Bersih 3.0 may have been a success in terms of its ability to attract more than 100,000 Malaysians from all walks of life to come out in support of clean and fair elections but the hard work of ensuring that this aspiration is translated into reality needs to continue.
All the more so when the Election Commission continues to ignore the many problems highlighted by the Malaysian Electoral Roll Analysis Project (Merap) with regard to the electoral roll.
The EC has recently published a booklet with 12 points of rebuttal against allegations that the electoral roll is full of problematic registrations and that the EC has been complicit in not cleaning up the electoral roll.
In this article, I want to highlight two of the rebuttals to illustrate that the EC has not been honest in its responses which reflects its inconsistent policies in managing the electoral roll and electoral boundaries.
EC's boundary 'correction' exercise
One of the EC's rebuttals is that it has not redrawn any electoral boundaries since such an exercise would require parliamentary approval. What the EC has been doing instead, is to correct previous errors of putting voters in certain localities in the wrong parliamentary and state constituency.
This was done after the introduction of a new Electoral Geographic Information System (Egis) in 2004. The EC used the power given to the chief registrar under the Section 25(3) of the Elections (Regulation of Electors) Regulations 2002 where it is state that:
"Where an error has resulted in any person being registered as an elector of a registration area which is not a registration area in respect of which that person should have been registered, the Chief Registrar may enter that person's name in the principal electoral roll or the supplementary electoral roll for the appropriate registration area."
This was the reason given for why Selangor Menteri Besar Khalid Ibrahim's voting areawas shifted from Petaling Jaya Selatan (P105) to Lembah Pantai (P121) - Section 16/2 is part of Wilayah Persekutuan rather than Petaling Jaya.
All the more so when the Election Commission continues to ignore the many problems highlighted by the Malaysian Electoral Roll Analysis Project (Merap) with regard to the electoral roll.
The EC has recently published a booklet with 12 points of rebuttal against allegations that the electoral roll is full of problematic registrations and that the EC has been complicit in not cleaning up the electoral roll.
In this article, I want to highlight two of the rebuttals to illustrate that the EC has not been honest in its responses which reflects its inconsistent policies in managing the electoral roll and electoral boundaries.
EC's boundary 'correction' exercise
One of the EC's rebuttals is that it has not redrawn any electoral boundaries since such an exercise would require parliamentary approval. What the EC has been doing instead, is to correct previous errors of putting voters in certain localities in the wrong parliamentary and state constituency.
This was done after the introduction of a new Electoral Geographic Information System (Egis) in 2004. The EC used the power given to the chief registrar under the Section 25(3) of the Elections (Regulation of Electors) Regulations 2002 where it is state that:
"Where an error has resulted in any person being registered as an elector of a registration area which is not a registration area in respect of which that person should have been registered, the Chief Registrar may enter that person's name in the principal electoral roll or the supplementary electoral roll for the appropriate registration area."
This was the reason given for why Selangor Menteri Besar Khalid Ibrahim's voting areawas shifted from Petaling Jaya Selatan (P105) to Lembah Pantai (P121) - Section 16/2 is part of Wilayah Persekutuan rather than Petaling Jaya.
The ‘correction of boundaries' according to the EC has affected 19,342 voters in the whole country.
I have three concerns with regard to the EC's rebuttal on this point.
Firstly, if the Egis system was introduced in 2004 after the 2003 delimitation exercise, why were these corrections not done until long after the 2008 general elections?
The following is the EC's statement in Bahasa Malaysia with regard to the introduction of the Egis system:
"SPR menyedari kedudukan beberapa lokaliti yang tersalah letak di dalam bahagian pilihan raya hanya selepas Sistem Egis (Electoral Geographical Information System) digunapakai dalam tahun 2004."
And again, when referring to the correction to the Section 16/2 locality and shifting it from PJ Selatan to Lembah Pantai:
"Apabila sistem Egis (Electoral Geographical Information System) dilaksanakan pada tahun 2004, SPR mula menyedari kesilapan penentuan kedudukan lokaliti berkenaan."
The fact that the EC only choose to make these boundary ‘corrections' more than four years (after the 2008 general election) after the introduction of Egis is very troubling and concerning.
Affected voters not informed
Secondly, what is more troubling is the fact that the EC did not notify the 19,342 voters which were affected by this exercise. They only wrote to the political areas in the areas which were affected.
Again, this raises concerns in regard to the intent of the EC to keep voters informed about their respective voting constituencies. Is it the responsibility of the politicians or is it the responsibility of the EC to keep voters informed of the changes made?
Most reasonable people would argue that the EC has the obligation to inform voters when they have been shifted from one voting constituency to another since this means that their political representative at the state (for the non-Federal Territories voters) and at the parliamentary levels have been changed.
The EC has since stated that it would send letters to each of these voters to inform them of these changes before the next general election. We hope that these letters will be sent in a timely fashion and that all of these affected voters will be duly notified.
Thirdly, the EC has not been consistent and transparent in this boundary ‘correction' exercise.
When they informed political representatives that localities have been shifted in and/or out of their respective constituencies, they failed to provide updated maps of the newly configured constituencies showing exactly where the new localities are placed.
This lack of transparency points to the possibility that the EC has not been transparent in terms of the standards used for these boundary ‘correction' exercises.
I shall firstly illustrate one case involving the movement of localities which does not make sense according to pre-existing electoral boundaries in the state seats of Seri Setia (Kelana Jaya parliament) and Kota Damansara (Subang).
Map 1 below highlights one of the areas affected by this boundary ‘correction' exercise in the state seat of Seri Setia (N32).
I have three concerns with regard to the EC's rebuttal on this point.
Firstly, if the Egis system was introduced in 2004 after the 2003 delimitation exercise, why were these corrections not done until long after the 2008 general elections?
The following is the EC's statement in Bahasa Malaysia with regard to the introduction of the Egis system:
"SPR menyedari kedudukan beberapa lokaliti yang tersalah letak di dalam bahagian pilihan raya hanya selepas Sistem Egis (Electoral Geographical Information System) digunapakai dalam tahun 2004."
And again, when referring to the correction to the Section 16/2 locality and shifting it from PJ Selatan to Lembah Pantai:
"Apabila sistem Egis (Electoral Geographical Information System) dilaksanakan pada tahun 2004, SPR mula menyedari kesilapan penentuan kedudukan lokaliti berkenaan."
The fact that the EC only choose to make these boundary ‘corrections' more than four years (after the 2008 general election) after the introduction of Egis is very troubling and concerning.
Affected voters not informed
Secondly, what is more troubling is the fact that the EC did not notify the 19,342 voters which were affected by this exercise. They only wrote to the political areas in the areas which were affected.
Again, this raises concerns in regard to the intent of the EC to keep voters informed about their respective voting constituencies. Is it the responsibility of the politicians or is it the responsibility of the EC to keep voters informed of the changes made?
Most reasonable people would argue that the EC has the obligation to inform voters when they have been shifted from one voting constituency to another since this means that their political representative at the state (for the non-Federal Territories voters) and at the parliamentary levels have been changed.
The EC has since stated that it would send letters to each of these voters to inform them of these changes before the next general election. We hope that these letters will be sent in a timely fashion and that all of these affected voters will be duly notified.
Thirdly, the EC has not been consistent and transparent in this boundary ‘correction' exercise.
When they informed political representatives that localities have been shifted in and/or out of their respective constituencies, they failed to provide updated maps of the newly configured constituencies showing exactly where the new localities are placed.
This lack of transparency points to the possibility that the EC has not been transparent in terms of the standards used for these boundary ‘correction' exercises.
I shall firstly illustrate one case involving the movement of localities which does not make sense according to pre-existing electoral boundaries in the state seats of Seri Setia (Kelana Jaya parliament) and Kota Damansara (Subang).
Map 1 below highlights one of the areas affected by this boundary ‘correction' exercise in the state seat of Seri Setia (N32).
Map 2 below highlights the localities which have been transferred from Seri Setia to Kota Damansara (N39).
The electoral boundary of Seri Setia is demarcated by the major highways highlighted in red. This is the pre-existing boundary which was used in the 2004 and 2008 general elections.
What is perplexing is that Saujana Resort, which lies on the southern side of the old Subang airport road (highlighted in red at the top of the map) has now been shifted to Kota Damansara, which lies on the northern side of the old Subang airport road.
I fail to see why Egis would show that this locality should be shifted to another constituency when it clearly falls within the boundary of the Seri Setia state seat.
What is also perplexing is that two localities that are to the east of the Jalan Monfort highway boundary (highlighted in red at the left of the map) - Pusat Bekalan Polis Shah Alam and Pangsapuri Melewar - have also been shifted to Kota Damansara.
If the EC had shifted these two localities to Kota Anggerik (N40), it could have at least been more defensible. It does not make sense for these two localities to be moved up to Kota Damansara when the Taman Glenmarie locality, which is closer to the Kota Damansara seat, remains in Seri Setia.
This is just one example of how the boundary ‘correction' exercise done by the EC does not seem to be using consistent and transparent standards which includes using natural boundaries like rivers and other man-made boundaries such as major roads to demarcate constituencies.
Another example which shows the EC inconsistency in adopting a consistent set of standards in its boundary ‘correction' exercise is the inclusion of a locality in Bukit Jalil, which should be under the Bandar Tun Razak parliamentary seat, not the Lembah Pantai parliamentary seat.
A new locality called POLIS (Pusat Tahanan Sementara Bukit Jalil - Kod Lokaliti: 1210017701) was created in 2011. As its name indicates, this police holding facility is located in Bukit Jalil.
But the almost 500 police postal voters in this locality were allocated to the Lembah Pantai parliamentary constituency, a marginal constituency which is likely to be contested by Federal Territories Minister Raja Nong Chik.
The EC would be hard pressed to explain why its boundary ‘correction' exercise failed to put this locality into the Bandar Tun Razak parliamentary constituency, especially since it falls outside the current boundaries of the Lembah Paintai parliamentary seat.
To ensure that this boundary ‘correction' exercise is not a stealth delimitation exercise, the EC should publish maps of the areas affected indicating clearly the localities which have been shifted as well as why they have been shifted.
42,051 doubtful voters in electoral roll
After initially promising to erase 42,051 voters whose information and status was not clear according to the National Registration Department (NRD) records, the EC retracted on this and said that they had no right to remove these names even though Section 25 2(g) of the Elections (Regulation of Electors) Regulations 2002 states that the chief registrar may "strik(e) out the name of any person who, in the opinion of the Chief Registrar, is not entitled, for any reason, to remain or be in the principal electoral roll or the supplementary roll".
This figure was later revised downwards by 1,248 voters to 40,803 when these names publicly displayed and the NRD was able to update their records accordingly.
However, the EC has failed to account for why 14,577 names were deleted in the Quarter 2 2011 because their records were not active in NRD - "rekod tidak aktif di JPN (NRD)". Not that these names were not deleted because the voters in question had died - that figure was reported in a separate category.
Hence, if in one quarter, the names of almost 15,000 voters were deleted because their NRD records were not active, why was it the case that over 42,000 voters whose status was not identifiable by the NRD in another quarter not dealt with in the same manner?
Again, this calls into question the EC's consistency in applying the same standards for the same cases.
What is also disconcerting is the fact that a very high percentage of the voters among the 42,051 not deleted as well as the 14,577 deleted voters were very old voters. For example, 13,183 or 90.4 percent of the 14,577 deleted voters were over the age of 70. Similarly, 33,421 or 79.5 percent of the 42,051 not deleted voters were above the age of 70.
It also raises questions about the utility of publicly displaying these records when most of the voters among the 42,051 are probably not very mobile and would not be able to ‘verify' the fact that they are still very much alive. And if they are dead and do not have family who happen upon this list, then there is no way that their names can be removed from the electoral roll.
It would be much more effective for the EC to form investigative teams, together with NRD, to try to track down as many of these voters as possible instead of just relying on public displays and relying on politicians and political parties to track down these voters.
As long as the EC is not pro-active in ‘going to the ground', the list of ‘dubious' voters, including those who are aged over 100 years old, will continue to populate the electoral roll. The identity of these voters can easily be used by irresponsible parties to cast a vote.
The electoral boundary of Seri Setia is demarcated by the major highways highlighted in red. This is the pre-existing boundary which was used in the 2004 and 2008 general elections.
What is perplexing is that Saujana Resort, which lies on the southern side of the old Subang airport road (highlighted in red at the top of the map) has now been shifted to Kota Damansara, which lies on the northern side of the old Subang airport road.
I fail to see why Egis would show that this locality should be shifted to another constituency when it clearly falls within the boundary of the Seri Setia state seat.
What is also perplexing is that two localities that are to the east of the Jalan Monfort highway boundary (highlighted in red at the left of the map) - Pusat Bekalan Polis Shah Alam and Pangsapuri Melewar - have also been shifted to Kota Damansara.
If the EC had shifted these two localities to Kota Anggerik (N40), it could have at least been more defensible. It does not make sense for these two localities to be moved up to Kota Damansara when the Taman Glenmarie locality, which is closer to the Kota Damansara seat, remains in Seri Setia.
This is just one example of how the boundary ‘correction' exercise done by the EC does not seem to be using consistent and transparent standards which includes using natural boundaries like rivers and other man-made boundaries such as major roads to demarcate constituencies.
Another example which shows the EC inconsistency in adopting a consistent set of standards in its boundary ‘correction' exercise is the inclusion of a locality in Bukit Jalil, which should be under the Bandar Tun Razak parliamentary seat, not the Lembah Pantai parliamentary seat.
A new locality called POLIS (Pusat Tahanan Sementara Bukit Jalil - Kod Lokaliti: 1210017701) was created in 2011. As its name indicates, this police holding facility is located in Bukit Jalil.
But the almost 500 police postal voters in this locality were allocated to the Lembah Pantai parliamentary constituency, a marginal constituency which is likely to be contested by Federal Territories Minister Raja Nong Chik.
The EC would be hard pressed to explain why its boundary ‘correction' exercise failed to put this locality into the Bandar Tun Razak parliamentary constituency, especially since it falls outside the current boundaries of the Lembah Paintai parliamentary seat.
To ensure that this boundary ‘correction' exercise is not a stealth delimitation exercise, the EC should publish maps of the areas affected indicating clearly the localities which have been shifted as well as why they have been shifted.
42,051 doubtful voters in electoral roll
After initially promising to erase 42,051 voters whose information and status was not clear according to the National Registration Department (NRD) records, the EC retracted on this and said that they had no right to remove these names even though Section 25 2(g) of the Elections (Regulation of Electors) Regulations 2002 states that the chief registrar may "strik(e) out the name of any person who, in the opinion of the Chief Registrar, is not entitled, for any reason, to remain or be in the principal electoral roll or the supplementary roll".
This figure was later revised downwards by 1,248 voters to 40,803 when these names publicly displayed and the NRD was able to update their records accordingly.
However, the EC has failed to account for why 14,577 names were deleted in the Quarter 2 2011 because their records were not active in NRD - "rekod tidak aktif di JPN (NRD)". Not that these names were not deleted because the voters in question had died - that figure was reported in a separate category.
Hence, if in one quarter, the names of almost 15,000 voters were deleted because their NRD records were not active, why was it the case that over 42,000 voters whose status was not identifiable by the NRD in another quarter not dealt with in the same manner?
Again, this calls into question the EC's consistency in applying the same standards for the same cases.
What is also disconcerting is the fact that a very high percentage of the voters among the 42,051 not deleted as well as the 14,577 deleted voters were very old voters. For example, 13,183 or 90.4 percent of the 14,577 deleted voters were over the age of 70. Similarly, 33,421 or 79.5 percent of the 42,051 not deleted voters were above the age of 70.
It also raises questions about the utility of publicly displaying these records when most of the voters among the 42,051 are probably not very mobile and would not be able to ‘verify' the fact that they are still very much alive. And if they are dead and do not have family who happen upon this list, then there is no way that their names can be removed from the electoral roll.
It would be much more effective for the EC to form investigative teams, together with NRD, to try to track down as many of these voters as possible instead of just relying on public displays and relying on politicians and political parties to track down these voters.
As long as the EC is not pro-active in ‘going to the ground', the list of ‘dubious' voters, including those who are aged over 100 years old, will continue to populate the electoral roll. The identity of these voters can easily be used by irresponsible parties to cast a vote.
Dr ONG KIAN MING is project director of the Malaysian Electoral Roll Analysis Project (Merap).
Total building maintenance:-
ReplyDeletejanitorial cleaning services
Total building maintenance:-
ReplyDeletejanitorial cleaning services