THE Election Commission (EC) is to embark on an exercise to redelineate electoral boundaries, a motion on which will likely be tabled in the March parliamentary sitting. Redelineation is done under Article 113(2)(i) of the Federal Constitution, which provides for it to take place at least every eight years, given demographic growth and other developments that may affect voters’ interests.
With concerns over the EC’s lack of credibility and in the absence of electoral reform, The Nut Graph asks Dr Wong Chin Huat what citizens should look out for.
TNG: Why is redelineation of electoral boundaries necessary, and what is involved?
Redelineation was last completed in the Peninsula, Sabah and Labuan in 2003 and in Sarawak in 2005. So the whole of Malaysia is now eligible for another exercise, after a mandatory interval of at least eight years.
The delineation process consists of two parts: apportionment and districting. Apportionment refers to the allocation of constituencies by an administrative unit, while districting means deciding how the boundaries should be drawn to create those constituencies within the administrative unit. From apportionment arises the problem of malapportionment, and from districting, gerrymandering.
For example, Article 46 now stipulates that Selangor has 22 parliamentary constituencies, while Johor has 26. However, in the 13th general election since our independence (GE13), Selangor had 2 million voters and Johor had only 1.6 million. An average parliamentary constituency in Selangor had 93,129 voters, while that of Johor had only 61,743, less than even two-thirds of Selangor. Such a disparity is called interstate malapportionment.
But even within Selangor, the largest constituency – Kapar – now has 127,012 voters, while the smallest – Sabak Bernam – has only 37,318, less than a third of the former. This is a clear instance of intrastate malapportionment.
Also, the current boundaries have serious gerrymandering issues. Since people who live in a common local jurisdiction are affected by its policies, it is only natural for a constituency to consist of electorate from the same city or district. The EC, however, has violated the provision in Section 2(d), Part 1, Thirteenth Schedule of the Federal Constitution, which stresses the need to preserve “local ties”. In Selangor alone, during the 2003 exercise, the EC created five of 22 parliamentary constituencies and two state constituencies that span across three local authorities.
For example, both P111 Kota Raja and one of its state constituencies, N50 Sri Muda, are made to span across, from west to east, the Municipality of Klang, the City of Shah Alam and the Municipality of Subang Jaya. Four other parliamentary constituencies and 11 more state constituencies span across two local authorities.
Redelineation is necessary to correct these instances of malapportionment and gerrymandering. In fact, correcting these anomalies should be its only function.
Amid the EC’s talk of balancing unequal-sized constituencies, what is the potential for gerrymandering to take place?
Talk of capping the size of urban constituencies at 100,000 voters is deceptive. In the 2003 exercise, the EC created a “malapportionment guide”, which classified constituencies into five classes: metropolitan, urban, semi-urban, semi-rural and rural, each with its range of electorate and land mass.
In the Peninsula, a metropolitan parliamentary constituency should have an electorate of between 70,000 and 90,000, and an area of 8-26km2; while a rural parliamentary constituency should have an electorate of between 20,000 and 29,000, and an area of 250km2.
Under these guidelines Baling was, however, created to have 72,387 voters across an area of 1,542.77km2. Now, is Baling a metropolitan or a super-rural constituency? Are the EC officials mathematically challenged, or do they think the rest of Malaysians are? The fact is, despite the numbers they put up, the EC does not have any objective measure of determining the urbanisation level of constituencies.
More seriously, as in the case of gerrymandering, the EC violates the constitution.
Section 2(c), Part 1, Thirteenth Schedule makes clear that “each constituency in a state ought to be approximately equal” in electorate size, and where malapportionment is allowed, it is only allowed in favour of rural constituencies because of “the greater difficulty of reaching electors … and the other disadvantages facing rural constituencies”.
In the 2013 general elections, the 133 Barisan Nasional (BN)-won parliamentary constituencies, many of which are in rural seats, had an average electorate of 46,860, while the 89 won by the Pakatan Rakyat (PR) had as many as 79,052 in average, nearly twice. Not surprisingly, the EC has systematically created smaller pro-BN constituencies and larger pro-opposition constituencies.
It, however, would be wrong to think that all supersized constituencies are created to pack opposition supporters and reduce the worth of their votes. A “beauty” of malapportionment is that it makes gerrymandering much easier.
For example, because of Umno’s low support among urban Malays, an urban constituency may need to be drawn extremely big to engulf enough Umno supporters so that the party may stand a chance to win. A classic example of that is N29 Sri Serdang, which Umno could only win by a thin, 45-vote margin in 2008, despite a total of 37,819 valid votes cast. In the 2003 redelineation exercise, N29 Sri Serdang was created with 36,989 voters, while its neighbour and the only other state constituency within the same P103 Puchong parliamentary constituency, N30 Kinrara, had only 19,304 voters.
In short, elections can be made unrepresentative by creative use of malapportionment and gerrymandering hand-in-hand. Large constituencies are created, sometimes to pack and dilute opposition voters, sometimes to include more BN supporters and increase the coalition’s chances of winning. When these don’t work, sometime voters get “teleported”, too (see below).
Can equalising the number of voters in all seats be done without increasing the number of parliamentary seats? A seat increase would need two-thirds approval in the Dewan Rakyat, which the BN does not have.
Redelineation can and should be separate from seat increase. In the United States, the House of Representatives today has 435 members as it did in 1911, even though the US population has tripled over the century.
Malaysia today has one parliamentarian for every 63,167 voters. By our standards, India’s Parliament should be 11,000-strong, but its size is capped at 552 members. To have a large parliament is not only costly but also bad for democracy. More lawmakers means less time for each to speak and be heard. A larger legislature also makes it harder for lawmakers to unite, and easier for the president or prime minister to divide and rule.
Without a seat increase, redelineation will have to be purely an adjustment of boundaries to correct excessive malapportionment and gerrymandering. On the other hand, a seat increase allows the EC to easily create winnable seats for the BN. Over time, Umno has contested more and more seats, from 40.26% of Parliament’s total in 1974 to 46.89% in 1986, and to 47.92% in 1995 and 53.42% in 2004.
In GE13, Umno and Parti Pesaka Bumiputera Bersatu (PBB), which is the BN’s second-largest party and the only other Muslim-dominated party, together won only 32% of votes. This, however, translated into 46% of seats. An increase of seats will likely allow Umno and PBB to win a simple majority of seats despite capturing a low percentage of votes.
It is, therefore, suicidal for the PR to support an amendment to Article 46 to increase the number of seats. The Article should, indeed, be amended, but instead of increasing the total seats, seats should be taken from over-represented states like Pahang, Perak and Johor and given to under-represented states and territories like Selangor and Kuala Lumpur.
Malaysians like to bemoan our bloated bureaucracy. Few realise that our Parliament is also bloated in the pace of 8-12% growth between every two elections. If we allow Parliament to constantly grow by 12% every 10 years, by about 2089, we would perhaps have 553 parliamentarians, one more than even India’s constitutional maximum!
Are there avenues for Malaysians to object to the new boundaries? What anomalies should people be looking out for in order to make informed objections?
The EC’s proposal of constitutional representation requires only a simple majority in the Dewan Rakyat to be passed. In other words, any party that can form a simple-majority government can have absolute power to redelineate the constituencies.
The constitution, however, provides an avenue of remedy for state governments, local authorities and any group of 100 or more affected voters to raise their objections within 30 days once the new boundaries are first displayed. The EC must hold a local inquiry to listen to the objections. Should the boundaries be altered, it would need to repeat the display-objection-inquiry process again, but it is only mandated to hold two inquiries only.
The grounds of objection can be either intrastate malapportionment if Section 2(c), Part 1, Thirteenth Schedule is violated (on approximate equality in size of each constituency), or malapportionment, where the “local ties” requirement in Section 2(d) is not adhered to.
Other than having constituencies spanning across a few local authorities, gerrymandering can also be found within the same local authority, where areas with unrelated interests are lumped together. One classic example is the N45 Selat Klang – where more than 80% of the electorate live at the edge of Klang township, sandwiched between the Klang River and Persiaran Raja Muda Musa, while the remaining are Chinese Malaysian fisherfolk in Pulau Ketam. The mainland portion and the islands are not even adjacent to each other! Common sense would suggest that Pulau Ketam would be better placed in N46 Pelabuhan Klang, which consists of mostly the maritime and fishery areas.
The third type of gerrymandering is the partitioning of neighbourhoods, such as housing estates and villages, leaving one half to one constituency and the other half to another. In addition to that is the “teleporting” of voters, where some voters are shifted to neighbouring constituencies on the electoral roll, even though their families who registered under the same address stay put.
One such example involves a Mr Ong who lives in 278-D, Kampung Abdullah, in Segamat. Before the 2003 redelineation, Mr Ong and his entire family were voters in Segamat (then code-numbered P125). After the redelineation, Mr Ong found himself staying put in Segamat (now P140) while his wife and siblings were all transferred to the newly created Sekijang (P141). Did the EC draw the constituency boundary in such a way that it cuts through the house, and even Mr and Mrs Ong’s bed, such that you have “one bed, two constituencies”?
Similarly, 39% of the villagers here were also “teleported”.
Citizens can help to stop malapportionment and gerrymandering by filing in their objections once the new boundaries are displayed. Unlike the 2003 and earlier exercises, where the objections mostly came from opposition parties, this time, civil society groups will play a main role in objecting to any flaws in the redelineation exercise.
Tindak Malaysia will launch their training in the middle of February. A Johor-based group called ENGAGE has organised a town hall meeting for constituents of P158 Tebrau and P159 Pasir Gudang to critically assess their constituency boundaries as a practice for the real inquiry should the need arise. Bersih is committed to providing legal assistance to any group of citizens who are unfairly excluded from the redelineation process, such as if their objections are unreasonably thrown out or dismissed in local inquiries. Malaysians really do have a choice in saving our constituency redelineation from gerrymandering, malapportionment and other irregularities.
Dr Wong Chin Huat is a Fellow of the Penang Institute, a Penang government think tank. He is passionate about electoral reform and is a pro bono consultant for the civil society group, ENGAGE, on the upcoming redelineation exercise.
Sunna Sutta says
I believe that prior to GE13 you put forth a number of scenarios, one of which was the best case scenario which saw PR winning by a slim majority of seats. Following GE13, I think you reanalysed the results which explained how BN won a simple majority despite losing the popular vote. According to some analysts, it appears that it is possible for BN to win a simple majority with as low as 35% of the popular vote because of blatant gerrymandering. Most analyses that were carried out showed that BN swept most of the thinly populated rural constituencies and won a number of semi-urban/semi-rural seats with thin majorities while PR ended up “wasting” their vote bank by winning most of the urban seats with thumping majorities. Nevertheless, it would appear that “rural” carries different meanings in West Malaysia and East Malaysia. In West Malaysia, most rural dwellers (especially on the West coast) are never more than 20 km (of paved roads) from the nearest shopping mall while many villages in East Malaysia are only accessible by river.
I wonder whether you could create a model where the highly populated urban constituencies are split into two while thinly populated rural seats remain as they are. I suspect that most urban constituencies can be split such that the urban poor (who are easily “persuaded” by BR1M and election handouts) and the middle class (who are more likely to vote PR) are neatly separated. If that happens, then BN earns a net gain of 1 seat for every split constituency while PR ends up with the same number of seats. I suspect that hidden hands are at work to move the so-called impartial EC to increase the number of parliamentary seats in order to dilute the effects of urbanisation with the associated easy access to social media and socio-economic mobility that increasingly favour PR.
If the above scenario materialises, will BN’s mandate be based on a much lower proportion of the popular vote than in GE13?