2022/104 “A Granular Analysis of the 2022 Johor State Polls: Implications for Malaysia’s Impending General Election” by Francis E. Hutchinson and Kevin Zhang

A recently installed banner featuring Malaysian Prime Minister Ismail Sabri promising the provision of subsidies totalling RM 77.7 billion in 2022, which would be the largest subsidy bill in Malaysian history to control inflation for the rising prices of goods. Photo taken on 23 October 2022 in Pontian, Johor, by Kevin Zhang of ISEAS – Yusof Ishak Institute.

EXECUTIVE SUMMARY

This Perspective looks at data from the 2018 and 2022 state elections in Johor to understand evolving trends in voter behaviour and what they portend for the upcoming Malaysian General Election. Rather than looking at state assembly seats, it analyses data from polling districts (PDs) within these seats for greater granularity.

Johor is a good bellwether state due to its level of urbanisation and ethnic heterogeneity, which map well onto national averages. It is also the first state to hold elections following the lowering of the voting age and introduction of automatic voter registration. The key findings are as follows:

  • The conventional wisdom that voter behaviour differs for parliamentary and state elections is only partially true. There is strong evidence that Barisan Nasional (BN) supporters consistently turn up to vote regardless of the type of election. For Pakatan Harapan (PH), voter turnout is much more variable.
  • While BN obtained a massive number of seats in the 2022 election, there is still worrying news for the coalition. Notwithstanding its formidable campaign machinery, high-profile appearances by Najib Razak, and an influx of younger Malaysians, the coalition did not break any new ground over the past four years.
  • The 2022 state election was disastrous for PH. The number of votes for the coalition plummeted due to disillusionment, pandemic-related fears, or difficulties for outstation voters to return. PH has areas of reasonably solid support in urban areas that enabled it to secure some seats, but, as with BN, it did not break new ground.
  • Compared to 2018, support for all PH component parties dropped. The Democratic Action Party (DAP) remained the most solid, with Parti Amanah Negara (Amanah) a surprising second. Parti Keadilan Rakyat (PKR) performed particularly poorly.
  • Despite netting few state seats, Perikatan Nasional (PN) did surprisingly well. PN ate into Pakatan Harapan’s support base. Of key importance, the Parti Islam Se-Malaysia (PAS) did better out of the partnership than did Parti Pribumi Bersatu Malaysia(Bersatu), with its votes shooting up dramatically from 2018 when it campaigned by itself. Bersatu performed slightly worse compared to 2018 when it was part of PH.
  • Unlike BN and PH which have key support bases, PN pulled in a significant number of votes from many different groups. However, under Malaysia’s first-past-the-post system, vote share alone does not translate into seats. If it campaigns by itself, PN is likely to struggle – but may still be a kingmaker.

ISEAS Perspective 2022/104, 27 October 2022

Download PDF Version

INTRODUCTION

Now that parliament is dissolved, Malaysia is barrelling towards a General Election. While COVID restrictions are being loosened and campaigning may revert to normal, important structural differences mean that the polls and ensuing results may differ significantly from those held in the past.

Since 2008, elections have been contested by two national coalitions, Barisan Nasional (BN) and Pakatan Rakyat/Harapan. While PAS played a regional role in the 2018 election, the upcoming parliamentary election will be the first to be contested by three national coalitions. In addition to multi-cornered fights producing unexpected results, it is likely that the parliamentary majority will be narrower than in the past. The ‘rules of the game’ have also changed, with the voting age lowered from 21 to 18 and with automatic voter registration (AVR) enabling more Malaysians to vote.

In order to explore the impact of these changes, this Perspective will compare and contrast voter preferences from the Johor state elections of 2018 and 2022. It will use an overlooked but highly valuable source of voting behaviour – polling district results. In particular, it will explore changes in support for the three coalitions and their component member parties across the two elections, before delving into support by urbanisation status.[1]

Johor is a good bellwether state for national trends and dynamics. With 3.8 million inhabitants, it is the third-largest state in the country, and its level of urbanisation is roughly at par with the national average. Johor has both large urban constituencies in and around Johor Bahru, as well as very rural areas along the eastern seaboard. The state is also diverse, and its ethnic composition maps well onto national averages.

Johor is also quite important politically, with parties such as UMNO, the Malaysian Chinese Association, and Bersatu having deep connections to the state. The state’s political trajectory has also closely mirrored the federal government’s. Johor is part of a group of ‘heartland’ states that were ruled continuously by BN up until 2018. That year, as with the federal government, PH secured control of the state. Following the Sheraton Move which saw Perikatan Nasional (PN) attain power, Johor was the first state to revert to PN/BN control. 

The 2018 state elections in Johor were held concurrently with the 14th General Election, which saw Pakatan Harapan win at the national and state levels. The 2022 state election was held in March and was the first election to take place following the lowering of the voting age and establishment of AVR.

Following this introduction, the next section will provide a broad overview of voting trends in Johor by analysing patterns in state seats from 2004 to the present. The subsequent section will then draw on polling district data to explore changes and continuities in voter behaviour across the two elections. The fourth and final section will draw out the implications for GE-15.

OVERVIEW OF STATE SEATS

Barisan Nasional has traditionally dominated Johor, enjoying an average of ten percent greater support in the state relative to the national average. This is due to several factors such as: the founding of UMNO in the state; the limited appeal of PAS due to the state’s traditionalist version of Islam; a legacy of large-scale rural development projects; and the dispersion of the Chinese electorate across more seats than in states such as Selangor and Penang.[2]

This translated into extensive control of the state assembly, with BN traditionally netting 90 percent of seats or more. This began to change in 2008 when Pakatan Rakyat, the precursor to Pakatan Harapan, made its first inroads in Johor (Table 2). This was led by the Democratic Action Party (DAP) in urban mixed constituencies. While an important change, this trend lagged notably in comparison to other urbanised states such as Selangor or Penang which fell to the opposition that year.

The trend was accentuated in 2013, when most urban and mixed seats that had been held by smaller BN parties, such as MCA and Gerakan, fell to Pakatan Rakyat. Nonetheless, the number of rural Malay-majority seats in the assembly meant that Barisan Nasional retained control.[3] Indeed, while the smaller BN parties suffered in 2013, UMNO performed largely as it had in the past. Seen from a historical perspective, 2018 was an aberration for BN and specifically for UMNO. From more than 30 seats, Malaysia’s grand old party was only able to secure 17 seats.

Insofar as PH is concerned, 2013 was key for the DAP as it secured 13 seats that year. In 2018, it only secured one additional seat, with the other PH member parties Amanah, Bersatu (then part of PH), and PKR securing nine, eight, and five seats, respectively. In addition to the urban centre around Johor Bahru and Muar which DAP has held since 2013, PH made headway for the first time in large swathes of North and Central Johor, which are much less urbanised (Map 1).

While the period 2008-2018 was one of steady progress for Pakatan Harapan, the 2022 state election was very much a slide backwards (Map 2). From 36 seats in 2018, PH retained a mere 12 four years later. The DAP weathered the onslaught in better shape, holding 10 seats while losing four. Amanah and PKR’s inroads were almost completely reversed, leaving one seat for each of the two parties. The Malaysian United Democratic Alliance (MUDA), which had an electoral understanding with PH in 2022, won one more seat. These seats were, almost in their entirety, urban centres.[4]

For the newly-formed Perikatan Nasional, the 2022 election was underwhelming. PAS equalled its 2018 performance of one seat and Bersatu’s holdings tumbled from eight seats to two.

For BN, the 2022 election seemed to herald a return to its glorious past. The coalition more than doubled its seats from 19 to 40, with UMNO netting no less than 33. Even the MCA and MIC, which had been virtually wiped out in 2018, seemed revitalised.

However, when votes – rather than seats – are analysed, the picture is rather different. Despite the passage of electoral reforms and the increase of Johor’s electorate from 1.8 million to 2.6 million votes, roughly the same number of people voted in the two elections, meaning turnout tumbled from 83 per cent to a mere 54 per cent. The election was affected by the pandemic, with health concerns prompting many to stay away, despite safety protocols.[5] There was also a cap on the number of people who could use the Vaccinated Travel Lane to cross from Singapore to vote.[6] And, despite postal voting being available for Johoreans resident overseas, only 7,800 ballots were requested – due to a lack of awareness as well as the short period for registration.[7]

Despite its impressive number of seats, BN garnered almost exactly the same number of votes in 2022 as it did in 2018. Bearing in mind that 2018 was the coalition’s worst electoral performance and took place amidst the widespread rejection of GST and the Najib Razak administration, it should worry the coalition that it only secured an additional 17,000 votes despite the total vote bank growing by 800,000.

Despite BN’s middling performance in terms of votes, the picture was very different for PH. 2022 was an unmitigated disaster, as the coalition netted less than half the number of votes it did in 2018. In contrast, PN did very well, coming within a hair’s-breadth of matching PH’s number of votes and more than tripling PAS’s total votes in 2018.

POLLING DISTRICT RESULTS

Analyses of Malaysian politics usually focus on parliamentary or state seats which, while determinant for securing power, aggregate voter preferences at a very high level. In practice, votes are collected at the polling district (PD) level. Given that individual votes are kept secret, PDs are the smallest unit of aggregated election data available for analysis. While Johor has 56 state seats, the state has 949 polling districts – giving us an average of 17 PD per seat.[8] Thus, using PD data is akin to magnifying the resolution by a factor of seventeen, allowing for more detailed findings. In addition, PDs can be grouped according to variables of interest, such as degree of urbanisation and ethnicity to explore voter preferences.

Map 3 provides an illustration of how the parliamentary seat Tanjong Piai as well as its state seats of Pekan Nanas and Kukup – all within Johor – are divided into polling districts. Each coloured polygon represents a PD, and as can be seen from the hues of lilac, ethnic composition varies tremendously across PD even within the same state seat.

Overall Support by Coalition and Party[9]

What, then, do the results at the polling district reveal for the three coalitions and parties?

At first glance, BN’s performance in 2018 and 2022 is indistinguishable (Figure 1). In both cases, the median vote share across PDs was under 50 per cent, although it was marginally higher in 2022. The top of the ‘box’ or upper quartile (marking 75 per cent of all PDs) remained unchanged at 60 per cent. There is a slight decline at the top of the upper whisker (which indicates BN’s best performing polling district) in 2022, although this still remained above 90 per cent. This remarkably similar performance across the two years indicates real discipline by BN grassroots networks at mobilising people to vote regardless of the occasion. The result is also impressive given pandemic worries and the greater number of parties contesting in 2022. 

Looking at the result by component party (Figure 2), the consistency in performance is provided by the largest BN party, the United Malays National Organisation (UMNO). This stabilising performance is magnified by the fact that UMNO contested in about two-thirds of seats in both elections. Turning to the other coalition partners, MCA did marginally better in 2022 than it did in 2018, perhaps due to votes that would have gone to its former coalition partner, Gerakan. The Malaysian Indian Congress (MIC) did substantially better in 2022 than in 2018, but the impact was limited given that it only contested in four seats.

While BN’s results denote consistency, PH’s indicate very variable performance (Figure 3). In 2018, the coalition performed spectacularly, with the median PD value just a shade below 50 per cent – marginally higher than BN’s. And the upper quartile (or 75 per cent) of PDs was 70 per cent – 10 per cent than BN’s. In some PDs, as indicated by the upper whisker, the coalition netted nearly 100 percent of votes. However, in contrast to BN which garnered at least 20 per cent of the vote in its worst-performing PDs, PH secured zero votes in a number of polling districts.

2022 was a catastrophe. The median value across the polling districts dropped by more than half – to a mere 20 percent. In other words, Pakatan Harapan obtained 20 per cent or less of the votes in half of the polling districts. What saved PH from a complete rout was its performance in the top quarter of PDs, with vote shares between 45 to 90 per cent. This one quarter of districts is largely concentrated in urban and mixed seats, which then translated into majorities in 13 seats.[10]

All component parties underperformed in 2022 relative to their 2018 showing. Even the DAP, the most consistent performer, saw its median vote share fall from almost 70 per cent to under 50 percent. In 2018, the party managed to net some votes even in the most inhospitable PDs where it contested. However, in 2022, it received close to zero per cent vote share in some districts. This was not confined to DAP, as all PH component members had PDs where they received almost zero per cent vote share.

In terms of median vote share for polling districts, the other parties performed much worse. PKR did not perform particularly well in 2018, receiving a median 30 per cent vote share. However, this collapsed to 10 percent in 2022. Interestingly, despite PKR being the lead party in PH, Amanah performed substantially better in 2018 and 2022. Its median vote share was 50 per cent in 2018, before collapsing to just under 20 per cent in 2022. Indeed, even the newly-minted party, Malaysian United Democratic Alliance (MUDA) performed better than PKR in 2022 – with a similar median value of ten per cent but a higher upper quartile.

Turning to PAS, it performed poorly in 2018, as it had in previous elections in Johor (Figure 4). The Islamist party only contested in 41 of the 56 seats and averaged about 10 per cent of the vote share in the PDs within the seats where it contested.

However, the partnership with Bersatu paid off in 2022. Through pooling candidates, Perikatan Nasional was able to contest all seats and actually had a median support level of 30 per cent, similar to PH. However, unlike PH, its top 25 per cent of PDs had a much lower vote share – meaning that it was not able to secure majorities in a significant number of seats. When the results for PN by party are analysed, PAS actually outperformed Bersatu, with a slightly higher median vote share (Figure 5). The contrast with its performance in 2018 is striking (Figure 4). 


Voter Preferences and Urbanisation Status

The Election Commission provides information as to the ethnic breakdown of each PD, and by looking at aspects such as physical size and population density, it is possible to establish the degree of urbanisation of each PD.[11] The proportions of rural, semi-urban, and urban PDs are as follows: 40.6; 28.6; and 30.9, respectively. It is possible to have rural PDs within an urban seat, and semi-urban PDs in rural seats.[12]

With regard to electoral performance in rural, semi-urban, and urban PDs across the two years for BN, three things stand out (Figure 6). First, as is to be expected, BN polls much better in rural districts and its level of support trends down in line with greater urbanisation. Second, its performance across the different types of PD in the two years is remarkably consistent – with little to no change. Third, despite UNDI-18 and automatic voter registration, there is no perceptible effect on BN’s level of support. 

Conversely, there is a very strong link between support for PH and urbanisation status, which also holds across both years (Figure 7). However, unlike for BN, support levels dropped across the board in 2022. While support dropped from a median of 70 per cent in urban areas in 2018 to a little under 50 per cent in 2022, there was enough majority support in sufficient PDs to secure a number of state seats. However, support dropped in semi-urban areas with median values under 50 percent; in rural areas it halved from 20 to 10 per cent. 

Figure 8: Vote Share by Polling Districts’ urbanisation status for PH in 2018 (left) and 2022 (right)

Turning to PAS in 2018, while support was consistently low across all types of PD, there is slightly more discernible support in rural areas (Figure 8). In 2022 and with the partnership with Bersatu, this relationship is more visible. As with BN, there is greater support for PN in rural areas, followed by semi-urban and urban areas, with median voter share descending from 30 percent in the case of the first down to 20 percent in the case of the third. This increase in support for PN across the three areas is roughly consistent with the vote share that PH lost in 2022 and is particularly marked in rural areas. This points to the unusual shift in votes in these areas to PH due to the role of senior Bersatu leaders in 2018 that was then lost in 2022.

IMPLICATIONS FOR GE-15

This analysis of voter preferences provides important indications of voter behaviour ahead of GE-15.

Given the remarkable consistency in BN’s performance across time and in the different types of PDs, there are grounds to anticipate that the coalition and its grassroots networks will be able to produce a similar performance in the upcoming parliamentary election. What is worrying, though, is that BN does not seem to have moved on from its nadir of 2018, with virtually no new supporters – despite the influx of nearly 800,000 eligible voters.

For PH, the heady atmosphere of 2018 has dissipated and its supporters are markedly less consistent than BN’s. The opposition coalition’s platform of good governance and institutional integrity plays well in urban areas. However, lacking the grassroots networks that BN does, PH is largely reliant on people to organise their own transport to vote. As with BN, the coalition does have solid pockets of support and this enables it to secure a certain number of seats. But, in 2022, turnout for the coalition was decimated. While BN can rest assured that its supporters will turn up to vote regardless of the occasion, PH needs to hope that its supporters treat state and parliamentary elections differently. It really needs to hold onto its urban pockets and work the ground in semi-urban constituencies to drive up support.

Despite flying below the radar, Perikatan Nasional did well in 2022. Its partnership with Bersatu allowed it to dramatically expand its scope and net 24 per cent of the vote. Broadly, while PH was affected by many people choosing not to vote, its performance was also really affected by part of its previous vote swinging to PN. However, unlike PH, PN lacks the unshakeable support in certain types of constituency that would allow it to secure seats.

Barring unforeseen events, BN’s performance will be solid but this ultimately depends on the relative success or failure of its competitors.

ENDNOTES

[1] This is the first of two Perspectives comparing the 2018 and 2022 Johor state elections. The second will examine voting patterns by ethnicity, turnout, and the lowering of the voting age.

[2] Hutchinson (2018) GE-14 in Johor: The Fall of the Fortress? ISEAS-Trends 2018-03, Singapore: ISEAS-Yusof Ishak Institute; Zhang, Choo, and Fong (2021) The Democratic Action Party in Johor: Assailing the Barisan Nasional Fortress, ISEAS-Trends 2017-01, Singapore: ISEAS-Yusof Ishak Institute.

[3] Hutchinson and Zhang (2022) The Johor State Election: A Spiderweb of Splits, Spats, and Surprises. ISEAS Perspective 2022-23.

[4] There are ten districts in Johor (inclusive of Johor Bahru district) and almost all the PH seats outside of Johor Bahru district are in district capitals. Mengkibol, Tangkak, Bentayan, Penggaram, Senai, Jemantah are within the district capitals of the Kluang, Ledang, Muar, Batu Pahat, Kulai and Segamat districts, respectively.

[5] See this link:

[6] See this link:

[7] See this link:

[8] Johor has 56 state seats, but the results from one state seat, N51 Bukit Batu, have been discounted for 2022 due to contested election results. Thus, the results from 2018 refer to 949 polling districts and those from 2022 refer to 933 districts.

[9] The terms used in the Box and Whisker figures are as follows:

[10] Skudai is an example of a DAP urban stronghold, which the party has held since 2008. In 2022, Marina Ibrahim from DAP won with a majority of about 14,000 out of the 45,000 votes cast. In three quarters of the 21 polling districts within Skudai, DAP had an average vote share of at least 60 per cent.

[11] The following parameters are used to determine the urbanisation status of polling districts: Global Human Settlement Layers, World Population Density, projected trends for cities from the Statistics Department of Malaysian, and area verification from the Town and Country Planning Department.

[12] For instance, Kota Iskandar is an urban seat which encompasses parts of Johor Bahru. Nonetheless, there are also rural PDs within Kota Iskandar such as Kampong Pulai, Tiram Duku, and Tanjong Kupang. The rural seat of Paloh also has semi-urban PDs including Bandar Paloh Utara, and Bandar Paloh Selatan.

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).

Download PDF Version