EXECUTIVE SUMMARY
- This Perspective analyses the results from the 2018 and 2022 Johor state elections to gauge the impact of ethnicity, turnout, and the youth vote on voter preferences.
- Voters in heavily Malay-majority areas in Johor were as motivated to vote in 2022 as they were in 2018 – and they largely chose Barisan Nasional (BN). The coalition even registered a small uptick in support in non-Malay majority areas in 2022.
- Pakatan Harapan’s support base is concentrated in ethnically-mixed areas. However, in 2022 the coalition suffered a marked collapse in support across the board, and the encouraging forays into the heavily-Malay majority areas that it made in 2018 were wiped out.
- Like BN, Perikatan Nasional drew a significant amount of support from Malay-majority areas in 2022, and it is possible that voters in these PDs supported PH in 2018.
- In 2022, those areas voting for Barisan Nasional and Perikatan Nasional were characterised by high levels of turnout – indicating a possible topping out of support. Conversely, support for Pakatan Harapan was strongest in areas with the lowest turnout.
- There was little impact from Undi-18, and 18-20 year-old voters did not turn out to vote in any discernibly different proportion than the average. However, 21-29 year-olds showed a slightly greater propensity to turn out to vote than the general public.
- Neither Barisan Nasional nor Pakatan Harapan enjoyed special favour among voters aged 21-29. However, Perikatan Nasional did register a small, but significantly greater amount of support among this age group.
ISEAS Perspective 2022/114, 16 November 2022
INTRODUCTION
Johor is a very good bellwether state, with urbanisation and income levels that are in line with national averages, as well as an ethnic composition that broadly mirrors Malaysia’s. The state’s political trajectory has also followed the country’s, with an uninterrupted Barisan Nasional (BN) reign until 2018, a fall to the Pakatan Harapan that year, and then a reversion to BN/Perikatan Nasional (PN) rule in 2020.
In March 2022, Johor became one of four governments to hold its election separate from the parliamentary election. However, unlike Sarawak, Sabah, and Melaka, Johor was the first state to hold its polls after the enactment of Undi-18, which lowered the voting age from 21 to 18, and the implementation of automatic voter registration. These reforms, along with population growth, saw Johor’s voter base expand from 1.8 million in 2018 to 2.6 million this year.
Drawing on polling district (PD) data, which is the smallest level of aggregated electoral data available, this Perspective compares and contrasts the results from the 2018 and 2022 state elections. This is done with one broad goal, which is to gauge how voter preferences have changed during this period and what this portends for the upcoming general election scheduled for November 19.
There are some 950 PDs within Johor’s 56 state seats, making an average of 17 PDs per seat. Studying results at this level allows for more detailed analysis and also permits PDs of similar characteristics to be grouped to facilitate exploration for patterns in voter behaviour across key attributes such as urbanisation status, ethnicity, or turnout.[1] Finally, we can examine how and whether voter preferences are affected when state elections are held separately from national ones.
This Perspective is comprised of five sections. Following this introduction, the second section will explore changes in voter preferences from 2018 to 2022 by the ethnic composition of polling districts. The third will explore the impact of the voting reforms in terms of turnout and levels of support for the different coalitions. The fourth will examine the impact of the youth vote, and the fifth will draw out the implications of these changes for the upcoming general election.
ETHNICITY AND VOTER PREFERENCE
Malaysia does not have information on how individual citizens vote. Consequently, the most effective way to analyse how voter preferences differ by variables such as urbanisation status and ethnicity is to group PDs into different categories and analyse prevailing patterns.
To explore the role of ethnicity on voter preferences, our study grouped Johor’s nearly 950 polling districts into the following four categories: mixed (<50% Malay voters); majority (51-75% Malay voters); large majority (76-87.5%); and dominant (>87.6%). Their relative importance is set out in Figure One.
Figure 1: Proportion of Polling Districts by Ethnic Category
Following this division, the median voter preference for Barisan Nasional (BN) across the four categories was plotted for 2018 and 2022 (Figure 2).[2] On aggregate, the proportion of Malay voters in a PD is strongly correlated with support for BN. Looking at the 2018 results, BN secured almost two-thirds of the vote share in ‘dominant’ and half in ‘large majority’ PDs. This high degree of support in these areas means that even in multi-cornered fights, BN has an almost insurmountable advantage. However, this advantage dissipates in ‘majority’ and ‘mixed’ PDs, with the vote share dropping to under one half and one quarter, respectively.
Figure 2: BN’s Vote Share by Ethnic Composition (2018 and 2022)
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
Looking across to the 2022 results, the former ruling coalition was able to reproduce an almost identical pattern – a testament to the strength of their grassroots machinery. That said, there are some very minor differences. For example, while the median PD vote share for BN decreased from 65 to 61 per cent in ‘dominant’ PDs, it increased slightly in ‘majority’ and ‘large majority’ areas. And, in mixed constituencies, its median PD vote share went up from 24 to 29 per cent.
The pattern of support for Pakatan Harapan shows the opposite tendency, as median vote share climbs in PDs with a higher percentage of non-Malay voters (Figure 3). In 2018, Pakatan Harapan obtained more than 70 per cent of the vote share in ‘mixed’ PDs, which constitute the largest proportion of all PDs. In addition, it obtained almost half of the vote share in ‘majority’ PDs, inching ahead of BN in this category. While it lost out to BN in ‘large majority’ and ‘dominant’ categories, it still managed to obtain roughly one-third and one-quarter of votes in these respective groups.
Figure 3. Support for Pakatan Harapan by Ethnic Composition (2018 and 2022)
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
However, in contrast to BN, the pattern of support for PH in 2022 was very different. Unlike BN’s consistent performance across all types of seats – PH lost roughly 20 per cent of its votes across all categories – including its bastion of ‘mixed’ areas. While PH did at least secure 50 per cent of votes in ‘mixed’ PDs, the coalition’s 2018 incursions into Malay ‘majority’ PDs suffered the most drastic fall of 26 per cent – from 48 per cent in 2018 to 22 per cent in 2022. Its support in ‘large majority’ and ‘dominant’ areas fell to 13 and 5 per cent respectively.
While support for BN remained static and plummeted for PH over the same period, the newly-formed partnership between the Islamic party, PAS and Parti Pribumi Bersatu Malaysia (Bersatu) in 2022 paid dividends (Figure 4). Four years previously, PAS only obtained about 15 median vote share per cent in ‘dominant’ and ‘large majority’ PDs – barely above the one-eighth threshold needed to retain election deposits. In ‘majority’ and ‘mixed’ areas, median PD vote share was even lower at ten per cent and four per cent respectively.
However, in 2022, PN obtained one third of votes in ‘dominant’ and ‘large majority’ PDs, and more than one quarter in ‘majority’ PDs. PN even secured 13 per cent vote share in ‘mixed’ areas. Consequently, while BN was able to take some votes from PH in mixed PDs in 2022, the real beneficiary of the swing away from PH was PN across all categories of PDs. And, PN also benefited from a swing away from BN in ‘dominant’ areas. Nonetheless, under FPTP electoral rules, should voting preferences remain unchanged, PN will struggle to win seats in the upcoming GE. This is because it does not attain majority support in any type of polling district, which then leaves it reliant on it prevailing in close-run multi-cornered fights.
Figure 4. Support for PAS (2018) and Perikatan Nasional (2022) by Ethnic Composition
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
TURNOUT
The 2018 election was characterised by a high turnout of 82.3 per cent nationally and 82 per cent in Johor. However, this collapsed to 54.9 per cent in the 2020 Johor state election. At one level, this is attributable to the fact that separate state elections traditionally command lower levels of interest than those held in tandem with national polls. Furthermore, the lowering of the voting age and automatic voter registration dramatically increased the state’s voter base. In addition, the 2022 state elections took place under COVID restrictions, which meant that many Johoreans working out of the state and who wanted to return to vote were unable to do so. While postal votes were allowed, there were complaints that the turnaround time was too short. Given these restrictions, it is worth investigating whether they had an observable impact on all types of voters, or only some.
Figure 5. Turnout According to Ethnic Category (2018)
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
Looking at different patterns of turnout across PDs according to their ethnic composition, there were very high and almost identical levels of turnout in 2018 (Figure 5). In 2022, while turnout decreased across all categories, it did so to differing degrees. Turnout for ‘dominant’ areas was the least affected, even though it decreased by 20 per cent to just above 60 per cent. Turnout decreased in line with ethnic homogeneity, with turnout in ‘mixed’ PDs decreasing by almost half, from the low 80s to the high 40s.
The decline of 20 per cent even in ‘dominant’ PDs indicate that COVID-19 concerns and the (relatively) low stakes of a state government election discouraged a substantial number of outstation voters from returning to cast their vote. That said, the 26 per cent gap in turnout between ‘dominant’ and “mixed’ PDs suggests that, apart from COVID-19 considerations, there were other factors which discouraged non-Malays to turn up for polls. Indeed, less than half of eligible voters in ‘mixed’ areas voted. One possibility could be widespread disillusionment towards PH among non-Malays but also an unwillingness to support either of the two alternatives, BN or PN.
Figure 6. Turnout According to Ethnic Category (2022)
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
To further probe this, we compared turnout by PD with the state-wide average of 54.9 per cent for 2022. PDs were grouped into four categories relative to this average: significantly below (<49%); marginally below (49-54.9%); marginally above (55-63%); and significantly above (>63%). The proportion of PDs across the four groups is as follows: 26.0, 22.7, 25.9, and 25.4.[3]
The distribution across the four categories clearly shows higher vote share for BN in areas with greater turnout, beginning at around 30 per cent and ascending to above 60 per cent in the PDs with the most turnout (Figure 7). Pakatan Harapan has the opposite relationship with turnout, registering the highest amount of support in the areas of lowest turnout. Its median PD vote share in areas with higher marginally and significantly higher turnout is dismal – at 14 per cent and 4 per cent respectively.
Figure 7. Support for BN and PH by Relative Turnout (2022)
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
How then does relative turnout in 2022 relate to support for Perikatan Nasional (Figure 8)? Broadly, voter preferences for the coalition mirror those of Barisan Nasional, with higher levels of support in areas that registered high turnout, and vice versa. The only difference is that vote share for PN tracks 15-20 per centage points below BN. While PN secured a lower vote share than PH in areas with low turnout, it did much better in PDs with higher turnout – indicating a more motivated support base than the other coalition.
Figure 8. Support for PN by Relative Turnout (2022)
Note: Values indicate vote share at the median polling district for each category, with outliers removed.
Thus, while the influx of new voters did not boost the number of votes and actually caused turnout to decrease for the state as a whole, both BN and PN obtained higher vote share in those areas with higher turnout. Conversely, PH experienced lower levels of vote share in these areas, only securing a large proportion of votes in PDs with low turnout.
YOUTH VOTE
While the previous section sought to explore the impact of turnout, this section will seek to establish whether there was a discernible impact from younger voters. To this end, the youth vote was divided into two cohorts: 18-20 year-olds and 21-29 year-olds. This division was made to distinguish the potential impact of Undi-18 from the preferences of younger voters in general.
Following this, two series of bivariate regressions were carried out (Annex One). The proportion of 18-20 year-olds and then 21-29 year-olds at the state seat level was run against voter turnout at the polling district level. This was performed to observe whether younger people were more likely to vote than Johoreans as a whole.
From there, the proportion of 18-20 year-olds and 21-29 year-olds at the state seat level was run against the vote share for each coalition at the polling district level. This was carried out to see if younger voters had markedly different preferences than the general population of voters.
Insofar as the Undi-18 vote is concerned, 18-20 year old Johoreans were no more or less likely to vote than the general population. At the same time, Johoreans in the 21-29 age bracket contrastingly showed a slightly greater propensity to vote than the population as a whole.
When matched against the vote share for the three coalitions, neither Barisan Nasional nor Pakatan Harapan benefited from more support from Johoreans in the 18-20 age bracket. While the same holds true for Barisan Nasional in the 21-29 age group, voters in this cohort were slightly less likely to vote for Pakatan Harapan.
Conversely, there is a stronger correlation between the youth vote and preference for Perikatan Nasional. Among 18-20 year-olds, there is a small, positive and almost statistically significant relationship with support for PN. At the same time, among 21-29 year-olds, this relationship, although also small, is positive and statistically significant – indicating that the newest coalition was able to harness slightly more support from young voters than its competitors did.
IMPLICATIONS
Of the three coalitions, Barisan Nasional is the steady performer, with its campaigning credentials honed by many decades in the electoral trenches. However, while virtually guaranteed to replicate the turnout and voting patterns seen in 2018 and 2022, it is not looking invincible. This is for two reasons. First, the types of areas that support BN were characterised by higher levels of turnout, indicating a topping out of support. Second, while consistently mobilising the faithful to vote, it made no new electoral inroads in the past four years – most notably among younger voters.
Nonetheless, there are some positive indicators for this most established coalition. The five per cent increase in support for BN in ‘mixed’ areas in 2022 is not insignificant. This is because under Malaysia’s first-past-the-post electoral (FPTP) rules, a party can win a parliamentary seat while receiving only around 40 per cent of the total votes in a multi-cornered fight. While BN can secure around 30 per cent in ‘mixed’ PDs in the upcoming general election, it needs to garner at least 50 per cent in Malay-majority PDs within the same parliament seat to win. Based on the 2022 state election, this is a plausible scenario. Conversely, should support for BN in Malay ‘majority’ PDs decline by a few percentage points, it risks losing these areas as well as corresponding parliament seats to its rivals.
Pakatan Harapan constitutes the biggest incognito of the three coalitions. Its heady performance in 2018 was followed by an electoral nadir in 2022. Beyond being hit by low levels of turnout in 2022 in its heartland areas, PH was hit by a shift away in support in Malay-majority primarily to PN, but also to a smaller degree to BN. Furthermore, it registered no particular favour among voters 18-20, and garnered lower levels of support among 21-29 year-olds.
In the upcoming GE, PH’s best chances lie in ‘mixed’ areas, and the coalition needs to retain as many of these PDs as possible. Its hope rests on the reservoir of votes it mobilised in 2018 and which were untapped in 2022. Should more voters turn up at the urns, they are more likely to vote for PH than the alternatives.
Perikatan Nasional is the youngest and most untested of the three. It has grown in size and influence, and managed a respectable performance in 2022. Most notably, it engineered a swing away from PH in Malay-majority areas, and more notably also managed to sap some support from BN in its heartland areas. It also seems to be more attractive to young voters than either BN or PH. However, the degree of attraction is not huge and as with all other aspects of PN’s support base – while promising – is yet to be determinant.
Traditionally, Malaysian elections have been between one established coalition, Barisan Nasional, and various opposition groups. Since 2013, this transitioned into a two-coalition model, with Pakatan Rakyat/Harapan now steadily in the fray. However, the next election will see three coalitions contesting, with all having led a national governing coalition in the recent past. This higher level of competition and the greater number of multi-cornered fights significantly increases the potential for dramatic denouements.
Annex One. Exploring the Youth Vote
| Dependent Variable | Independent Variable | Result |
| Voter Turnout (PD level) | Proportion of 18-20 year-olds (state seat) | Not statistically significant: Adj r2 (0.000); T-stat (1.245); P value (0.214) |
| Voter Turnout (PD level) | Proportion of 21-29 year-olds (state seat) | Statistically significant, positive, and very small: Adj r2 (0.006); T-statistic (2.629); P value (0.009) |
| Vote Share for BN (PD level) | Proportion of 18-20 year-olds (state seat) | Not statistically significant: Adj r2 (0.001); T-stat (-1.271); P value (0.204) |
| Vote Share for BN (PD level) | Proportion of 21-29 year-olds (state seat) | Not statistically significant: Adj r2 (0.001); T-stat (-0.491); P value (0.624) |
| Vote Share for PH (PD level) | Proportion of 18-20 year-olds (state seat) | Not statistically significant: Adj r2 (0.000); T-stat (-0.763); P value (0.445) |
| Vote Share for PH (PD level) | Proportion of 21-29 year-olds (state seat) | Statistically significant, negative, and very small: Adj r2 (0.003); T-stat (-2.029); P value (0.043) |
| Vote Share for PN (PD level) | Proportion of 18-20 year-olds (state seat) | Not statistically significant: Adj r2 (0.002); T-stat (-1.805); P value (0.071) |
| Vote Share for PN (PD level) | Proportion of 21-29 year-olds (state seat) | Statistically significant, positive, and very small: Adj r2 (0.014); T-stat (3.833); P value (0.000) |
ENDNOTES
For endnotes, please refer to the the pdf document here.
| ISEAS Perspective is published electronically by: ISEAS – Yusof Ishak Institute 30 Heng Mui Keng Terrace Singapore 119614 Main Tel: (65) 6778 0955 Main Fax: (65) 6778 1735 Get Involved with ISEAS. Please click here: /support/get-involved-with-iseas/ | ISEAS – Yusof Ishak Institute accepts no responsibility for facts presented and views expressed. Responsibility rests exclusively with the individual author or authors. No part of this publication may be reproduced in any form without permission. © Copyright is held by the author or authors of each article. | Editorial Chairman: Choi Shing Kwok Editorial Advisor: Tan Chin Tiong Editorial Committee: Terence Chong, Cassey Lee, Norshahril Saat, and Hoang Thi Ha Managing Editor: Ooi Kee Beng Editors: William Choong, Lee Poh Onn, Lee Sue-Ann, and Ng Kah Meng Comments are welcome and may be sent to the author(s). |