e.g. column ‘Mahathir Only’) and their overlapping audience (the column labelled ‘Both’).

wp_mahathirpm_chart28

 

Age Group Mahathir Only (%) Both (%) Najib Only (%) Mahathir Max Reach, % = Mahathir Only + Both Najib Max Reach, % = Najib Only + Both
>50 20.75 35.85 43.40 56.60 79.25
41-50 19.35 24.19 56.45 43.54 80.64
31-40 15.38 21.54 63.08 36.92 84.62
21-30 10.53 20.53 68.95 31.06 89.48
13-20 10.42 16.67 72.92 27.09 89.59

wp_mahathirpm_chart29

 

Age Group Mahathir Only (%) Both (%) Najib Only (%) Mahathir Max Reach, % = Mahathir Only + Both Najib Max Reach, % = Najib Only + Both
>50 17.11 44.74 38.16 61.85 82.90
41-50 14.94 33.33 51.72 48.27 85.05
31-40 11.11 30.00 58.89 41.11 88.89
21-30 7.14 32.14 60.71 39.28 92.85
13-20 7.87 26.97 65.17 34.84 92.14

 

Between February and May 2018, the combined audience of users interested in PM Najib and Tun Mahathir grew from 4.83 million users to 7.12 million users. That is a net growth of 2.29 million users.

In February 2018, Tun Mahathir was able to reach 56.60% of his combined audience with PM Najib. In May 2018, his reach has increased to 61.8%.

For other age groups however, Tun Mahathir is still not able to reach 50% of his combined audience with PM Najib. For users aged 21-30 years he is only able to reach 39.28% of the combined audience. That means the remaining 60.72% of the audience of young potential voters is not liking, commenting, posting or sharing content related to Tun Mahathir on Facebook.

This is a bad sign for Tun Mahathir when it comes to getting the support of the youth and voters aged 31-40 years as well.

5.4.2 Reach by Age Group (Malay Speakers)

The charts below show the divide in interest between the combined audiences of PM Najib and Tun Mahathir by age group for May 2018.

wp_mahathirpm_chart30

 

Age Group Mahathir Only (%) Both (%) Najib Only (%) Mahathir Max Reach, % = Mahathir Only + Both Najib Max Reach, % = Najib Only + Both
>50 12.73 41.82 45.45 54.55 87.27
41-50 11.43 34.29 54.29 45.72 88.58
31-40 13.33 27.33 59.33 40.66 86.66
21-30 8.33 28.75 62.92 37.08 91.67
13-20 6.94 26.39 66.67 33.33 93.06

 

Compared to the Max Reach statistics for the total audience, Tun Mahathir’s Max Reach with Malay speakers is lower for every age category and PM Najib’s Max Reach is higher for every age category. Our conclusion for Malay speakers is that Tun Mahathir is only doing well with users aged 51 years and above.

5.4.3 Reach by Age Group (Chinese Speakers)

The charts below show the divide in interest between the combined audiences of PM Najib and Tun Mahathir by age group for May 2018.

wp_mahathirpm_chart31

 

Age Group Mahathir Only (%) Both (%) Najib Only (%) Mahathir Max Reach, % = Mahathir Only + Both Najib Max Reach, % = Najib Only + Both
>50 27.27 50.00 22.73 77.27 72.73
41-50 23.53 41.18 35.29 64.71 76.47
31-40 20.83 29.17 50.00 50.00 79.17
21-30 16.13 29.03 54.84 45.16 83.87
13-20 13.13 28.28 58.59 41.41 86.87

 

Compared to the Max Reach statistics for users of all languages, Tun Mahathir’s Max Reach with Chinese speakers is higher for every age category, and PM Najib’s Max Reach is lower for every age category.

Tun Mahathir performs much better with Chinese speakers, being able to reach at least 50% of users in age groups 31-40, 41-50 and 51 years and above. This reinforces the point we made earlier, that interest in Tun Mahathir on Facebook is being supported by Chinese speakers.

 

6. Malaysia’s Twitter Demographics Overview

Since 2014 we have been building up a database of profiled Twitter users in Malaysia which we use to filter tweets and perform opinion-based analyses on political topics. This database has now grown to 922,062 users as of December 2017. Our profiling at the moment is location-based only, with estimated racial profiling for selected users. This has proven reliable for determining which state a user resides in, but less reliable for which suburb (taman) they live in. We currently assign users a location where they most likely live, work or study in based on their tweeting history.

We have also been developing a racial profiler to classify users by their ethnic group based on their interests. In addition to that we performed a study on a sample of 108 thousand users to determine Malaysia’s most followed users and the ethnic distribution of their followers.

Following redelineation we have matched 919,547 of our users’ commonly found locations against polling districts, which we can use to ‘tag’ them to their associated State Assembly Seat (kerusi Dewan Undangan Negeri, or DUN) and Federal (Parliament) seat.

Based on our research, Malaysia has between 2 million active Twitter users. The majority of users are between 18 – 30 years old, with a concentration of users in their early 20s. An estimated 70% – 80% are ethnic Malays.

This makes Twitter a valuable resource for studying the opinions of Malay youth on current issues, as well as marketing products and services that target this demographic.

From past research work done in 2013, we have classified the Federal seats in Malaysia based on their level of urban development – Urban; Semi-Urban; or Rural. These categories are described as:

  • Rural = villages (kampungs) / small towns / farmland distributed within the seat. Rural seats tend to be physically large with a low population.
  • Semi-urban = larger towns and/or numerous small towns, may include villages as well
  • Urban = cities where a majority of the seat is covered by some form of urban development

The chart below shows the distribution of 222 Federal seats by urban development category.

wp_mahathirpm_chart32

This chart shows the number of seats won by each party categorised by the level of urban development.

wp_mahathirpm_chart33

Data for the above charts:

Urban Development Category BN PAS PKR DAP Total
Rural 108 11 3 3 125
Semi-Urban 20 6 15 13 54
Urban 5 4 12 22 43
Total 133 21 30 38 222

 

For comparison, this is the distribution of Twitter users in our database in seats by urban development category.

wp_mahathirpm_chart34

By comparing the distribution of users with the number of seats by urban development category, we can see how imbalanced Twitter (and social media in general) is:

  • 22% of the users are based in rural seats, representing 56% of Federal seats
  • 30% of the users are based in semi-urban seats, representing 24% of Federal seats
  • 48% of the users are based in urban seats, representing 20% of Federal seats

The chart below shows how Twitter users are distributed based on seats won by each party in GE13:

wp_mahathirpm_chart35

It is interesting to note that BN has the largest share of users at 34%, and both PKR and DAP combined have 50% of Malaysia’s Twitter users in their 68 seats, representing 30.63% of the total 222 Federal seats.

7. Interest in Tun Mahathir on Twitter

We have been tracking mentions of Tun Mahathir on Twitter since March 2016. By filtering mentions of Tun Mahathir against our database of profiled Twitter users, this allows us to monitor trends in interest by Federal seat.

7.1 Tweets per month

The chart below shows the number of users tweeting or retweeting content that mentions Mahathir per month from March 2016 – May 7th 2018.

wp_mahathirpm_chart36

Based on this we can see that Tun Mahathir was most popular in November 2016, which can be attributed to his attendance of the Bersih 5 rally.

Following his nomination as the PM candidate for PH, interest in Tun Mahathir dropped in February 2018. It started to pick up again before hitting a peak of 32,970 users in April.

For political analysis we normally focus first on the number of users to reduce the impact of passionate supporters who may tweet or retweet multiple times (its best to analyse tweet amounts separately). We also prefer to focus on non-retweets. This allows us to identify periods when a conversation topic prompted more discussion rather than just retweets.

The next chart shows the number of users tweeting about Tun Mahathir, not including retweets of tweets that mention Mahathir.

wp_mahathirpm_chart37

The statistics for both graphs are listed below, along with the share of users in KL/Selangor for the above graph.

Month Users (including retweets) Users (no retweets) Users (no retweets) based in KL / Selangor, %
Mar-16 35,431 5,901 51.40
Apr-16 12,789 2,270 51.15
May-16 13,833 2,235 47.02
Jun-16 19,682 2,225 53.75
Jul-16 18,216 2,215 52.87
Aug-16 14,141 2,277 48.57
Sep-16 22,967 2,194 52.32
Oct-16 31,006 2,334 51.24
Nov-16 47,906 3,300 51.09
Dec-16 15,154 2,014 51.74
Jan-17 25,406 2,123 49.98
Feb-17 7,835 1,441 48.58
Mar-17 11,696 1,513 50.30
Apr-17 6,835 1,135 50.57
May-17 15,240 1,462 52.39
Jun-17 6,092 1,066 54.97
Jul-17 11,402 2,630 53.69
Aug-17 18,696 2,809 50.48
Sep-17 29,616 2,359 50.45
Oct-17 10,427 1,494 52.21
Nov-17 19,954 1,586 52.77
Dec-17 24,598 2,106 55.37
Jan-18 17,187 3,283 54.61
Feb-18 5,391 1,370 54.31
Mar-18 15,276 2,182 53.90
Apr-18 32,970 4,518 52.99
May-18 20,196 3,302 52.36

 

The level of interest from users outside of KL and Selangor has remained above 50% since March 2017. Social media in Malaysia has an inherent bias towards KL and Selangor, so ideally the percentage of users in these 2 states should stay within 44% – 48% depending on the issue that is talked about. This is a bad indicator for Tun Mahathir’s popularity because it means not enough people outside of KL and Selangor are talking about him.

Based on the graph (not including retweets) we can see that Tun Mahathir’s most popular month was March 2016 during the Citizens’ Declaration, November 2016 (Bersih 5), January 2018 (declaration as PM candidate) and April 2018 (the general election).

7.2 Interest by Urban Development Category

The graph below shows the number of users tweeting about Tun Mahathir (not including retweets) in Federal seats based on their classification of urban development from March 2016 – May 7th 2018.

wp_mahathirpm_chart38

The patterns are similar but the levels are different. Predictable there are more users tweeting about Tun Mahathir in urban seats compared to rural seats.

When we look at the average number of tweets per user, this will tell us how popular Tun Mahathir was with users in each seat type. A high average would mean they are more passionate about the current issue associated with Tun Mahathir. The graph below shows this.

wp_mahathirpm_chart39

In March 2016, users in rural seats had higher average tweets/user compared to users in semi-urban seats. For most of this period the average tweets/user in rural seats is close to the average tweets/user in semi-urban seats.

However after Tun Mahathir was confirmed as the PM candidate, the average tweets/user in rural seats dropped. The GE14 campaign period has done nothing to drive up interest from users in these seats. This is precisely the area where Tun Mahathir was expected to make an impact. Based on the number of users, interest has gone up in rural, semi-urban and urban seats. But people are not motivated enough to say much about Tun Mahathir.

These statistics provide a high-level view. To get better context a detailed reading of what is being said would be needed. Given our time constraints we examined tweets by users in rural seats from April 28th – May 7th, but found no clear pattern of support for Tun Mahathir.

7.3 Interest in Seats Won by Each Party in GE13

The graph below shows the average tweets/user for in Federal seats based on which party won the seat in GE13, from March 2016 – May 7th 2018.

wp_mahathirpm_chart40

Twitter users in PAS seats show a low level of interest. This is significant because not all PAS seats are rural – this trend covers Shah Alam, which is one of the most densely populated seats in Selangor. Twitter users in BN seats also showed declining interest in April and May. PH needs to win PAS and BN seats, so the lack of conversation about Tun Mahathir is not a good indicator.

8. Twitter Sentiment Analysis for PM Najib and Tun Mahathir

By using automated sentiment analysis we can label tweets as ‘positive’ or ‘negative’. This approach is at best 70% accurate. By limiting our analysis to tweets and avoiding retweets, we can reduce the impact of mislabelled tweets.

We applied our system to tweets from users in Malaysia that mention PM Najib. The chart below shows the percentage of users tweeting about PM Najib that mentioned PM Najib in a negative tweet, from December 2013 – May 7th 2018.

wp_mahathirpm_chart41

The highest point in the graph is 51.17% in December 2015, which means that 51.17% of users tweeting about PM Najib mentioned him in a negative tweet. It does not mean that the negativity was directed towards him, just associated with his name.

Looking at the historical trend-line (in red) it is clear that negative sentiment started to rise following the 1MDB scandal. While it started to cool down in 2016, since December 2017 the percentages have started to climb from 40.27% of users to 49.30% of users who mentioned PM Najib.

This rising trend is not a good indicator for PM Najib when it comes to evaluating support from the youth represented on Twitter. Due to a time constraint we are not able to present positive sentiment trends for PM Najib at this time.

For Tun Mahathir there has been an increase in negative sentiment as well. The chart below shows the percentage of users tweeting about Tun Mahathir that mentioned Tun Mahathir in a negative tweet, from March 2016 – May 7th 2018.

wp_mahathirpm_chart42

The percentage of negative users increased from 42.92% of users in February 2018 to 50.54% of users in May 2018.

The percentage of positive users increased from 40.66% of users in February 2018 to 34.00% of users in May 2018.

While automated sentiment analysis is not 100% reliable, the impression is that both PM Najib and Tun Mahathir have not been receiving much expressed praise in the run-up to, and during the general election campaign.

 

9. Conclusion

Based on the research work in this report, the following conclusions are what we consider most relevant for the 14th General Election.

The Youth Have Trust Issues with Tun Mahathir

Based on our Twitter studies, we have determined that Tun Mahathir has a trust deficit with young people between 21-30 years old of all races. This was observed in our analyses in March 2016 and December 2017.

He is considered by many users to be untrustworthy and expected to prioritise his own personal agenda. A majority of Malay youth do not support him as the PM candidate for Pakatan Harapan, and do not expect him to honour any deals such as the interim PM plan.

This is an image problem for Tun Mahathir that can be remedied by apologising for specific allegations made against him, and to take responsibility for the allegations without trying to blame others. Youth of all races expect him to acknowledge his track record, good and bad, and not just focus on the good.

The low level of interest in Tun Mahathir from youth in PAS seats, BN seats and Rural seats further reinforces the findings of our Twitter studies. His nomination may have turned the youth away in the seats that PH needs to win.

Young people are ready for a change of government but nominating Tun Mahathir cost PH some support. Having Azmin Ali or Muhyiddin Yassin as the PM candidate, with Tun Mahathir supporting the campaign would have been a better alternative for winning the support from the youth.

Pakatan Harapan’s reputation with the youth has also suffered due to their choice of Tun Mahathir as the PM candidate. PH may have lost 20% – 31% of their strong supporters among the Malay youth.

PH supporters were not able to provide a good explanation for why Tun Mahathir needs to be the PM candidate. Use of phrases such as ‘we need Tun Mahathir to win rural Malays’ is unlikely to win over Malay voters as it sounds manipulative, like PH parties are just using Tun Mahathir instead of respecting him as their leader.

Between May 1st – May 8th 2018, Tun Mahathir’s audience of Malay speakers on Facebook dropped by 10 thousand users. Among users aged 21-30 years, the drop was 20 thousand users. For users in older age groups particularly 51 years and above, the numbers increased.

On the eve of polling day, we would have expected the Malay-speaking youth on Facebook to be more interest in Tun Mahathir, not less. PM Najib on the other hand, gained a total of 300 thousand users, 100 thousand of which were aged 21-30 years old.

 

Statistics for Tun Mahathir:

Age Group May 1st May 8th Change (users) Change (%)
13-20 240,000 160,000 -80,000 -33.33
21-30 890,000 870,000 -20,000 -2.25
31-40 610,000 640,000 30,000 4.92
41-50 320,000 350,000 30,000 9.38
>50 300,000 330,000 30,000 10.00
Total 2,360,000 2,350,000 -10,000 -0.42

 

Statistics for PM Najib:

Age Group May 1st May 8th Change (users) Change (%)
13-20 670,000 700,000 30,000 4.48
21-30 2,200,000 2,300,000 100,000 4.55
31-40 1,300,000 1,400,000 100,000 7.69
41-50 620,000 660,000 40,000 6.45
>50 480,000 510,000 30,000 6.25
Total 5,270,000 5,570,000 300,000 5.69

 

In chart form:

wp_mahathirpm_chart43

One likely exception to our findings of low interest from Malay youth will be for youth in Kedah. Twitter users consistently expressed higher levels of support for Tun Mahathir in March 2016 and December 2017, compared to users in other states. In our December 2017 study, support for PH is highest in Selangor (28.19%), Kuala Lumpur (28.21%), Johor (30.88%) and Kedah (35.48%).

Based on our December 2017 – January 2018 study, 39% of Malay youth reject PH, 28% reject BN and 9% reject both options. This left PH in a weak starting position at winning the support of the Malay youth.

The Nature of Anti-DAP Sentiment is changing among the Malay Youth

One notable change in anti-DAP sentiment is that Malay youth have moved away from focusing on DAP’s Chinese leadership and Chinese membership base, and focused more on DAP’s track record of governance in Penang.

This was observed on a national level, meaning that DAP’s track record in Penang affects the reputation of PH on national level. However DAP’s use of the PKR logo during GE14 will help negate any potential loss of votes that could be attributed to their track record. Anti-DAP sentiment cannot be expressed at the ballot box if voters don’t know which candidate is a DAP candidate.

No Malay Tsunami, But Likely an Older Generation Shift of Support for PH

From trends observed through Facebook statistics, the ‘Mahathir effect’ has been good at drawing more interest from Chinese speakers of all ages and Malay speakers aged 50 and above. PH may experience a swing of support from the older generation of voters, while sustaining their GE13 levels of support from the Chinese electorate. However there is no indication of a ‘Malay Tsunami’ based on our observations of Twitter and Facebook data.

The following chart shows the distribution of voters by age group based on the 2017 4th Quarter electoral roll. Voters aged 51 and above make up 36.37% of the total electorate of 14.968 million voters. They are the largest segment of voters.

wp_mahathirpm_chart44

Out of 5,443,606 voters aged 51 and above:

  • 2,515,441 voters (46.21%) are Malay
  • 1,958,889 voters (35.99%) are Chinese

Based on GE13 voting patterns, we can isolate and divide voters by ethnic group and likely voting patterns. The charts below show how Malay voters and Chinese voters are divided based on their GE13 voting patterns and other factors such as location, age and race for new voters.

wp_mahathirpm_chart45

It is quite clear that Chinese voters were already very pro-Opposition, which we expect to translate into Pakatan Harapan votes.

Malay voters are a bit more complicated to analyse as many of these older Malays would have voted PAS in previous elections. If Facebook is a reliable indicator for this age group then Tun Mahathir can reach 56.90% of Malay voters who are above 50 years old. Without other indicators it is not possible to say how much of the 66% share that is Leaning BN will swing to Pakatan. Our expectation is that 40% – 50% of Malay voters in this age group will still favour BN.

Based on the 2017 4th Quarter roll, in this coming election there will be 2.39 million new voters. Out of that total, 1,192,488 voters (49.87%) are Malay voters aged 21-30 years. This is the demographic that we have been evaluating through our Twitter analyses. The graph below shows the number of new voters by age and ethnic group, for the 3 largest ethnic groups – Malays, Chinese and Indians.

wp_mahathirpm_chart46

In our December 2017 study we found that Malays in this age group were more likely to be anti-PH than pro-PH. Based on the May 8th Facebook statistics, Malay speakers in this age group were not showing much interest in Tun Mahathir. This does not mean they will vote overwhelmingly for BN – it is likely their votes will be split between BN, PAS and PH with less support for PH. Or they might not vote at all. Voter turnout could be a serious issue for the youth this election.

Advertisements

Written by politweet

May 15, 2018 at 3:34 am

%d bloggers like this: