For the first time, the incumbent government has lost the popular vote nationally (in 2008, it was only in the peninsula). The BN coalition has still managed to hold on to power.
This piece, in a series analysing the election results, looks at the concerns raised regarding the electoral process and the potential impact these issues may have had on the final results.
In analysing the fairness of any polls, one asks whether the irregularities in the process could have affected the final outcome. Were the problems enough to change which coalition would have formed government?
These issues will be debated and assessed in the days and weeks ahead. Let me share some preliminary observations that suggest that in this election, some things appear not to be quite right.
Integrity of electoral roll
This was the longest wait for an election, and both sides were extremely active in registering new voters, especially in the urban areas where the party machinery was well honed.
Even factoring in the more robust voter registration efforts, changes in electoral procedures to register people where they live rather than where they are from, population demographics, and possible housing developments in different seats, the increased numbers in the electoral roll are significantly not in line with historical patterns of voter registration. This out-of-line pattern is in every state, except Negri Sembilan.
The figure that stands out in voter increase occurred from 2004 to 2008 in Sabah. The questions about the electoral roll in Sabah have been long standing, and are the subject of the ongoing royal commission of inquiry into immigrants.
These increases from 2004 through 2008 are by any measure - huge - in places such as Liburan, where caretaker Chief Minister Musa Aman's state seat is located, in Semporna, the seat of Shafie Apdal and in Ranau currently held by Ewok Ebin.
Yet, after 2008, while the numbers have dropped, there is still on average 21% new voters in Sabah seats, a high number not in line with demographic trends. Migration appears to continue be a factor shaping voter numbers in Sabah in this GE13, despite calls to tighten the flows.
We also find that new voters have flooded states like Selangor, Pahang, Terengganu and Johor in GE13. The average increase in voters nationally between 2004 and 2008 was 8.2%. In the run-up to GE13, the voters registered doubled to 19.4%. The national and statewide averages however obscure the differences among different seats within states. It is clear that some seats have been special recipients of new voters.
Much has been made of the 28% of new voters in Lembah Pantai. This seat is actually on the low side compared to others. Consider the whopping 61.5% increase in Tapah, recently re-won by BN, or Subang with 52% new voters, won by Pakatan with a larger majority this election but shaped heavily by Pakatan's registration of new voters.
A total of 90 seats, or 41% of all parliamentary seats, have more than 25% new voters. Many of these were in races with tight contests in 2008, and continued to have tight contests in GE13. The new voters have advantaged the opposition in urban areas, but benefitted the BN in rural and semi-rural areas or in states where the machinery of the opposition is comparatively weak, such as Johor.
Such races also won by BN that had large number of voters include Cameron Highlands (20%), Pasir Gudang (39%) and Tebrau (45%) in Johor. While some of the increase in the latter two seats might be explained in part by development, bizarrely there are sharp increases in voting populations in the remote interior state of Pensiangan (33%) and remote coastal seat of Kota Marudu (32%) in Sabah. These abnormal high increases raise questions.
The placement of new voters is even more intriguing when studying the actual polling stations results. Many new voters are concentrated in more less populated areas within constituencies, often in rural and semi-rural seats.
This is where the questions over the large number of unexplained voters grouped in bunches in places like Bachok (21% new voters and won by PAS with less than 1% margin) and Bukit Gantang (29% of new voters and won by PAS with 2% margin) come in.
It appears that the localised remote placements of new voters may have had an impact. For example, the placement of 3,600 new voters in a remote Felda scheme occurred in Segamat, which was won by the BN with a 1,217 majority. The voting in this Felda scheme was over 90%, with one stream at 99%. In 2004, the voter turnout in this area was much lower.
This spike pattern of voter turnout in particular polling stations was found in Terengganu in 2004, when the BN wrested back the state, and questions were raised at that time as well.
Spike patterns out of line
This GE13 spike in voter turnout at the local level is being witnessed in specific places across the country. With the national level of turnout at 85%, the spike patterns that are well out of line with historic patterns of voting behaviour raise questions, even accounting for the overall rise in participation and voter turnout.
Another pattern in the placement of new voters beyond tight races involves prominent leaders getting large shares of new voters, such as Najib Razak's own seat Pekan with 38% new voters, or Rompin represented by Jamaluddin Jarjis at 29% new voters. It remains unclear why these largely rural constituencies would have such large voter increases.
Generally out-migration areas such as Perak and Pahang receiving large numbers of new voters does not conform with population patterns. Why are places with people leaving to work outside get sharp increases in voters?
The lack of clear transparent explanations on why voters are registered in some areas in such high numbers this election, compared to past patterns in these areas, understandably raises questions.
Many seats that were lost by the opposition or were in tight races have large number of new voters, including, including Tanah Merah (24%) and Balik Pulau (25%), although in some cases the opposition picked up or retained seats with large voter increases in these seats, such as Kota Raja (47%) and Kuala Nerus (25%), among others.
This issue of voter registration and voter turnout levels needs further study, with more information on who are these new voters and their pattern of voting. The fact is that the polling station results will show the spikes at the local level and careful study will tell us statistically the impact of these new voters on electoral outcomes.
The Electoral Commission (EC) and electoral administration as a whole are facing a real trust deficit. A reliable electoral roll is essential for any fair elections. Repeatedly questions have been raised about the veracity of many new voters.
Election watchdog Merap and others have time and again drawn attention to the questions of electoral roll integrity. Before the polls, these matters were essentially ignored or dismissed. To date, the scope of phantom voters and questionable placement is not fully known. Now the results themselves will show the impact at the local level.
This is why the sharing of all results through the Borang 14 is essential in order to make a systematic and thorough assessment. Preliminary reviews of results are already raising red flags as they have shaped the outcomes at both the parliamentary and state levels.
Early and postal voting
Queries about the early and postal voting have also emerged. Here the question is about double voting, with individuals having the opportunity to vote twice. Postal voting numbers increased in this election. Historically, there have always been questions about the veracity of postal voting, with reports questioning that this voting is secret and others arguing over the accuracy of the results.
There have been improvements in recent years over postal voting involving polling agent access to this process in many locations. Yet, even with these improvements, questions about whether postal voting is fair and accurate remain.
In this election, further questions emerged over the numbers and placement of these postal voters in different constituencies. Many tight races, such as Sibu, had increases in postal voters. In some cases, the list of names of new postal voters were reportedly not provided openly.
Early voting, an estimated 240,000 people, is also a new addition for this election and being queried. Early voting includes many Malaysians within Malaysia, such as the wives of army officers and journalists who can vote before polls.
There was not a clear distribution of the list of early voters provided nationally, and in some cases even individual candidates were not able to access the names of who were the postal and early voters.
No clear explanation was given to why some constituencies received early voters and others did not. Importantly, this information was not properly shared so that it could be verified. Furthermore, there were unexplained instances when the numbers of early and postal voters increased. In Lembai Pantai, for example, the number stated was 200, but 600 showed up. How did this happen?
Given the reality that the indelible ink was in many cases not indelible, the possibility of double voting exists. On voting day there are numerous reports of individuals finding out that someone had voted fraudulently using their name, leading to concerns also about electoral disenfranchisement.
There were also reports of non-Malaysians being transported to the polling stations by buses and even flown in, some of these believed to be phantom voters. The scale and impact of these on the results is not yet clear, but given the combination of a non-transparent early and postal voting process in various locations and non-indelible ink issues on election day, and sightings of non-Malaysians in contentious seats, troubling questions are being raised.
The close results make these issues and questions more salient. A total of 72 of seats (or 32%) were won by less than 10% margins of turnout. Twenty percent of seats, 44 seats, were won with less than a 5% margin. The closeness of these races could easily have come down to a few voters. These razor-thin margin seats were won by both sides, but given the questions raised about the process of voting in these close seats, they need to be carefully reviewed.
To date, the total number of seats affected by either non-transparent new voter increases and early voting allocations and unexplained incidents of disenfranchisement appears to be more than the actual margin of victory for the BN. These reports need to be properly vetted and verified, but fundamental questions remained.
A spoilt-vote victory
Finally, this brings up the questions on the election night itself. There are queries surrounding the recounts and spoilt votes. How many recounts which overturned the results at the last moment were there? In Perak, for example, three state seats - three is a famous number in Perak - Alor Pongsu, Manjoi and Pangkor results were overturned at the last minute. Questions were also raised at Kamunting as well.
The need for transparency at the final count is essential for a fair election. When the EC asks people to leave and there are new ballot boxes seen outside of a polling station, as was reported in Lembah Pantai, there are questions. It is not fully clear what exactly happened with the recounts in Perak and elsewhere - as there were numerous recounts nationally this election - but the climate of distrust that has permeated the assessment of the election process raises doubts.
In the days ahead, a better sense of the numbers and recounts will emerge. With reports of sudden changes in the voting results such as Bentong and Labis, questions are being raised. Many people cannot understand how a result that was statistically a large margin ahead could be overturned. These need to be clarified, particularly in Bentong where the margin was larger.
Part of the problem is that in some cases, the number of spoilt votes exceeded the actual majority in places where recounts took place. Here are some of the seats at the parliamentary level where this happened: Kuala Selangor, Cameron Highlands, Bachok, Bentong, Sungai Besar, Kota Merudu and Baram won by the BN and Sepang and Kuala Nerus won by the opposition. Another seat with high spoilt votes is Segamat, at 950.
What distinguishes these close recounts from the famous cases of Lim Kit Siang and Karpal Singh losing in 1999 with more spoilt votes than the majority, is the process of the administration of the indelible ink in this election - before marking the ballot paper - thus staining the papers and contributing to higher spoilt votes.
This pattern of higher spoilt votes than actual margins of victory was also replicated at the state level as well in many areas, where only a few seats mattered for who should win state power. The process of administering the ink appears to have had an impact on the results in some areas.
It is important to be careful when reviewing the election results and not rush to judgement about what happened and why. It is also important to see the election holistically. The focus here has not touched on the use of money in the campaign, which was rampant, labelled ‘bombing' in Sabah, or the mainstream media reporting.
The aim has been to raise the preliminary questions revealed in the results and the impact actual numbers of voters associated with the election. As the evaluation of the election moves forward, the call to answer these questions will only increase and intensify. Further study and analysis is essential.
Nevertheless, from the non-indelible ink and spikes in voter turnout to being not allowed to vote, concerns with the electoral process itself are not sitting right with many in the public, and this is not just supporters on one side or another. Transparent and truthful answers are both needed and welcomed.
DR BRIDGET WELSH is associate professor of political science at Singapore Management University. She is travelling around Malaysia to provide her GE13 analyses exclusively to Malaysiakini. Bridget can be reached at email@example.com.