Blog of the International Journal of Constitutional Law

Interethnic Vote Pooling, Institutional Frailty, and the Malaysian Elections of 2018

[Editor’s Note: This is the second entry in our symposium on “Constitutional Implications of the Malaysian Tsunami.” The introduction to the symposium is available here.]

Donald L. Horowitz, Duke University

In the early 1950s, as Malaya was approaching independence, the British decided to conduct the first elections ever held in the country at the town council level, starting with Kuala Lumpur.  Not long before, the United Malays National Organization (UMNO) had had a leadership crisis.  Its founding president, Dato’ Onn bin Ja’afar, had resigned when the party failed to agree to accept non-Malay members and decided to continue as a Malay-only organization.  Onn had gone on to found the multiethnic Independence of Malaya Party (IMP) to contest the local elections.  The IMP looked to be a formidable competitor to UMNO in Malaya’s towns, most of them with non-Malay—especially Chinese—majorities.  The leadership of the Malayan Chinese Association (MCA), founded a few years earlier, was favorable to the IMP, but the MCA leader in Kuala Lumpur, H.S. Lee, decided to join with UMNO’s leader in KL to form the UMNO-MCA Alliance, which, by pooling Malay and Chinese votes, managed to defeat Onn’s IMP and went on to do so in one council election after another all over the country.[1]

The Alliance became a permanent preelectoral coalition that, with the later addition of the Malayan Indian Congress (MIC), went on to sweep all but one of the 52 seats contested at the 1955 national elections that preceded independence.  By pooling the votes of Malays, then barely half the population but at the time a greater fraction of the voting population, with the votes of Chinese and Indians, the coalition overwhelmed the opposition. Even after more non-Malays became citizens, the Alliance also won the first post-independence election in 1959. In each case, it put up a single slate of candidates in single-member constituencies across the country. With the aid of some parties in Sabah and Sarawak after those Borneo states joined with Malaya to form Malaysia in 1963, the Alliance—later called the Barisan Nasional or BN—won the next dozen general elections as well.  In 2018, however, it lost to a competing coalition, the Pakatan Harapan (or Coalition of Hope), which practiced the interethnic vote pooling pioneered by the Alliance.  Embedded in this practice are some questions of constitutional and electoral-system design.

By 2018, many forces had undermined the Barisan’s winning formula of pooling the votes of Malaysia’s various ethnic groups and had especially eroded the support of Chinese and Indians. Even as early as 1959, when many Chinese had just become citizens, the then-prime minister, Tunku Abdul Rahman, effectively sent many new citizens off to Chinese opposition parties by abruptly failing to meet some claims advanced by the MCA.  A decade later, when Chinese opposition parties celebrated certain state-level victories at the polls, they touched off a violent Malay reaction that produced a 21-month suspension of parliament and a new program of preferences for Malays in business, education, and employment.  That program, in turn, drove more non-Malays to the opposition, further weakening the contribution the MCA and MIC could make to the ruling coalition in succeeding elections.

With those restraints lost, policy outcomes became more and more one-sided in the 1980s and ‘90s, further alienating non-Malay voters; about half of Chinese voters supported the opposition in most of these elections.[2]  And as this alienation progressed, a crony-driven politics of patronage became increasingly evident in the UMNO-led government.  Still, with heavily gerrymandered constituencies and with coopted allies in Sabah and Sarawak, UMNO and its partners could command large parliamentary majorities despite the slow-moving consolidation of Chinese opposition in one party, the Democratic Action Party (DAP), on one side, and Malay opposition in the Islamic party (PAS), on the other.

That is how things stood until the Asian financial crisis struck in 1997, creating a major split between the prime minister, Tun Mahathir Mohamad, and his dismissed deputy, Dato’ Seri Anwar Ibrahim, soon to be imprisoned.  The opposition to Mahathir took up the cry of “Reformasi,” also heard in neighboring Indonesia, where President Soeharto succumbed to similar grievances pertaining to political favoritism and management of the economy.  In Malaysia, Mahathir survived, but the reform movement eventually galvanized a new opposition coalition among Anwar’s multiethnic Parti Keadilan Rakyat (People’s Justice Party, or PKR), PAS, and the DAP, supported by sympathetic civil society movements that crossed ethnic lines.

The DAP and PAS soon fell out, however, over Islamic religious issues, and in 1999 and 2003 many Chinese voters returned to vote for the BN.  But over the long haul, the trend toward a dramatic decline in Chinese (and, increasingly, Indian) support for the ruling coalition was unmistakable.  With Chinese disaffected and Malay voters divided among UMNO, PAS, and PKR, it was possible for the BN to falter.  It lost the popular vote in 2013 because of what the regime termed “a Chinese tsunami.” Still, the BN kept its majority of seats by dint of radical gerrymandering to pack Chinese into large constituencies, so they elected disproportionately fewer candidates than their numbers would justify, and Malays into smaller constituencies, so they elected many more.

By 2018, however, several elements combined to augment the opposition. First, failing Chinese support had made it imperative for the UMNO-led BN to shore up its Malay support.  This it did by extreme, even Wahabbist, appeals to Malay religious conformity and corresponding suppression of minority religious activity—among Shia, Ahmadis, and (conspicuously) among Christians, using the heavy hands of Islamic bureaucracy and cooperative courts.[3]  Non-Malays—not just Chinese and Indians but Christians, especially (but not only) in Sabah and Sarawak—felt even more alienated.  Second, several large scandals involving misappropriation of large sums of government funds were uncovered, although wrongdoing was denied by the regime.  Third, a split in PAS led to the creation of Amanah (Parti Amanah Negara, or National Trust Party), a moderate Islamic splinter party that joined the opposition. Fourth, the unresolved scandals led retired prime minister Mahathir to form a new party appealing to Malay voters and to take it, too, into the opposition. Mahathir’s Parti Pribumi Bersatu Malaysia (PPBM, or Malaysian United Sons of the Soil Party) became part of Pakatan Harapan (PH), together with  PKR, DAP, and Amanah.  With Mahathir and Anwar reconciled, and the former prepared to be the interim prime minister after the 2018 elections, it was agreed that, should PH form the government, the new government would immediately seek a pardon for Anwar, who would then be released from prison, and wait in the wings to become prime minister.

There was a big question, however.  Malaysia’s are plurality elections.  PAS, impressed by UMNO’s newfound Islamic piety and willingness to cooperate with its radical Islamic agenda, was prepared to run third-party candidates who could siphon Malay opposition votes away from the Pakatan and allow UMNO or other BN candidates to win seats on much less than a majority of votes.  To defeat any such strategy, the four-party Pakatan coalition would have to coordinate its interethnic vote pooling, so that even if its Malay or non-Malay candidates could not win alone, their ethnic allies could compensate.  They had to do this in the teeth of a gerrymander even more radical than any in the past, one the BN accomplished with the aid of a compliant electoral commission and a mostly accommodative judiciary—about which more shortly.

As interethnic vote pooling was the key to the Alliance’s early success, so was it a very important component of the Pakatan Harapan’s victory in 2018.[4] This time the Pakatan won just short of half the popular vote, but by far the largest share of votes and a solid majority of the 220 seats in the Malaysian parliament:  122 (counting the seats of an ally in Sabah) to the Barisan’s 79 and PAS’s 18, plus three won by minor candidates. (The Pakatan majority increased further after the election with the accession of some non-PH winning candidates.)

It is too soon to do a definitive analysis of vote pooling in 2018, but a few illustrations of its powerful impact can be observed.  In Pulai, in Johor, previously UMNO’s most reliable state, where Malay voters were 49.4 percent, Chinese 39.2, Indians 10.2, and others 1.2, Amanah’s Salahuddin  Ayub defeated UMNO’s Nur Jazlan Mohamad by 63.1 percent to 30.2 percent, with a PAS candidate polling a mere 4.9 percent.  Salahuddin had a strong multiethnic majority. In Nibong Tebal, in Penang state, Malays were 47.9 percent of the electorate, compared to 35 percent Chinese and 16.8 percent Indians.  There the PKR Malay candidate defeated an UMNO Malay and a PAS Malay by 56.1 to 31.1 and 10.9 percent, respectively. Or Batu Pahat, Johor, with 54.8 percent Malays, 43.5 percent Chinese, and 1.3 percent Indians, where the UMNO candidate appears to have won a majority of Malay votes (28,035), and a PAS candidate received 8,173, but the PKR Malay candidate won a clear majority of all votes (54.8 percent) by adding non-Malay support to that of Malays.  In these and other constituencies, the UMNO-led BN paid a heavy price for alienating non-Malay voters.

It was, however, not merely Malay candidates of the Pakatan who were aided by the DAP.   Consider the narrow victory won in Bentong, in Pahang, by a DAP candidate who defeated Dato’ Seri Liow Tiong Lai, the president of the MCA and Minister of Transport in the BN government. The DAP candidate won with a plurality of 46.7 percent to Liow’s 43.0 percent and a PAS candidate’s 10.3 percent. The constituency has 46.6 percent Malay voters to 41.7 percent Chinese and 9.2 percent Indians.  In Pahang, the odds are that most Malay voters are pro-BN, and that would have put the DAP candidate at a disadvantage, which might, at least hypothetically, have been overcome by DAP solidarity and the slight majority held by non-Malay voters in the constituency.  To defeat the MCA president, however, would, in such a heterogeneous constituency, have required as many PKR and Amanah votes as could be mustered.

Two even stronger cases of Malay votes for DAP Chinese candidates were in Bangi, in Selangor, and Seremban, in Negri Sembilan. Both were won by thumping majorities.

Bangi is a constituency in which gerrymandering reversed the ratio of Chinese to Malay voters from 49:39 in 2013 to 39:49 (with 11 percent Indians) in 2018.  Yet the DAP’s incumbent M.P. Ong Kian Ming won a thumping 65.1 percent majority in the latter year, nearly doubling the combined vote totals of a BN and a PAS candidate.

An equally strong case for vote pooling to benefit the DAP candidate is provided by the Seremban seat, where the DAP’s Anthony Loke won a landslide victory, with a huge margin of 30,694 votes out of the 92,870 votes cast in a constituency that is 48.6 percent Malay, 36.8 percent Chinese, and 13.2 percent Indian.  Loke received 9,000 more votes than there were non-Malay voters, not all of whom would have been his supporters in any case.  These surplus votes would have been delivered almost exclusively by Malay voters committed to Pakatan parties.

Since DAP many candidates stood in strongly Chinese constituencies, where, with few exceptions, they easily defeated BN Chinese candidates, it is harder to show the full extent to which they were helped by pro-Pakatan Malay voters, without a much more fine-grained analysis of individual voter behavior.  But what we have so far is a strong indication that interethnic vote pooling was indeed reciprocal among Pakatan partners.[5] It certainly was so in constituencies where leaders of the constituent parties, such as PKR’s Wan Azizah Wan Ismail and DAP’s Lim Kit Siang racked up very strong majorities, with the support of voters loyal to their allies in the coalition.

Were vote-pooling victories decisive in producing the overall Pakatan majority?  Bearing in mind that they are only possible in heterogeneous constituencies, it is impossible to know without a more careful statistical dissection.  Still, it is notable that the majority of the Pakatan’s seats were won in heterogeneous constituencies: 73 of the 83 seats where no ethnic group comprised more than 70 percent of the electorate.[6]  Perhaps the coalition would have won many of those seats without herculean efforts to pool votes. But that issue elides a more fundamental point.  In such constituencies, vote pooling is a rewarding, vote-maximizing strategy, but it has an important prerequisite. The agreement to pool votes across group lines, with the party identified with one group vouching for candidates of another, is not possible without compromise and mutual moderation on divisive ethnic issues.[7]

In accordance with this prerequisite, the Pakatan has made a point of being in the moderate center[8]—on religious issues, where it has not supported Islamist exclusivism, and on ethnic issues, where it has substituted ketuanan rakyat (supremacy of the people) for ketuanan Melayu (Malay supremacy), touted by some UMNO leaders.  That sort of moderation, coupled with the reciprocal value of the votes to be exchanged, is the glue that holds interethnic coalitions together and makes them rewarding at the polls. That value has been demonstrated in Malaysia, just as it was in the 1950s and ‘60s.

In plural societies with a strong need for compromise, majorities tend to want their way, and rewards for moderation are often in short supply, however necessary they may be. Those rewards are easier to come by when groups live in close proximity to each other and there are votes to be pooled. The best way to encourage vote pooling is to make constituencies heterogeneous where the demography is conducive and to apportion them fairly.  That requires in turn a suite of institutions dedicated to fairness and the rule of law, institutional virtues that far too often prove easy to cast aside.

What this means concretely is that independent electoral commissions, operating under lawful mandates and resistant to manipulation, are indispensable.  So, too, are courts that are faithful to those mandates. Creating these institutions is one thing; keeping them when they come under pressure has proven to be another—and not merely in Malaysia.  Malaysia’s opposition tsunami of 2018 was created despite exceptional efforts at electoral manipulation and only because the electorate was unusually aroused.  In Malaysia, and elsewhere as well, methods need to be found to shore up rule-of-law institutions that are critical to electoral democracy.

Suggested citation: Donald L. Horowitz, Interethnic Vote Pooling, Institutional Frailty, and the Malaysian Elections of 2018, Int’l J. Const. L. Blog, Jun. 22, 2018, at:

[1] For this history, see Donald L. Horowitz, Ethnic Groups in Conflict (University of California Press, 2000), 398-410.

[2] Harold A. Crouch, Government and Society in Malaysia (Cornell University Press, 1996) 71; Thomas B. Pepinsky, ‘The 2008 Malaysian Elections: An End to Ethnic Politics?’ (2009) 9(1) Journal of East Asian Studies 87.

[3] See Dian A.H. Shah, Constitutions, Religion and Politics in Asia: Indonesia, Malaysia and Sri Lanka (Cambridge University Press, 2017).

[4] All inferences about cross-ethnic voting are vulnerable to ecological fallacy, the hazard of extrapolating from aggregate to individual behavior, which is why vote pooling in the 2018 election awaits a fuller statistical analysis, with ecological fallacy ruled out. This does not, however, preclude probabilistic judgments in cases where the numbers could not otherwise add up.

[5] See the interesting results in Raub, in Pahang, where a DAP Malay candidate easily defeated an MCA candidate and a PAS candidate in a Malay-majority constituency.  Perhaps Malays were inclined to vote for a Malay candidate, but more likely Pakatan parties directed Malay supporters to the DAP candidate.

[6] A point made by winning DAP candidate Ong Kian Ming in a media statement: Ong Kian Ming, ‘A Truly Malaysian Tsunami’ (Dr Ong Kian Ming, 17 May 17 2018) <> accessed 14 June 2018.

[7] The literature on this approach, now called centripetalism, is prodigious.  See, e.g., Benjamin Reilly, “Centripetalism and Electoral Moderation in Established Democracies,” 24 Nationalism and Ethnic Politics 201 (2018); Donald L. Horowitz, “Conciliatory Institutions and Constitutional Processes in Post-Conflict States,” 49 William and Mary L. Rev.  1213-48, at 1214-26 (2008).

[8] See books by the Pakatan Secretary-General and by a DAP leader: Saifuddin Abdullah, New Politics 2.0: Multiracial and Moderate Malaysian Democracy (Institut Darul Ehsan, 2017); Liew Chin Tong, Middle Malaysia: Centre Ground Is Battle Ground (Genta Media, 2013).


Leave a Reply

Your email address will not be published. Required fields are marked *