Delineation and malapportionment
Why seats have wildly different voter counts; rural weightage.
Say you live in Bangi and your cousin lives in Putrajaya, twenty minutes down the highway. At the last general election, GE15 in 2022, Bangi had 303,430 registered voters for a single MP. Putrajaya had 42,881 registered voters for its own, separate MP. Both of you got exactly one Member of Parliament. But your cousin's vote did roughly seven times more work in deciding who represents them than yours did. That's malapportionment — the same "one voter, one vote" principle we normally assume, quietly broken by unequal constituency sizes.
How does this happen in a country with a written constitution? The Constitution actually used to guard against it. At independence in 1957, the rule was that no constituency could have more than 15% more or fewer voters than the average for its state — a small, deliberate allowance for the fact that a rural MP might have to cover a much bigger, harder-to-reach area than an urban one — the Constitution still says in so many words that "a measure of weightage for area ought to be given" to rural seats. But the Constitution is amended by Parliament, and the ruling Alliance government widened that 15% gap to 33% in 1962, then removed the limit entirely in 1973. Since then, there has been no constitutional ceiling on how lopsided a state's constituencies can get. The body that draws the lines, the Election Commission, does not have the final say either: since 1962, its recommended maps have to be approved by a simple majority vote in the Dewan Rakyat.
Two things have made the gap worse recently rather than better. First, boundaries only get redrawn periodically — Peninsular Malaysia's maps were last redrawn in 2018. That exercise ran from September 2016 through two rounds of public display and local inquiries into objections, with the last of them, in Selangor, closing in February 2018; the EC then sent its final report to the Prime Minister on 9 March, the Dewan Rakyat approved it on 28 March by 129 votes to 80, and it was gazetted within about a day of that vote, roughly six weeks before polling. Bersih described the final stage as rushed; Deputy Prime Minister Ahmad Zahid Hamidi told the House the tabling was in line with the Constitution and that the courts had ruled in the EC's favour. Second, in between redraws, the country lowered the voting age to 18 and switched on automatic registration, adding about 5.8 million voters to the rolls by the end of 2021. Those new voters skew younger and more urban — exactly the kind of voter who ends up in a seat like Bangi. So Bangi's electorate grew nearly 70% between 2018 and 2022 while the map around it stood still.
Why does any of this matter beyond the numbers? Because Malaysia elects Parliament using first-past-the-post — whoever gets the most votes in each constituency wins that seat outright, regardless of the margin (see first-past-the-post-and-its-effects). At the 2013 election, GE13, Barisan Nasional got about 47% of the vote, less than the opposition's 51%, but won 60% of the seats and formed the government (see ge13-2013). Those numbers are not disputed; what causes them is. The academics Steven Oliver and Kai Ostwald, who modelled both the 2013 and 2018 results, put a substantial share of that gap down to malapportionment — when seats are unevenly sized and one side's support clusters in the smaller, overrepresented seats, that side can win a majority of seats while losing the popular vote. The Election Commission's then chairman, Mohd Hashim Abdullah, argued the opposite in 2018: that election history showed boundary changes were not what decided any candidate's win, pointing to the opposition holding Selangor and Penang under maps his own commission had drawn.
Whether this is a deliberate design to entrench power or a defensible attempt to give sparsely populated areas a fair hearing is genuinely disputed. The Constitution provides for rural area weightage, and EC and government officials have defended the process, timetable and constitutional criteria rather than a party advantage. This page has not identified a government or EC speaker who defends the scale of current disparities, such as the roughly 10:1 Bangi/Igan comparison. Groups like Bersih and academics who've studied the boundary changes in detail (see bersih-rallies-2007-2016) say the scale of the gap, and who benefits from it, goes far beyond what "rural difficulty" can explain. Both accounts are on the record, but they do not supply equivalent arguments about the present ratios; the elector-count disparities themselves are not in dispute.
That is the answer. Everything below is the working: each claim the explanation rests on, labelled with how certain it is and linked to where it came from. You can stop here.
The working
What the sources say
12 claims, 12 of them carrying a linked source. Each one shows how certain it is — that label is written in the knowledge base alongside the claim, not decided here.
How to read the labels
- established
- multiple independent reliable sources, and uncontested.
- contested
- reliable sources disagree, or official and independent accounts differ.
- uncertain
- genuinely unknown at the time of writing.
establishedHow delineation legally works
The Election Commission (EC/SPR) is constitutionally required, under Article 113(2) of the Federal Constitution, to periodically review the division of the country into federal and state constituencies and recommend changes to Parliament. The rules it must follow when drawing boundaries are set out in the Thirteenth Schedule of the Federal Constitution. The EC's recommendations are not self-executing: since 1962 (see below) they must be tabled and approved by a simple majority in the Dewan Rakyat, which can accept or reject them but not amend the maps directly.
established
The Thirteenth Schedule requires the EC to consult the public: where it provisionally proposes changes affecting a constituency, it must gazette the proposal and publish it in a local newspaper, and representations may be made within one month. A local inquiry is required when an objection comes from the state government, an affected local authority, or a body of at least 100 people on the current electoral rolls of the affected constituencies; one resident alone cannot trigger that statutory local inquiry.
establishedThe rural-weightage rule: origin and erosion
As originally enacted at independence in 1957 (based on the Reid Commission's draft), the Constitution capped the allowed deviation in constituency size at 15% above or below the state average — a genuine but modest allowance for rural constituencies to have somewhat fewer voters than urban ones, justified by the difficulty of representatives reaching scattered rural populations. A 1960 EC redelimitation exercise applied this 15% cap.
New MandalaNew Mandala, ibid. — established
established
In 1962, alarmed that the 1960 exercise had reduced the rural advantage, the Alliance government passed a constitutional amendment that (a) widened the permissible deviation from 15% to 33⅓%, and (b) — more consequentially — stripped the EC of final decision-making power over constituency boundaries, transferring that power to Parliament (simple majority), where the EC's recommendations could be accepted, rejected, or sent back, but were no longer binding on their own.
how the source page qualifies this: established; established, echoing academic consensus
New Mandala, ibid. — establishedMalaysiakini, "'Rural weightage' is why BN keeps on winning," malaysiakini.com/letters/79349, undated — established, echoing academic consensus
established
In 1973, a further constitutional amendment removed the deviation cap altogether, deleting the constitutional wording that had limited rural constituencies to "as little as one half" the electors of an urban one. After 1973 there was no constitutional ceiling on how much smaller a rural seat could be relative to an urban seat within the same state. Academic and civil-society sources describe the practical result as constituencies within the same state varying by a factor of four or five to one, a gap that has since grown further.
New Mandala, ibid. — establishedNew Mandala, ibid. — established
establishedCurrent elector-count disparities
Using the electoral roll for GE15 (2022): the Bangi federal constituency (Selangor) had 303,430 registered voters, the largest in the country, while the Putrajaya federal constituency had 42,881 registered voters — a ratio of roughly 7.1 to 1, not the frequently repeated round figure of "300k vs 35k / ~8x." The commonly cited "~8x" claim is in the right order of magnitude but should be stated as ~7x based on the actual GE15 roll. Kai Ostwald's analysis uses the extreme national pair of Bangi (303k) and rural Igan, Sarawak (~28,290 electors), a disparity exceeding 10:1 — the single largest gap among all 222 federal seats, and a different pair from the Bangi/Putrajaya comparison.
how the source page qualifies this: established outlet; established, peer-reviewed-adjacent academic source
established
Bangi was created from parts of Serdang and Hulu Langat in the 2018 delineation, with 178,790 electors at its first election. The implementation of Undi18 and automatic voter registration before GE15 increased Malaysia’s electorate by over 40% between the 2018 and 2022 general elections, with the largest increases concentrated in already-large constituencies.
established
Nationally, Kai Ostwald's analysis of the GE15 roll finds the largest federal constituency (303k voters, i.e. Bangi) had over 10 times the electors of the smallest (~28k), and that the 75th-percentile constituency (~117k voters) was nearly twice the size of the 25th-percentile constituency (~62k voters) — meaning the disparity is systemic across many seats, not just a Bangi/Putrajaya curiosity.
Kai Ostwald, ISEAS Perspective 2024/30, as above — established
contestedDelineation timeline
The most recent delineation exercise for Peninsular Malaysia was completed in 2018. The EC began its review on 15 September 2016. The Thirteenth Schedule's display-and-objection stages ran over the following 18 months: local inquiries into objections were completed for every affected peninsular state during 2017 except Selangor, where court proceedings delayed the process; the EC published its second display for Selangor on 15 January 2018 with objections due by 14 February 2018. The EC submitted its final report to the Prime Minister on 9 March 2018, the Dewan Rakyat approved the new boundaries on 28 March 2018 by 129 votes to 80, and the order was gazetted and brought into force within roughly 24 hours of that vote — roughly six weeks before polling day at GE14 (9 May 2018). The Constitution requires the EC to conduct a review at intervals of not less than eight years. On that timeline, Peninsular Malaysia and the Federal Territories became due for a fresh review in March 2026; Sabah's last exercise was in 2017, making it due in February 2025; Sarawak's last peninsular-style exercise (for federal seats) was in 2015, making it due in August 2023. EC chairman Abdul Ghani Salleh said on 1 January 2024 that the eight-year rule was the reason Sarawak alone was then eligible: "The last review of federal seats previously took place in 2018, just before the 14th general election, while that for Sarawak was done in 2015, and Sabah in 2017," with the peninsula and Sabah becoming eligible only in 2026 and 2025 respectively. The Sarawak exercise is also tied to Sarawak's 2022 decision to expand its state assembly from 82 to 99 seats. Minister in the Prime Minister's Department Azalina Othman Said said in February 2026 that the EC would decide on a peninsular redelineation only after the eight-year period expired on 9 March 2026. In July 2026, Azalina said the EC was in the final stage of preparing recommendations for Sarawak, but the constitutional review had not formally begun: Article 113(9) fixes its commencement at publication in the Gazette of the notice under Thirteenth Schedule section 4.
how the source page qualifies this: established for the dates; the same article's characterisation of the process is an advocacy assessment, treated as contested below; established for the Selangor display and objection dates, attributed to Bersih for its assessment; established; established, direct EC statement
Bersih-affiliated commentaryThe StarThe Edge MalaysiaThe Edge MalaysiaThe StarThe StarFree Malaysia TodayFederal Constitution of MalaysiaSelangor JournalMalaysiakini, "How GE14 is being rigged through redelineation," malaysiakini.com/news/417270, 2018 — established for the dates; the same article's characterisation of the process is an advocacy assessment, treated as contested below
contestedThe FPTP interaction
Malaysia elects its Dewan Rakyat using first-past-the-post in single-member seats (see). The GE13 (2013) vote and seat totals are not in dispute: Barisan Nasional won 47.38% of the popular vote against Pakatan Rakyat's 50.88%, yet BN took 133 of 222 seats (59.9%) to PR's 89. How much of that gap is caused by malapportionment, as opposed to the ordinary distorting effect of first-past-the-post or to where each side's voters happen to live, is an analytical question rather than an arithmetic one. Steven Oliver and Kai Ostwald's study of the 2013 and 2018 elections attributes a substantial share of the partisan bias in the result to malapportionment, arguing that support concentrated in many small overrepresented seats can convert into a parliamentary majority even when the nationwide popular vote is lost. EC chairman Mohd Hashim Abdullah rejected the general proposition in March 2018, saying the history of elections had proven that redelineation was not the determining factor in the victory of any candidate, and pointing to opposition victories in Selangor and Penang under boundaries drawn by the same commission. See ge13-2013.
how the source page qualifies this: established election database, for the vote and seat totals; contested/academic assessment, attributed to Oliver and Ostwald; contested, attributed to Mohd Hashim Abdullah
International Foundation for Electoral SystemsSteven Oliver and Kai OstwaldMalay Mailfirst-past-the-post-and-its-effectsfirst-past-the-post-and-its-effects
establishedUndi18's effect on disparity
The 2019 constitutional amendment lowering the voting age to 18 and introducing automatic voter registration (Undi18/AVR) added roughly 5.8 million voters to the roll by December 2021, ahead of GE15. Because this expansion happened without any accompanying boundary redraw (the last redraw predates it, in 2018), and new voters skew younger and more urban, it widened rather than narrowed the disparity between fast-growing urban/suburban seats like Bangi and static or shrinking rural seats.
how the source page qualifies this: established, author is a Bersih-affiliated civil society figure; established, independent academic corroboration of the same trend
Thomas FannKai Ostwald, ISEAS Perspective 2024/30, as above — established, independent academic corroboration of the same trend
uncertain
As of August 2026, the Election Commission had not announced formal commencement dates for Peninsular Malaysia or Sabah; it was preparing the Sarawak exercise. The eight-year interval permits a review but does not make commencement automatic.
What is still argued over
Where reliable sources disagree
These are set out side by side, with each position attributed to whoever holds it. No winner is picked.
contested
Whether present-day malapportionment reflects deliberate partisan design or a defensible rural-representation principle is contested, and sources divide roughly along institutional lines.
established
The official/process position starts from the constitutional text itself. Section 2(c) of the Thirteenth Schedule directs that elector numbers "ought to be approximately equal except that, having regard to the greater difficulty of reaching electors in the rural districts and the other disadvantages facing rural constituencies, a measure of weightage for area ought to be given to such constituencies" — that is, the Constitution provides for rural area weightage as compensation for a representational burden, not as a party-political allowance.
how the source page qualifies this: established as the text and stated rationale of section 2(c
contested
Named officials have defended the exercise in those terms. Election Commission chairman Mohd Hashim Abdullah said on 13 March 2018 that there was no manipulation of the electoral system or of the boundaries for GE14, that the redelineation was carried out "not in the political interests but in the interests and for the convenience of voters and to facilitate the service of the elected representatives," that allegations of rigging were baseless claims intended to confuse the people, and that court decisions had found the EC to have complied with the constitutional provisions.
how the source page qualifies this: contested, attributed to Mohd Hashim Abdullah
contested
Winding up the Dewan Rakyat debate on 28 March 2018, Deputy Prime Minister Ahmad Zahid Hamidi said the EC had taken into account the objections raised by various parties and dealt with them, that the courts had decided in the EC's favour, and that this showed the EC had carried out its duties according to the principles in the Federal Constitution; on the speed of the tabling he said that "what was done was in line with the Constitution," citing section 9 of the Thirteenth Schedule and the Prime Minister's discretion to table the EC's final report. Prime Minister Najib Razak told the same sitting that the government disagreed with parts of the EC's report but had not interfered with it or changed its findings, because the government respected the EC's power under the Constitution.
how the source page qualifies this: established, for the 129-80 vote; contested, attributed to Ahmad Zahid Hamidi; contested, attributed to Najib Razak and Ahmad Zahid Hamidi
The StarThe Edge MalaysiaMalaysiakinithe tabling of the EC report
established
On the current delay, EC chairman Abdul Ghani Salleh (1 January 2024) attributed it to the eight-year rule rather than to any discretionary choice, and Minister in the Prime Minister's Department Azalina Othman Said said in February 2026 that the peninsular decision could only follow the expiry of that period on 9 March 2026. Former EC chairman Abdul Rashid Abdul Rahman, describing the procedure in the same February 2026 report, said a review once begun must be completed within two years and that a one-month appeal period is opened in which members of the public may write to the EC, each appeal carrying one hundred signatures.
how the source page qualifies this: established, direct EC statement; established, attributed to Azalina Othman Said and Abdul Rashid Abdul Rahman; established, official EC statement
The StarThe Edge Malaysia, 1 January 2024, as cited above — established, direct EC statementSelangor Journal, January 2024, as cited above — established, official EC statement
uncertain
No government or EC speaker identified by this page defends the scale of the present disparities — for example, the roughly 10:1 Bangi/Igan comparison. The official statements above defend the constitutional rural-weightage principle, the procedure and the timetable; they should not be read as a defence of any particular ratio.
contested
The Bersih / Tindak Malaysia / academic-critique position holds that the scale of disparity seen since 1973 — and especially the sharpest cases — has gone far beyond anything a "difficulty of contact" rationale can justify, and reflects deliberate boundary design by the long-dominant BN/UMNO government to advantage its own strongholds. Civil society coalition Bersih has for years listed constituency redelineation as one of its core demands for electoral reform, arguing the 2018 exercise in particular was rushed through and drawn to entrench incumbent advantage. Political scientist Wong Chin Huat (Sunway University) has characterised the scale of some disparities as making a mockery of the rural-weightage justification, contrasting seats like Baling (~107,213 voters) and Alor Setar (~80,272 voters) — both nominally "rural-adjacent" — to argue the principle is applied inconsistently in ways that track political convenience rather than genuine geography. Kai Ostwald and Steven Oliver's academic work goes further, describing a documented historical pattern in which "UMNO systematically reshaped Malaysia's electoral process," including "reducing the size of electoral districts in areas where the coalition enjoyed strong support" — an explicit claim of deliberate partisan engineering, not incidental rural allowance, and finds Malaysia's malapportionment level to be among the highest of any middle- or high-income democracy.
how the source page qualifies this: contested, attributed to Bersih; contested, attributed to Wong; contested/academic assessment, attributed to Ostwald; contested/academic assessment, attributed to the authors
Malay Mail / FMT reporting on Bersih statements on the 2018 redelineationMalaysiakini, "How to electorally marginalise the Malays," malaysiakini.com/news/511794, citing Wong Chin Huat — contested, attributed to WongKai Ostwald, ISEAS Perspective 2024/30, as above — contested/academic assessment, attributed to OstwaldOliver and Ostwald, SSRN paper, as above — contested/academic assessment, attributed to the authors
contested
Neither position is adjudicated by any court or independent body in a way that settles the "why" question; this page presents both, attributed.