Elections

74 posts

‘Dubious’ voters may decide GE13

(Also published on Malaysiakini and Selangor Times)

In the first part of this article, I highlighted five problems with the electoral roll which were not addressed in the report by the Parliamentary Select Committee (PSC) on Electoral Reform.

In this second part, I will highlight five additional problems with the electoral roll, all of which concern the highly problematic area of postal voting among army and police personnel.

5. Postal voters who are registered using their regular ICs

Army personnel, who are postal voters, have IC addresses beginning with T. Police personnel, who are postal voters, have IC addressing beginning with R, G and I.

If any of these army or police voters were previously registered as voters using their civilian IC numbers, their civilian registration entries should be deleted from the electoral roll.

NONEWhile this process may have taken place for most of these postal voters, the possibility that these voters appear twice on the electoral roll cannot be discounted. This possibility was not investigated by the PSC on Electoral Reform.

The analysis conducted by Mimos (Malaysian Institute Of Microelectronic Systems) on behalf of the PSC was only restricted to looking for voters who were registered under the same IC number.

Conducting this analysis requires massive amounts of computing power since the name and date of birth of each army and police postal voter has to be matched with the name and date of birth of every single entry in the electoral roll (currently numbering more than 12 million).

Even researchers, political parties and politicians who are interested in investigating the presence of this specific problem in the electoral roll would not have the capacity and know-how to conduct this analysis.

Table 3 below shows two examples of army voters who were also registered as voters using their civilian ICs. The civilian entries were deleted after their inclusion was referred to the Election Commission (EC).

NONE

This problem is not restricted to the presence of one person registering under his or her own civilian IC as well as his or her army/police IC.

It is also possible for the IC of army or police voter to be used by his or her spouse. If this occurs, it would be much harder to detect since it would not be the same name and date of birth registered under two ICs – one army/police and the other civilian.

If the spouse is registered twice, one using his or her own IC, and once using the civilian IC of his or her army/policy spouse, then the same name would appear twice but having to different dates of birth.

Of course, one way to detect this is to check for the mistake in the gender indicated by the IC and the gender of the voter as indicated by the EC records.

The screenshot below shows one such example of a wife who are registered as a voter using the IC number of the husband. The name of the voter is female as is their gender identity in the EC records but her IC number ends with an odd number which indicates that the holder of the IC is male.

When the IC number was inputted into the MyEG (Malaysian E-Government) website, the name of the male postal voters appeared.

After a total of 42 cases were discovered across five parliament seats in Negeri Sembilan (Seremban, Rasah, Telok Kemang, Rembau and Tampin) in Q2 (Quarter 2) of 2011, the EC was obliged to delete these names.

NONE

This was a very time-consuming task since the civilian IC addresses of the postal voters in question had to be ‘tracked down’ using other databases (since they are not given in the EC website). In this particular instance, the MyEG database was used because NRD (National Registration Department) no longer allows open access to their website to check for the validity of ICs.

If one’s wife could be allowed to register using the husband’s IC which has a different name, one wonders if another person could register using any police/army postal voter’s civilian IC address?

Again, this possibility was not addressed in the PSC report. There are currently more than 200,000 army and police postal voters. To ensure that their civilian ICs are not being used by other people, a thorough investigation needs to be conducted using the civilian IC numbers of these army/police postal voters.

6. Spouses of police who are registered as postal voters

According to section 2 of Elections (Registration of Electors) Regulations 2002, spouses who qualifies to be registered as postal voters are as follows:

  • Spouse of a serving member of regular naval, military or air force of Malaysia, the Commonwealth or other country;
  • Spouse of a person in the public service of the Malaysian government or any state/local authority or statutory body who is living with the serving spouse outside the boundaries of Peninsular Malaysia or Sabah or Sarawak;
  • Spouse of a person engaging in full-time studies at any university, training college or any higher educational institution outside the boundaries of Peninsular Malaysia or Sabah or Sarawak and who is living with the student at the time for application for registration as a parliamentary or state elector.

The spouse of a member of the police force is not eligible to be registered as postal voters.

But in our analysis of the postal voters in the June 2011, we found there were over 4,000 spouses of police voters who were registered as postal voters. Table 4 below shows a sample of spouses of police voters who are also registered as postal voters.

NONE

7. Spouses of army/police voters who are of the same gender

In this category, spouses of army/police who are of the same gender – i.e. same sex marriages – were identified.

For example, in the Setiawangsa June 2011 electoral roll, Harisah binti Ab Ghani (IC: 770130035668) was listed being married to Zaini binti Hamzah (IC: T1110543).

When Zaini binti Hamzah’s IC was searched in the EC website, it was found that she had now become Zain bin Hamzah/Lelaki and had been shifted to the Ketereh parliamentary constituency in Kelantan.

In the most recent check, Zain was switched back to Zaini binti Hamzah. Both Zaini and Harisah’s current voting constituency is unknown as both are listed as being ‘processed’ in the EC website.

NONE

NONE

NONE

NONE

8. Army and police voters who are above the retirement age

According to the army’s website, the retirement age for army personnel is 55 (maximum). For certain grades, it is lower than 55. For other civil servants, the mandatory retirement age is 58.

A limited search of the June 2011 electoral roll revealed over 44 names of police postal voters who were above the age of 58.

For example, Wan Rasidy bin Roni (IC: RF151304), born on Nov 19, 1900, which makes him 112 this year, is registered in Balai Polis Lumut in the Lumut parliamentary constituency.

NONE

9. New army and police postal voters who are above the recruitment age

The maximum entry age for army recruits is not more than 30 years of age.

However, an investigation into the voters who were cut from the electoral roll in Q3 2011 because they had joined the army showed more than 200 recruits who were more the maximum age of 30.

Of equal or perhaps greater concern is that these over-aged recruits, who changed their place of voting in the 3Q of 2011, were then moved to another constituency in 4Q 2011.

For example, Abu Talib bin Ahmad (IC: 690625086571), aged 42, was taken out of the 3Q 2011 electoral roll because he had joined the army. He was originally registered in the parliamentary district of Kuala Kangsar in Perak as a regular voter with a civilian IC address.

According to the 4Q 2011 supplementary electoral roll, this voter, who was then registered in the parliamentary constituency of Jeli in Kelantan, was then transferred to the parliamentary constituency of Ketereh as a postal voter.

In other words, in the space of six months, this voter, who was not only over 30 years of age, had also experienced two changes in his voting constituency from Kuala Kangsar to Jeli and from Jeli to Ketereh.

Abu Talib bin Ahamd, aged 42, was registered as a new army postal voter in 3Q 2011.

NONE

Examples such as these raise the possibility that over-aged voters have been illegally ‘recruited’ as army postal voters and then quickly transferred from one electoral constituency to another in successive quarters in order to ‘mask’ the origin of these suspicious army recruits.

In the Q4 2011 data, the EC no longer gave details of why voters were deleted from the electoral roll, which meant that voters who joined the army and police could no longer be detected. One wonders why EC altered the structure of the data released by quarter from 3Q 2011 to 4Q 2011.

How much difference will these ‘problematic’ voters make?

At the end of the day, one should ask the question of how many ‘problematic’ voters are in the electoral roll. The simple answer is that we really do not know.

The cases highlighted under these 10 categories number approximately 100,000. But even within these categories, an exhaustive analysis has not been conducted because of time and manpower limitations.

NONEThese problems could potentially be the tip of the iceberg. They do not include addresses with a large number of registered voters. They do not include the 42,000 voters whose IC numbers could not be found in the NRD database.

They do not include other categories such as the insertion and removal of registered voters who do not appear in any of the quarterly electoral roll updates. They do not include other categories of possible electoral manipulation which have not yet been analysed or investigated such as the registration of voters in non-existent addresses.

The inclusion of these categories could easily increase the number of problematic voters to 400,000 names or approximately 1,800 votes divided over 222 parliamentary constituencies.

Considering that 35 seats out of 222 were won or lost by less than 2,000 votes, these problematic voters can potentially affect the balance of power in the next general election.

History repeats itself?

If one does not think that the presence of these problematic voters is not sufficient to swing the overall election results, I point to two past precedents where sudden increases in the number of voters in the electoral roll in states which were lost by BN which probably helped win the BN win back these states.

The two states are Sabah after the 1990 elections and Terengganu after the 1999 elections. We see the same pattern being repeated in the state of Selangor leading up to the 13th general election.

The number of voters in Sabah increased by 17.6% from 1990 to 1995 compared to the national average of 13.0%. This was just after PBS left the BN coalition prior to the 1990 Sabah state elections. PBS controlled the Sabah state government from 1990 to 1994.

In the state of Terengganu, the number of voters increased by a whopping 17.7% from 1999 to 2004 compared to a national average of 7.3%. Again, it is noteworthy that the state of Terengganu had fallen to the opposition during the 1999 general elections.

In 1999, the opposition won 28 out of 32 state seats and all eight parliamentary seats. In 2004, after this massive increase in the number of voters, the opposition won only one parliamentary and four state seats, an almost total reverse of the 1999 general election results.

In Selangor, according to the Q4 2011 electoral roll, the number of voters has increased by over 340,000 to more than 1.9 million voters since the 2008 general election. This represents an increase of 21.8% compared to a national average of 16.3%.

Some of these increases have occurred in areas which limited increases in the number of new housing estates and population inflows.

NONEFor example, the seat of Hulu Selangor, a marginal parliamentary seat, saw an increase of more than 17,000 voters, or 27.1%, from the 2008 general election to the Q4 2011 electoral roll.

Finally, for those who use the fact that the opposition managed to win a historic number of parliamentary seats and control of five state governments in the 2008 general election to say that the electoral roll is relatively clean and accurate, one could easily respond by saying that the opposition could have won even more seats if the electoral roll did not include these problematic voters.

The upcoming 13th general election is expected to be one of the most closely contested electoral battles in our nation’s history. This increases the importance of having a relatively clean and transparent electoral roll to ensure that the results are a fair representation of the will of the electorate.

Without a strong commitment by the authorities, including the PSC on electoral reform, EC, NRD and the government of the day to firstly identify and then address these problems, the accuracy of the electoral roll cannot be ascertained with any degree of confidence.

Part 1: 10 problems in EC’s electoral roll


ONG KIAN MING holds a PhD in political science from Duke University. He is currently a lecturer at UCSI University.

10 major problems in EC’s electoral roll

(Also published on Malaysiakini and Selangor Times)

The Parliamentary Select Committee on Electoral Reform report, which was released earlier this week, highlights 22 recommendations on how to improve the electoral process.

I will not go through each and every recommendation or discuss the overall quality of these recommendations since others including the Bar Council chairperson and the Bersih steering committee have already done so.

bersih announcing 3rd rally 040412 ong kian mingWhat I will focus on in this two-part article is the many problems which are to be found in the electoral roll. To summarise, the report fails to acknowledge significant problems that have to do with the electoral roll, many of which are already well known, and seems to limit the scope of checking the accuracy of the electoral roll to a few not very useful parameters.

Section 10.6 of the report recommends that an independent agency such as Mimos (Malaysian Institute Of Microelectronic Systems), under the supervision of a newly-established PSC, be given the responsibility of checking the accuracy of the electoral roll.

Unfortunately the examples listed in Sections 10 and 11 of the report are not wide-ranging enough. For example, Mimos was given the task of checking to see if there are duplicate IC numbers in the electoral roll. Not surprisingly, this search failed to turn up any such entries.

Mimos only did a search for duplicate entries using the new IC (identity card) numbers and not the old IC numbers. Nor did Mimos did a search to see if the civilian IC numbers of army and police postal voters (who have their own unique army and police IC numbers) appear on the electoral roll.

As will be clear later in this article, these are but some of the methods in which the electoral roll can be manipulated and abused.

The only significant problem area identified in the report was the presence of addresses with more than 50 registered voters. There were 938 addresses with between 51 to 100 registered voters and another 324 with more than 100 registered voters.

For these cases, the report recommends that the Elections Commission (EC) display these names in public so that the voters in question can come forward to help EC clean up the electoral roll in these addresses.

The Merap project

Any electoral roll in any country will have its share of problems. Voters move and do not update their voting constituency. Voters who have died may not be taken off the electoral roll. Resourceful politicians often shift supporters into their own voting constituency.

These problems may be unintentional administrative errors or they may arise as a result of purposeful fraud. Both types of problems need to be addressed.

But systematically identifying and addressing these problems require a lot of resources not just in terms of time and personnel, but also in terms of computing power and knowledge.

The problems which will be highlighted in these two articles were uncovered as a result of an ongoing academic study called the ‘Malaysian Electoral Roll Analysis Project (Merap)’, whose purpose is to identify problems with the electoral roll and to make recommendations on how to deal with them.

These results took three part-time researchers approximately three months to compile. These findings are not exhaustive and are possibly only the tip of the iceberg of the underlying problems with the electoral roll.

A total of 10 problems with the electoral roll will be examined – five in this first part and another five in the second part.

The 10 problems are:

1) Voters who are above 85 years old.

2) Inconsistencies in the gender indicated by the IC number and EC data.

3) Voters with the same name, and some with the same/similar date of birth.

4) Voters who have IC addresses with the state of birth as ‘7x’, indicating that they are born overseas.

5) Voters in Selangor and Kuala Lumpur in the third quarter and fourth quarter of 2011 who do not have house addresses, even though other newly-registered voters in the same locality have house addresses.

6) Postal voters who are registered using their regular ICs.

7) Spouses of police who are registered as postal voters.

8) Spouses of army/police voters who are of the same gender.

9) Army and police voters who are above the retirement age.

10) New army and police postal voters who are above the recruitment age.

1. Voters who are above 85 years old

An analysis of the third quarter (3Q) of the 2011 electoral roll revealed 65,543 voters who were aged 85 and older. In addition, more than 1,000 voters were listed at 100 years old and above. 19 voters were born before 1900!

This is not to say that all of these voters are fraudulent voters that have been included in the electoral roll by irresponsible parties. Rather, the number of elderly voters is a good place to start in terms of identifying voters who may have died but for whatever reason, have not been deleted from the electoral roll.

This is an area of concern since the ICs of dead voters can be used by others to cast a vote on their behalf.

The oldest voter identified was Wong Kwan Moy (IC: 530823040010) who was listed as being born in 1853, which makes this person 149 years old in 2012.

This voter is registered in Setiawangsa in Kuala Lumpur. It is possible this could be typo and that the voter was born in 1953 rather than 1853 but this needs to be verified.

NONE

But this cannot be said of Tey Kim (IC: 900206740068) who was listed as being born in 1890. This cannot be a typo since voters who are born in the 1990s do not have old IC numbers, which this voter clearly does (K844769).

If Tey Kim was born in 1890 and she is still alive, she would be 122 this year.

NONE

The EC would probably respond by saying that they rely on the National Registration Department (NRD) to provide them with names of recently deceased people who then can be deleted from the electoral roll.

But there is nothing to prevent the EC from conducting their own periodic investigations on voters, especially those who are above 100 years old, to ensure that these voters are still alive and residing in their voting constituency.

2. Inconsistencies in the gender indicated by the IC number and EC data

An IC ending with an odd number means that the person is a male, while an IC ending with an even number means that the person is female.

An analysis of the 3Q 2011 electoral roll found 15,855 voters whose IC numbers indicate a different gender from that listed by EC.

Among these 15,855 voters, there were 72 “bin”(son of) who are listed as female and 76 “binti” (daughter of) who are listed as male.

For example, Mashitah binti Hashim (IC: 550901095040) was listed as a male while Ali bin Ariff (IC: 880726265197) was listed as a female.

These ‘gender inconsistencies’ point to the possibility that a person of a different gender could have used these ICs to register themselves as voters.

NONE

NONE

While these instances could be legitimate errors on the part of NRD, it makes sense to conduct more analysis, including field visits and discussions with NRD to ensure that these are merely clerical errors.

In addition, the presence of these sorts of inconsistencies also raises the question on whether polling agents can object to the identity of these voters on the basis that their listed gender differs from that indicated by the IC number.

These sorts of inconsistencies also highlights the need for NRD to get more involved in the process of clarifying these inconsistencies and discrepancies and play a more pro-active role in helping the EC maintain a clean and updated electoral roll.

3. Voters with the same name and some with the same/similar date of birth

Many Malaysians share the same name and as such many same/similar names are bound to appear in the electoral roll.

But when Malaysians with the same name and the same or similar date of birth appears, often in the same constituency or in the same state, this should certainly raise concerns.

The presence of such voters raises the possibility that they have been issued more than one IC number in order for them to vote multiple times.

For example, an analysis of the 2008 general election electoral roll for Terengganu found 369 voters named “Fatimah binti Ismail” and 346 named “Fatimah binti Abdullah”.
Among the 369 “Fatimah binti Ismails”, 10 pairs were found to have same/similar date of births.

Among the 346 “Fatimah binti Abdullahs”, 13 pairs were found to have same/similar date of births.

NONE

NONE

This seems like too much of a coincidence especially when one considers that Terengganu experienced a significant percentage increase in the number of voters from 1999 to 2004, after the state fell to PAS in the 1999 general elections.

This example illustrates how conventional methods of data mining conducted by Mimos for the PSC on Electoral Reform are insufficient to detect problems in the electoral roll.

More refined searches such as looking for same names with similar or same date of births (the first six digits of the IC number) are required.

4. Voters who have IC addresses with the state of birth as ‘7x’, indicating that they are born overseas

Voters who are not born in Malaysia have ‘71′ as their 7th and 8th digits in their IC number. There has been much speculation that non-Malaysian have been given citizenship and hence, voting rights, in order to have them vote for the ruling government.

Malaysiakini recently reported on a certain ‘Mismah’ being originally listed as a permanent resident and registered as a voter and then later being listed as a citizen in the NRD website.

A preliminary analysis of the September 2011 electoral roll for voters in Selangor turned up 30 voters with ‘Indonesian’ sounding names (only one name). For example, Juwairiyah (IC: 560317715158), is a ‘71′ voter in the Gombak parliamentary constituency.

NONE
Table 1 below shows the full list of these 30 names.

Table 1 below shows the full list of these 30 names.

NONE

This is not to say that those who the digits 71 in their ICs and who appear on the electoral roll are necessarily suspicious voters. But the precedent of illegal immigrants in Sabah, many of them from Indonesia and Philippines, being given fake ICs to register as voters should make one especially wary of such cases.

There were over 167,000 ICs with the state of origin digits listed as ‘7x’ (71, 74 and 75) in the 3Q 2011 electoral roll.

Our findings are thus only a small sub-sample of this larger dataset. While many of these names are of Malaysian citizens born in China and India, much more in-depth analysis needs to be done in order to ascertain the percentage of voters among this dataset, which are legitimate citizens and those who have been given IC numbers through less than legitimate means.

5. Voters in Selangor and Kuala Lumpur in 3Q and 4Q of 2011 who do not have house addresses, even though other newly-registered voters in the same locality have house addresses

One would not expect that voters who are newly registered in Kuala Lumpur or Selangor would still be registered in addresses without house numbers.

Given the 100 percent urbanisation rate in Kuala Lumpur and an almost 100 percent urbanisation rate in Selangor, it is surprising that some voters would still have IC addresses without house numbers.

What is even more surprising is that there are still voters who are registered in addresses without house numbers in the same locality, where other newly-registered voters are registered in addresses with house numbers.

For example, according to Table 2 below, Syahrul Ali Mansur (IC: 691023715099) is a newly-registered voter in 3Q 2011 in Kampung Puah, which is in the Wangsa Maju parliamentary constituency.

Table 2 also lists nine other newly-registered voters in 3Q 2011 in the same locality which do have house numbers.

NONE

Voters who fall into this category are still being compiled from 3Q 2011 and 4Q 2011 in Kuala Lumpur and Selangor.

In Part 2 of this article, five problems which have to do with postal voters will be examined in greater detail.

Part 2: ‘Dubious’ voters may determine GE13 outcome


ONG KIAN MING holds a PhD in political science from Duke University. He is currently a lecturer at UCSI University.