685,000 Voters Would Give Barisan Alternatif 12 More Seats

· The Barisan Alternatif would add 12 parliamentary seats had the 685,000 newly registered voters been allowed to vote in the 29 Nov 1999 general election.

· Five Opposition heavy-weights would have entered the Parliament, and one Cabinet Minister lost.

· While the 0.685m represent only 7% increase to the 9.6m eligible voters, they can increase Opposition seats by 27% from 45 seats to 57, or from 23% of the parliament to 30% of the parliament.

· A common-sense estimate supports the assumptions that 80% of these "ineligible" voters would vote for the BA. This study also fully discounts for actual turnout and spoiled vote ratios.

· A 90% preference by the "ineligible" would have enabled BA to deny BN's 2/3 majority.

· The results compel the BA and the NGOs to stress a future strategy of new voter education and registration, because they will make a big difference.

Which parliamentary seats are affected? By how much?

Winners and losers of parliamentary seats - with increasing % preference by "ineligible" voters for BA:

% by			Seat	Official winning candidates	Coali-	Official	Adj.	Potentially real	Coali-
685k	Seat name	State	Code	- Potential losers	Tion	Majorty	Majorty	Winner	Tion2
65%	Bukit Bendera	Penang	05P47	Chia Kwang Chye	BN	104	(937)	Lim Kit Siang	BA
	Jelutong	Penang	05P49	Lee Kah Choon	BN	775	(132)	Karpal Singh a/l Ram Singh	BA
	Lumut	Perak	06P71	Kong Cho Ha	BN	605	(270)	Zaman Huri Haji Samsudin	BA
	Pekan	Pahang	07P80	Mohd Najib bin Hj Abd Razak	BN	241	(319)	Ramli Mohamed	BA
	Temerluh	Pahang	07P84	Mohd Sarit bin Hj Jusoh	BN	213	(615)	Ab Wahab bin Ismail	BA
	Gombak	Selangor	08P91	Zaleha bt Ismail	BN	803	(387)	Dr Mohd Hatta Md Ramli	BA
	Shah Alam	Selangor	08P98	Dato Ir Hj Mohd Zin Mohamed	BN	1,440	(69)	Ezam Mohd Nor	BA
70%	Bandar Tun Razak	FedTerr	09P112	Tan Chai Ho	BN	1,224	(16)	Chandra Muzaffar	BA
	Rasah	Nsembilan	11P118	Datuk Goh Siow Huat	BN	1,288	(321)	Chen Man Hin Dennis	BA
80%	Sabak Bernam	Selangor	08P86	Dato Hj Zainal Dahalan	BN	901	(75)	Dr Ghazali Basri	BA
	Titiwangsa	FedTerr	09P107	Dato Dr Suleiman Mohamed	BN	1,513	(21)	Mohamad Noor Mohamad	BA
	Lembah Pantai	FedTerr	09P109	Shahrizat binti Abdul Jalil	BN	1,417	(128)	Zainur Zakaria	BA
85%	Ipoh Timor	Perak	06P61	Thong Fah Chong	BN	2,127	(321)	Ngeh Koo Ham	BA
	Jerantut	Pahang	07P77	Tengku Azlan Ibni Sultan Abu Bakar	BN	1,463	(111)	Syed Ibrahim bin Syed Abd Rahman	BA
	Bintulu	Sarawak	15P190	Tiong King Sing	BN	1,400	(103)	Chiew Chiu Sing	BA
90%	Arau	Perlis	01P3	Mastika Junaidah Husin	BN	1,586	(77)	Abd Aziz Hanafi	BA
	Bagan Serai	Perak	06P55	Zainal Abidin Zin	BN	1,584	(54)	Ahmad bin Awang	BA
	Beruas	Perak	06P65	Lim Kheng Heang @ Lim Keng Yaik	BN	1,455	(122)	Yew Teong Chong @ Iu Teong Chong	BA
	Pj Utara	Selangor	08P94	Chew Mei Fun	BN	2,481	(326)	Ronnie Liu Tian Khiew	BA
	Kapar	Selangor	08P99	P. Komala Devi	BN	2,860	(285)	Dr Dzulkefly Ahmad	BA
	Batu	FedTerr	09P104	Ng Lip Yong	BN	2,297	(165)	Sanusi Osman	BA
95%	Serdang	Selangor	08P96	Dato Yap Pian Hon	BN	4,099	(430)	Dr Wong Ang Peng	BA
100%	Tanjung Karang	Selangor	08P87	Haji Noh Haji Omar	BN	2,075	(46)	Dr Md Yusoff Abd Wahab	BA
	Total swing up to		23	OPP total seat would be	68

The table shows how many more MPs the BA would gain, who, where, and by how much adjusted majority if the 685,000 prefer the BA by a progressive range of 65%-100%.

· In the base case assumption that 80% of the 685,000 prefer the BA, the BA would gain 12 more seats.

· At 90% threshold, the BN's 2/3 majority would be broken as the BA gains 21 on top of its 45.

· Had they voted 100% for the BA, the increase for the BA would be 23 seats.

Coalition	BA and Opp	BN	Total	BA in parliament	BA increase
Current seat	45	148	193	23%
Adjusted seat	57	136	193	30%	27%

· Even at a low 65% preference, there would be major seat "turnovers" - three Opposition leaders would have been returned to the parliament, namely Lim Kit Siang, Karpal Singh, and Ezam Mohn Nor.

· At 70%, another two Opposition heavy weights would be included: Chandra and Chen Man Hin.

· Two Cabinet Ministers' seats are at risk. At 60% "ineligible" preference for BA, Najib would be out of the parliament. At 90%, Lim Keng Yaik would be out.

It is an eerie observation that just one electoral issue could turn things around for many opposition leaders. A suspicious mind would have imagined an invisible hand taking every excuse to slice the opposition margin so desperately the 685,000 disenfranchisement was but one of many underhanded practices.

Why is this important? Who should act?

· 685,000 registered for 7 months but still could not vote in the MSC era.

· This represents only one of many potential election abuses, yet this analysis shows it alone is already a significant swing factor, and should be a grave concern to the nation.

· Had all the alleged election abuses been dealt with, the combined impact on the election outcome could have been decisively different.

· The people may have been improperly represented in the parliament by 12 MPs,

· The immediate arbitrageur must be the judiciary because the integrity of the Parliament is in question. The court must take action to resolve any complaints, such as Dr. Jomo's suit against the Election Commission. And if there is any remaining doubt regarding the assumption about 80% of ineligible voting for the BA, then let them speak for themselves: Let them vote. That is the whole point.

· The ultimate arbitrageur, however, must be the people, who should take these alleged abuses into account in the next election to ensure they vote not only for their preferred candidates, but also to counterbalance and correct any abuses.

· It will be worthwhile for the BA to stress a future strategy of new voter education and registration, because they will make a big difference.

· NGOs have a strong basis to pressure the Election Commission on transparency and accountability.

What are the assumptions and data? How is this done?

· The original data contain for each parliamentary seat: the constituency name (eg, Padang Besar), code (eg, 01P1), state (eg, Perlis), candidate names (eg, Azmi bin Khalid), Party, Coalition, Voter, Turnout, Spoiled Votes, Vote Won, Majority Vote Count, which are all used in the analysis. The file contains other details not used in this analysis, and they include Deposit Forfeit, previous election results (Winning Candidate, Party, Coalition, Voters, Majority), race composition (Malay, Chinese, Indian, Others), Chinese names if available, and brief economic/employment description of each constituency.

· In this analysis, the first assumption is that 685,000 potential voters would be spread into each parliamentary constituency in the proportion of their individual voter bases.

· For example, seat 01P1, Padang Besar of Perlis has 32,478 voters, or 0.34% of whole-nation voter base of 9,566,188. Hence 0.34% of 685,000, or 2,326 are allocated to this constituency.

· The second assumption is that the same turnout ratio and spoiled vote ratio for each constituency from the 29/11/99 actual election results would apply to the 685,000 potential voters.

· For example, turnout ratio for 01P1 was 74.7%, and spoiled vote ratio was 0% (an exceptionally low figure). In another example, spoiled vote ratio for 01P2 was 1.6%, closer to national average of 2.0%.

· The 685,000 voters allocated to each constituency, such as the 2,326 for 01P1 (Padang Besar), are discounted by the individual constituency's turnout ratio and spoiled vote ratio, yielding for 01P1, eg, 1,737 "potential voters, adjusted for turnout and spoiled votes."

Summary of overall voting statistics

	Total					Spoiled/
	seats	Voter	Turnout	Turnout%	Spoiled	Turnout
West Mal, KL and Labuan total	145	7,992,125	5,835,301	73.0%	124,635	2.1%
East Mal total	48	1,574,063	961,954	61.1%	13,621	1.4%
Mal total	193	9,566,188	6,797,255	71.1%	138,256	2.0%

· The third assumption is key. In the Base Case, 80% of these largely young voters allocated to each constituency (discounted for turnout and spoiled votes) are assumed to vote in favour of the BA. Had the split been 50-50%, it would make no difference to the results.

· This 80% assumption is well-supported by a common-sense estimate which yields a range of 76-84% preference for the BA among the "registered but ineligible" voters. This is based on the further assumptions that:

(a) 200,000 of the 685,000 represent the normal, "yearly harvest" of garden-variety voters, and are likely to vote in line with the actual pattern of the 1999 election, ie, 56.5% BN : 43.5% BA, or, 113,000 BN and 87,000 BA.

(b) the remaining 485,000 would represent the "newly-aware" voters agitated by the recent political events, and are likely to vote in favour of the BA by 90-100%, or 434,000-485,000.

(c) hence among the 685,000 ineligibles, 521,000 (87,000+434,000) up to 572,000 (87,000+485,000) are likely to vote for the BA, representing 76-84% of 685,000.

(d) This estimation only establishes the credibility of the 80% base-case assumption. Actual calculation still runs by details for each constituency, discounted for turnout and spoiled votes.

· Example, in the Base Case where 80% of these "ineligible" registered voters prefer the BA, the discounted 1,737 of Padang Besar are split into 347 estimated to prefer the BN and 1,390 for BA.

· Majority vote count is used/adjusted to simplify calculation. The 4,519 majority vote won by Azmi bin Khalid (BN) in Padang Besar is increased by 347 and reduced by 1,390, yielding an adjusted - and reduced - majority of 3,477.

· Not so lucky for 12 other BN candidates who officially won the election - if 80% of ineligible voters are able to vote for the BA, as shown on the first page. Five heavy-weight opposition leaders would be returned into the parliament. (Note the Adjusted Majority shown on the table on page 1 refers to the Adjusted Majority at the percentage of preference shown. At a higher percentage of preference by the ineligibles, the Adjusted Majority would have become larger. See next paragraph.)

· The first to cross the invisible hurdle would be Lim Kit Siang, who would have won if only 52% of the allocated ineligible voted for the BA, turning a 104 vote loss to 35 vote majority. For illustration, at 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 100% "ineligible" voter preferences, his majority would have grown to 243, 590, 937, 1284, 1630, 1977, 2324, 2671, 3018, and 3365.

· There is a what-if table showing scenarios ranging from 50% to 100% preference.

Are the assumptions conservative or aggressive? What might go wrong?

· Admittedly, the greatest argument will surround the assumed percentage (80% in Base Case) of the registered-but-ineligible voters who would vote for the Barisan Alternatif.

· Definitive answer can only come by putting the question to a vote by these disenfranchised. The ultimate test of this assumption is a vote. Which is why the law suit for 685,000 to vote is crucial.

· There is also high likelihood that the 485,000 "newly-aware" voters are predominantly rural, small, constituencies. This means in reality, the 485,000 will have greater impact on these small constituencies than allowed by this analysis. In other words, the proportional spread of 685,000 by each constituency's voters' weight is too conservative compared to the likely, real-life, skewing of "newly-aware voters" towards the small and easily-swung rural constituencies.

· If the 685,000 are spread evenly into 193 constituencies, the average is still a large 3,549. There are 55 constituencies where the majority vote counts are less than 3,549 (but of course not all would turnout out to cast unspoiled votes for the BA).

· The use of actual turnout and spoiled vote ratios are conservative because the presumably enthusiastic, newly-aware voters are likely to yield higher turnout ratios and lower protest-type spoiled votes.

· On this matter, the what-if scenarios test if the turnout+unspoiled votes are higher or lower by 5% of actual. The results are meaningful but minor.
Where do the data come from? Can I see them?

The sources are publicly available as of mid December 1999.

· Most of the data came from Star newspaper's website: http://polls.thestar.com.my/results/results.html.

· An example page of details is http://polls.thestar.com.my/results/01/01p.html

· Some confirmation and Chinese name data are sourced from Kwong Wah Yit Poh's site at http://www.kwongwah.com.my/ge/welcome.htm

· Original data are on multiple pages, then extracted and organised into Excel format.

· Three Excel files are available upon request with links available to check the calculation. The first contains this analysis that supports this MsWord document; the initial cut includes only parliamentary data (ineligible.xls, zipped 0.17MB). The second contains full parliamentary and state data, original webpages and Excel formula linkages (ge1999.xls, zipped 1.7MB, large). A third is a summary, with value-only, no-link, Excel file, with full parliamentary and state election results (ge1999-s.xls, zipped 0.2MB). But due to the large size to email, not all requests could be responded immediately.

What if this and that?

How many more seats would the BA and opposition have gained?

… depends on the percentage of the 685,000 who would vote in favour of the BA. The What-if table also allows a variation of 5% on turnout-plus-unspoiled votes. For example, if 85% of 685,000 voted for BA and turnout+unspoiled vote is higher than 29/11 by 5%, then the BA would gain 19 additional seats.

Additional seats for BA/OPP
		Assumed non-eligible voting for Opp (after adjusting for turnout and spoiled votes)
If turnout had		50%	67%	75%	80%	85%	90%	95%	100%
been lower	-5%	0	7	9	11	14	18	22	22
or	0%	0	8	9	12	15	21	22	23
higher by this %	5%	0	8	9	13	19	22	22	24

What would be the total seats held by BA and other Opposition parties?

A total of 57 rather than 45 in the Base Case of 80% preference by 685,000. But the maximum could have been 69 if all the 685,000 favours BA and the turnout is high enough.

Adjusted total number of BA/OPP Seats
		Assumed non-eligible voting for Opp (after adjusting for turnout and spoiled votes)
If turnout had	54	50%	67%	75%	80%	85%	90%	95%	100%
been lower	-5%	45	52	54	56	59	63	67	67
or	0%	45	53	54	57	60	66	67	68
higher by this %	5%	45	53	54	58	64	67	67	69

At what point would BN's 2/3 majority be denied?

When preference for BA is 90% or greater by these young disenfranchised. Bear in mind the 685,000 registered-but-ineligible voters represent only one of several questionable issues regarding the electoral process. Even as a stand-alone issue, it harbours a real potential for denying the BN its 2/3 majority.

BN 2/3 majority denied?
		Assumed non-eligible voting for Opp (after adjusting for turnout and spoiled votes)
If turnout had		50%	67%	75%	80%	85%	90%	95%	100%
been lower	-5%							DENIED	DENIED
or	0%						DENIED	DENIED	DENIED
higher by this %	5%						DENIED	DENIED	DENIED

Cheah Kah Seng, 03 Jan 2000 ineligible.doc, Supported by ineligible.xls