(Also published on Malaysiakini)
In two previous articles, I highlighted 10 problems associated with the electoral roll as part of the preliminary findings of the Malaysian Electoral Roll Analysis Project (Merap), a research effort to identify and understand problems with the existing electoral roll.
Firstly, approximately 3.1 million voters were identified as potential non-resident voters by the National Registration Department (NRD) in 2002. This data was given by NRD to the Election Commission (EC) but no action was taken by the EC to assess the magnitude of this problem and to identify ways to rectify it.
Secondly, using the EC’s own data which lists the nationality or ‘bangsa’ of each voter, approximately 65,000 voters were identified as having foreign nationalities. Of these, close to 90 percent or 58,000 had IC numbers which indicate that they were born in Malaysia.
In addition, approximately 49,000 of these voters came from one state alone – Sabah – which has a well-documented history where ICs were given to illegal immigrants in order to allow them to register as voters.
Thirdly, by comparing the electoral rolls in Quarter 4 (Q4) 2010 and Quarter 3 (Q3) 2011 and by cross-checking them with the quarterly updates in the first to third quarters in 2011, it was found that there were 106,743 voters removed from the electoral roll without public display and another 6,762 voters added to the electoral roll during the same period, also without public display.
Fourthly, important information provided by the EC in the Q1 to Q3 2011 quarterly roll updates such as the reasons for the deletion of names from the electoral roll were omitted in the Q4 2011 quarterly roll. This omission immediately raises concerns about the possibility of concealing important information by the EC in order to prevent detailed analysis from being conducted.
Adding the figures which I found from the initial preliminary analysis to the figures highlighted in this article, I have identified approximately 3.4 million cases where further investigation needs to be conducted by the EC in order to verify the legitimacy of these voters because this is the pool of problematic registrations that currently tars the integrity and accuracy of the electoral roll.
1) 3.1 million potential non-resident voters
In 2002, the information systems section of the NRD (Bahagian Perkhidmatan Sistem Maklumat) conducted an exercise called Projek SPR where the IC addresses of all voters in Peninsular Malaysia were compared with the voting constituencies of these voters.
The main purpose behind this project was to identify all the voters whose IC addresses did not correspond to the constituencies that these voters were registered in.
For example, if a person’s IC address was somewhere in Bukit Gasing in Petaling Jaya, which is located in the PJ Selatan parliamentary constituency but if this person is registered as a voter in the Shah Alam parliamentary constituency, then this voter would be classified as a potential non-resident voter (pengundi luar kawasan).
A total of 3,082,482 million voters were identified as potential non-resident voters, meaning that their IC addresses did not correspond with their voting constituencies. This constituted 37 perrcent of the then 8.3 million voters in Peninsular Malaysia in 2002.
Of course, not all of these 3.1 million voters are non-resident voters. For example, a person’s IC address may be that of his parent’s home in Bukit Gasing, but he may be currently residing in a house in Shah Alam and registered as a voter in that constituency.
Article 119 (1) of the Federal Constitution states that “Every citizen who (a) has attained the age of twenty-one years on the qualifying date; and (b) is resident in a constituency on such qualifying date or, if not so resident, is an absent voter, is entitled to vote in that constituency in any election to the House of Representatives or the Legislative Assembly…”
This means that only those who are residents in a particular constituency when they register as voters are qualified to vote in that particular constituency.
The presence of such a large number of voters whose IC addresses do not correspond with their voting constituencies clearly shows that the problem of non-resident voters is a serious one, even if only a small proportion of these 3.1 million voters were not residents in these constituencies at the time of their registration.
The possibility of these voters not being resident voters is further enhanced when one considers the lax standards required before 2002 in order to ‘qualify’ as a resident of a constituency. All someone needed to do prior to 2002 was to show some sort of proof that he was a resident in a particular area – an electricity or phone bill for example.
A number of politicians whom I interviewed on an informal basis told me that the requirement for this proof of residency was often waived. This meant that many strategic politicians could pad their respective constituencies with supporters, friends and family, even though they did not live in that constituency.
One manifestation of this voter padding exercise is the presence of houses with a large number of registered voters.
According to analysis done by Mimos (Malaysian Institute Of Microelectronic Systems), as instructed by the parliamentary select committee on Electoral Reform, 938 houses with between 51 and 100 registered voters each and 523 houses with more than 100 registered voters were identified in the most recent electoral roll.
If we include the 2,071 houses with between 21 and 50 registered voters each and the 3,949 houses with between 11 and 20 registered voters, there are approximately 260,000 voters who are registered in houses with 11 or more registered voters.
And this excludes localities with no house addresses (kampung or longhouses, for example). This is a legacy system left behind by the pre-2002 electoral roll where non-resident voters were registered in addresses that they did not live in.
Perhaps the most worrying aspect of this NRD analysis is that the EC, as far as we know, made no effort to identify the magnitude of this problem – how many of these voters were actually non-resident voters – nor did the EC seem to take any measure in order to rectify this problem in the pre-2002 electoral roll.
2) 65,000 ‘foreigners’ in the electoral roll
The EC collects comprehensive data on the ‘bangsa’ or nationality of individual voters. This data is not distributed to political parties, not officially at least, but contains valuable information that can be used to identify problematic registrations.
Table 2 below lists the nationalities which I classified as ‘foreign’. I then extracted all the voters from the Q3 2011 electoral roll which were categorised as belonging to either one of these nationalities or ‘bangsa’.
I identified approximately 65,000 ‘foreigners’ on the Q3 2011 electoral roll using this list of ‘foreign’ nationalities. Out of these 65,000 voters, I found that approximately 58,000 or 90 percent had IC numbers which indicated that they were born in one of the states in Malaysia.
This was quite unexpected, since I presumed that most of these foreigners would have IC numbers which would indicate that they were born outside Malaysia rather than in Malaysia.
After all, if we met 10 random ‘Italians’ who are Malaysian citizens, wouldn’t we expect most of them to be born outside Malaysia and subsequently became naturalised Malaysian citizens? Instead, we find the opposite phenomenon, whereby a majority of these foreign voters are born in Malaysia instead.
More worrying is the fact that approximately 49,000 voters out of 65,000 are foreign voters, or 75 percent, and they are in Sabah, where cases of illegal immigrants being given Malaysian ICs has been well-documented.
Out of these 49,000 foreigners in Sabah, almost 48,000 or 97 percent are ‘Malaysian born’.
Table 3 below shows the distribution of these 65,000 foreigners by state and the number and percentage of these voters with IC numbers that indicate that they were born in Malaysia.
Armed with this kind of information, the EC can easily do an audit of these foreign voters in Sabah, perhaps in cooperation with NRD, to verify if their identities are indeed legitimate and that they are resident voters. As it is, the presence of these voters and the high percentage of them who are Malaysian-born certainly raise serious concerns.
3) Mysterious deletions and additions
In every quarterly update of the electoral roll (Rang Daftar Pemilih Tambahan, or RDPT), the names of those who are newly registered voters, those who have changed their addresses and those whose names have been deleted from the electoral roll in a constituency have to be publicly displayed in a location within that constituency, usually in a government building.
The number of voters who have been added and deleted from the gazetted electoral roll before and after this public display and after the possible objections to the inclusion and the exclusion of some names must add up.
For example, if the gazetted roll had 10,000 voters before the quarterly update, during which 1,000 voters were added and 500 voters were deleted, the gazetted roll after public display should contain 10,500 voters.
My team of researchers compared the number of voters in the Q4 2010 with the Q3 2011 electoral roll and cross checked the numbers with the Q1, Q2 and Q3 2011 electoral roll updates. What we found was that 106,743 voters in the Q4 2010 electoral roll no longer appeared in the Q3 2011 electoral roll and also did not appear in the Q1 to Q3 2011 electoral updates as deleted voters.
This means that they were deleted from the electoral roll without public display. In addition, we found that 6,762 new voters were found in the Q3 2011 electoral roll but these new voters did not appear in any of the quarterly updates in 2011, meaning that these voters were added without public display!
The presence of such a high number of cases of additions and deletions without public display raises doubts as to whether these voters were removed or added under suspicious circumstances. It also raises questions about the number of voters that were removed/added without public display in the quarters prior to 2011.
4) Information on deletion of voters missing
The data given by the EC to the political parties for the Q1 to Q3 2011 quarterly updates had detailed information with regard to the reasons as to why a voter was deleted from the electoral roll.
These reasons include: death, change of address, having joined the police or army, losing one’s citizenship, losing the status of being a postal voter and doubtful identities, just to name a few.
Tables 4 to 6 below show the number of voters added and deleted and the reasons for these deletions for the Q1 to Q3 2011 electoral roll updates.
But all of a sudden, the reasons as to why voters were deleted from the electoral roll were no longer included in the Q4 2011 electoral roll update.
What is more shocking is no voter was listed as being deleted as a result of death. The only data given is the number of voters who were deleted from the Q4 2011 electoral roll as a result of a change in address.
The summary of the Q4 2011 electoral roll update are given in Table 7 below.
One cannot help but wonder if the EC wanted to exclude the voters who had been deleted from the electoral roll because of the intense scrutiny given to the earlier Q1 to Q3 2011 electoral roll updates.
For example, the 42,051 voters whose identity was questionable because of their details could not be verified in the NRD database was revealed in the Q3 2011 update (highlighted in yellow in Table 6 above).
In addition, the sudden increase in the number of deletions in Q3 2011 because of voters entering into the army/police force (highlighted in blue in Table 6 above) also may have drawn unwanted attention as to where these new army/police postal voters were being transferred to.
In-depth audit of electoral roll needed
The data shown in this article raises many concerns about the EC’s desire to ensure a clean, accurate and transparent electoral roll.
The EC has clearly failed to act on important data with regard to problematic registrations that were handed over to it by NRD back in 2002 – the 3.1 million potential non-resident voters – and also data that it collected on its own – the 65,000 foreign voters in the Q3 2011 electoral roll.
In addition, the EC seems to have concealed important data from public knowledge, such as the nationality or ‘bangsa’ of the individual voters, deleting and adding voters without public display and excluding the reasons why voters were deleted from the electoral roll.
The EC should conduct an in-depth audit of the electoral roll, including further investigations into selected samples, and then consider various solutions to these problems, including automatic registration of all voters according to their IC addresses.
With the presence of these problems in the electoral roll, it is clear that the EC chairman’s claim that Malaysia has world’s cleanest electoral roll is one that is not based on reality or facts.
ONG KIAN MING holds a PhD in Political Science from Duke University. He is a lecturer and political analyst at UCSI University. He is also the project director of the Malaysian Electoral Roll Analysis Project (MERAP).