The need to quantify racially-weighted vote manipulation – an informal response to Clive Kessler
The most telling of these is the finding by the Merdeka Centre, as reported by Malaysiakini, that ‘only 11 percent of Malays polled just before the 13th general election said the protection of Malay community interests and the community’s political clout was a concern’; the economy was their dominant preoccupation.
Clive Kessler’s recent article, ‘Malaysia’s GE13: What happened, what now?’, is a plausible and insightful account of what was going through UMNO’s collective mind with respect to their GE13 strategy. In particular, his observation that rather different and even contradictory messages were delivered to international and domestic audiences is crucial to understanding UMNO’s general mindset and the campaign as a whole, and it is indeed a pity that these audiences do not seem to have compared notes.
Nevertheless, while understanding how UMNO thinks may be important and interesting, in the long run it is much more important to understand what was and is going through the minds of the voters themselves.
Kessler’s main argument in this direction is that, having first malapportioned the rural Malay vote in West Malaysia into disproportionate importance, UMNO won this vote – and the election – by creating a ‘Malay siege mentality’, to the detriment of PAS. This claim is based mainly on the increase of UMNO and the decrease of PAS seats in Parliament; one could also argue that it is supported by Thomas Pepinsky’s quantitative analysis, showing positive correlation between the percentage of ‘Malay’ voters and the percentage of votes for Barisan Nasional for West Malaysian parliamentary constituencies. These analyses are necessarily incomplete, as they do not study in any detail the kingmaker in GE13: East Malaysia, with its 57 parliamentary seats (out of 222 in total). Moreover, important data exist which contradict Kessler’s story.
The most telling of these is the finding by the Merdeka Centre, as reported by Malaysiakini, that ‘only 11 percent of Malays polled just before the 13th general election said the protection of Malay community interests and the community’s political clout was a concern’; the economy was their dominant preoccupation.
In addition, Wong Chin Huat argues in a data-based analysis that GE13 malapportionment was partisan rather than rural/urban or based on race. (See also related articles on fz.com from June/July of this year). For example, he points out that the urban constituency of Alor Star has many fewer voters than Baling, which is rural, but is inclined to vote for the opposition and has a history of social activism. Wong points out further that the three Pakatan Rakyat parties – PKR, DAP and PAS – received similar numbers of votes; however, for reasons which remain unclear, PAS ‘paid’ the most votes per seat, followed by PKR then DAP. This led to the imbalance in the number of seats won by each party. So PAS did not lose votes in GE13, but they lost seats. In fact PAS’s share of the total vote increased from 14.05% to 14.77% while UMNO’s remained roughly constant (29.45% vs. 29.33% in 2008).
I would like to suggest that a hitherto neglected factor in GE13 analyses – the fact that the ‘manipulability’ of the vote is far from race-blind – will go some way towards reconciling several superficially contradictory analyses, such as those cited above, and yield additional insight into the thought processes of Malaysian voters.
It is easy to see that racially-weighted vote manipulation was a potentially significant factor in GE13. After decades of institutional racism, most public institutions, including the civil service and armed forces are now almost entirely made up of Malaysians officially classified as ‘Malay’. Other government-linked, centrally-controlled institutions engineered to be ‘Malay-dominated’ include Federal Land Development Authority (FELDA) settlements and public universities, particularly ‘bumiputera-only’ UiTM (which has 120,000 students). These institutions provide both means and opportunities for UMNO/BN to introduce pressure mechanisms operating on large blocs of voters who ‘just happen’ to be disproportionately ‘Malay’. Depending on the efficiency and extent of these mechanisms, they could significantly affect quantitative psephologic analyses such as Pepinsky’s: ‘statistical inclinations’ of voters labelled as ‘Malay’ to vote for UMNO/BN may exist for reasons quite unrelated to any function of their free will – or even their ‘ethnicity’ or ‘ethnic (siege) mentality’ as such – but simply because, by design, they formed the overwhelming majority of voters who were ‘in the wrong place at the wrong time’, e.g. in a naval base on (early) polling day. In other words, analyses such as Kessler’s may be naive in assuming that the only ‘racial’ levers UMNO/BN have are diffuse ‘ethnic psychological’ ones. In general, analyses of GE13 have neglected the effects of vote manipulation, whether racially-weighted or not, taking the official results at face value.
To be more specific, when one speaks of ‘vote manipulation’, there is most obviously outright fraud. Almost 400 000 civil servants and armed forces personnel and their spouses participated in advance or postal voting, which is seen as being easier to manipulate than regular voting. For instance, ex-armed forces personnel have alleged that votes were cast ‘on their behalf’. It was also recently reported that some Election Commission officers (almost always drawn from the civil service even though this does not in principle have to be the case) were offered RM100 for their ballots. Thus, given the large number of voters involved, efficient and discreet manipulation of early and postal voting by itself may already account for a large part of any difference in voting patterns between different ‘races’, and may have even won the elections for UMNO/BN.
In addition to fraud, civil servants and armed forces personnel (~1.5 million voters out of a total of 13.3 million) face pressures of various kinds to vote a certain way – or not to vote – and never to oppose ‘the government’, which in certain contexts is code for the BN coalition. The ‘Akujanji’ that all civil servants have to sign is an explicit example of such pressure. In the Bersih 2.0 preliminary report on the elections, it was noted that ‘a military personnel, Major Zaidi was demoted for making a report about the ineffectiveness of the indelible ink’.
The fear-mongering has been so successful that well-educated, middle- and high-ranking civil servants have been known not to vote, fearing repercussions for their careers ‘if they find out that I voted for the opposition’. Even long-standing loyalties can be weakened by hierarchical pressures: Recently, after the school magazine had been printed, the principal of a school in my area went through every copy and tore out a page sponsored by the local Member of Parliament, ‘dengan ingatan tulus ikhlas’. The principal and the MP are (or perhaps were) childhood friends. (The MP is, obviously, from the federal opposition).
Just before GE 13, while visiting friends in a neighbouring village, I was asked by a university graduate in complete seriousness whether they, as a civil servant, were allowed to vote for the opposition. Such stories are legion. One also needs to bear in mind the existence of spillover effects on retirees and family members who are not included ~1.5 million figure above, especially spouses participating in early or advance voting. As another example, for a certain period of time my grandparents on one side (who were peasants) felt that they ‘should vote for the government’ (i.e. the BN in their minds) as many of their children were ‘working for the government’. This partly the result of continuous, long-term efforts by UMNO/BN to create a one-to-one identification in the minds of Malaysian between themselves and ‘the government’, e.g. by putting up signs saying ‘Satu Lagi Projek Kerajaan Barisan Nasional’ next to road works.
Pressure mechanisms in place in other ‘Malay-dominated’ public institutions (including FELDA and public universities) as well as the possibility that other vote manipulation strategies (such as vote-buying and closing polling stations early) were racially-weighted will also need to be addressed in a full analysis of the situation.
What would have happened, and what might happen in future elections if these various vote manipulation techniques had been removed or at least rendered much less effective? On a global level, considering only postal and advance votes, a recent Merdeka Centre study strongly suggests that the BN would have lost GE13. As a case study, the parliamentary constituency of Lumut, where Malaysia’s main naval base is located, could be of interest to inquiring journalistic or academic minds. During the last elections, rumours on the ground had it that the PKR candidate, retired admiral Mohamad Imran, as an ex-Navy insider was able to effectively counter vote-manipulation strategies usually employed for the Navy vote. He won his seat with 55.6% of the vote against BN/MCA’s Kong Cho Ha, the incumbent since 2004. In 2008, Kong Cho Ha had won in Lumut with 50.3% of the vote. Dare one suggest that the main reason for this upset might be fraud- and intimidation-minimisation rather than ‘racial concerns’? Dare one further suggest that other non-racial, non-religious factors may have also played important roles? One notes for instance well-organised local opposition to development projects in the Lumut area, based on environmental concerns. In addition, Mohamad Imran is seen by many locals as cleaner, more sincere, ‘truly religious’ (this from a non-Muslim local) and as having less baggage compared to Kong Cho Ha.
A Malaysian at Sciences Po recently wrote a fine article on public suicides in France, in which the reluctance of the SNCF (France’s national state-owned railway company) to ‘profile’ railway suicides was noted. One official was quoted as saying, ‘It’s a taboo subject. We don’t have any study on the profile – so to speak – of people who kill themselves….It is delicate to interpret this. We should avoid hasty interpretations.’
One sometimes wonders if a small dose of this caution would not be salutary in analyses of Malaysian elections, especially race-based ones hastily done the day after the polls.
If one must nevertheless conduct race-based analyses of Malaysian elections, given the large number of voters potentially involved, it would only be intellectually honest to try to take into account the effect of racially-weighted vote manipulation – both in the form of fraud as well as undue influences through the institutions mentioned above (and likely others that have escaped my attention). Most commentators, if not all, have neglected to do this.
Apart from numerically estimating what the election results might have been in the absence of various forms of racially-weighted vote manipulation, among the questions one might attempt to answer is whether, as suggested by Wong Chin Huat, the electoral map was drawn with purely partisan interests in mind (in which case one might have expected all three PR parties to have had similar vote/seat ratios), or whether – more sinisterly – mechanisms were put in place to create a perception of racial polarisation for casual observers and ‘day after the polls’ analysts, reinforced by well-timed talk of ethnic tsunamis and ‘reconciliation’. Or perhaps both.
Before leaving this topic, I note another possibly statistically important factor for ‘race-based’ analyses: allegations that some voters, especially indigenous peoples, were prevented from voting [Read the following linkes: Link 1. Link 2. Link 3. Link 4]. This could also lead one to ask to what extent the perception that East Malaysia is a BN fixed deposit is well-founded, and whether there was similar ‘orientally-weighted’ vote manipulation and to what extent this was effective. The upcoming Bersih 2.0 People’s Tribunal on GE13, and any evidence this might yield, will be of great interest for developing these lines of thought.
It is, I admit, rather bad form to throw out ideas in this fashion without having done some of the number-crunching and on-the-ground investigation suggested. In my defence, my appointment is in physics and I have very limited time to work on Malaysian issues. These ideas are thus ‘up for adoption’ and I hope they will find good homes.
Coming back to Kessler’s article, he further suggests that his insights were missed by foreign journalists as well as the Malaysian pundits they had access to, the latter tending to be city slickers disconnected from the rural reality and fond of using words such as ‘discourse’ and ‘narrative’. Speaking only for myself, as someone who may have let slip even ‘hermeneutic’ on occasion, I am for the record from Malaysia’s rural agricultural heartland, where my family has lived and tilled the land for generations. One school I attended had a dropout rate of 50% and was classified as ‘sekolah perintis’ – a euphemism if there ever was one. When I was a child, food was rationed in our home – and I am not talking about meat, which we almost never saw. So, some of us rural folk are out there, saying things, even if we do not always mention our roots.
Speaking therefore as your friendly neighbourhood country bumpkin, while it may be the case that many city-dwellers are disconnected from life in the country, the reverse is not necessarily true. Many rural families have children or other relatives who live and work in the city and through them news, goods and ideas make their way out into the country; this is increasingly the case as Malaysia continues to follow a trend of rapid urbanisation. (72% of the population lived in urban areas in 2010, with an annual rate of increase of 2.4%. In addition, ~1 million Malaysians live abroad, about half of these in Singapore). It was precisely to make use of these ties to the city and the outside world that Haris Ibrahim has been running a ‘balik kampung, bawa berita’ campaign. Proportionally fewer rural folk may be able to wax lyrical on notions such as ‘separation of powers’ and ‘independence of the judiciary’; however, ideas such as justice, honesty, kindness, generosity, openness and solidarity are universal values which can be understood by all. Some even say that these values are more present in the countryside, with its strong social networks, than in urban contexts. Let us not forget that, after all, the Baling Incident of 1975 happened in rural Baling and not urban Alor Star.
