Malaysia’s General Election 2008: a comprehensive analysis

Welcome to this first article on my blog.This blog is mainly focused on analysis of political events in Malaysia, especially the recently concluded 12th General Election results. My analytical approach is very comprehensive and firmly based on empirical evidence from publicly available statistical data collected before and after the recent election. My early qualifications, namely B. Econ. and MBA followed by deep interest over the years to test all kind of data in the fields of finance and politics has enabled me to build up a reasonably large data base which are handy to analyze to enable me to get some interesting insights into the current political or financial events.

We have heard a lot of comments on the outcome of the 12th general election over the television and the printed media. Blames were thrown on the Barisan National leadership over the election debacle. Many issues were cited, such as rises in prices of consumer goods, fuel and houses as well as political issues stirred by hindraf and issues on the integrity of the judiciary. There has been a lot of guessing works on the impact of these issues after the election, though Barisan National had actually grossly underestimated these issues. Moreover, other issues later cropped up regarding the choice of election candidates, in particular, dropping of many incumbents in favour of new faces

I have found it very exciting to analyse these issues in a scientific and objective way using a multiple regression model that accomodate many of these factors. The results that I derive, I believe will convincingly answer many questions that we all have been asking. For ease of understanding this article, I will first dwell into the questions, issues and the relevant answers, while leaving the explanations on the regression model at a later stage. It is hoped that this article will generate enough interest among the public at large as well as among the academicians, whom I hope will provide fair and reasonable feedback, comments and constructive criticism, which can hopefully improve our understanding of the impact of election issues on the political, and hence, economic stability of our beloved nation. The regression findings are shown below.

Finding No. 1. How did the Malays vote in the 2008 general election, for or against Barisan National and UMNO, and at what percentage. This issue is critical since Malay votes form the majority in the Peninsular Malaysia. Continued support by the Malays means continued Strength of Barisan National to form a stable government. In my finding, the majority of Malay votes in Peninsular Malaysia, ie 52% were specifically identified as cast for BN. This is a drop of 10 percentage points from the 2004 general election when 62% of Malays identified as voting for Barisan National. On the surface, the shift in Malay votes was less drastic compared to the shift in the chinese and Indian votes. However the actual percentage level of 52% is still critical unless BN is able to excute a correct strategy to improve support among the Malays, while at the same time regaining confidencce and support of other races. Alternatively, ultras in UMNO may prefer to drum up support focusing on racial and religious issues with a view to regain the votes that have deserted, including from PAS. PAS has to realise that, being with the Pakatan Rakyat, it cannot claim to be more religious than UMNO as, like UMNO, they now cannot ignore non muslim issues and demands. Of course, the first strategy is the best option to ensure a stable government supported by all races.

Finding No .2. Chinese votes shifted more than Malay votes, as in the 2008 election 38.1% of Chinese votes could be identified for BN, mainly MCA, compared with 56.4% in the 2004 general election. This is a shift of 18.3 percentage point away from Barisan national. I feel Chinese voting pattern, as usual, is largely a seasonal factor. Obviously, it was affected by the prevailing anti government sentiment at that time. Moreover, should Pakatan Rakyat win,the Chinese based party could play a pivotal role in the Pakatan Rakyat government.

However, this will not be advantageous to the Chinese community, since UMNO and Bumiputera parties can in fact still form the government with nominal representation of the Chinese and Indian communities, leaving both communities in the opposition. Such a scenario is likely o be viewed by the Malays as a challenge to Malay political dominance. This will easily play into the hands of Malay ultras who will drum up racial issues, at the detriment of PKR and PAS, in an emergency general election should there be one.

Finding No 3. The biggest shift in votes, occured as expected, among the Indian community. A very low percentage of Indian voters of approximately 8.3% probably voted for BN, mainly MIC in 2008 election compared with 72.4% in 2004 general election(even such a low 2008 figure could not be verified in this analysis as indicated in the model below). Therefore, I feel it is now of utmost importance for the government to address this issue. As for the Indian community, they may still be wondering how their votes could have this big an impact on the election outcome. But this phenomenon ia not expected to bring much good to the Indian community as they have lost meaningful representaion in the government. Worst still, they don’t control the agenda of Pakatan Rakyat either.

Finding No.4. Another factor also determined the number of votes that BN obtained, namely the status of the BN candidate in his constituency, ie. whether he is a new candidate or whether he is an incumbent and for how many terms. My finding indicates that the votes BN received is positively correlated with this factor, generally the longer the candidate served as YB in his constituency, the more votes he could bring in. However I could not isolate the votes he obtained by race, and further, such votes are usually not large and only impact the voting results in marginal constituencies. As a matter of illustration, for any constituency, a new BN candidate is expected to bring in 0 votes in his own right and 1200 votes if he has four terms as a YB.

Finding No. 5 The percentage of votes that a Barisan Nasional candidate obtained in the previous election has a very important positive impact on the number of votes that BN obtained in the 2008 election.This is very much expected since votes obtained in the previous election can be taken as a measure of voters loyalty to a political party. My regression model however could not segregate the votes racewise, hence figures obtained are in the form of votes without racial breakdown. Again as a matter of illustration, where BN had obtained 65% votes in previous election, this factor is expected to account for 20,000 BN votes in the 2008 election.

Finding No.6. About 19% of votes that BN obtained in 2008 is classified as error terms which could not be explained by the statistical model derived in the analysis.

Satistical Analysis

For the benefits of those familiar and skilled in statistical analysis, I have based the above findings on empirical testing of a hypothesis that the number of votes obtained by Barisan National in Peninsular Malaysia is dependent on a few variables, ie. the relative composition of Malay,Chinese and Indian voters, plus two other variables namely the popularity of the candidate and the traditional strength of BN in a particular constituency. Fairly large samples were taken from Parliamentary constituencies in Peninsular Malaysia both for the 2008 and 2004 elections.Multiple regression analysis were tested on a few models, and the model that best fit the data was selected.

Update 17 September 2009

Small errors were discovered in the data entries of the results for 2008 General Election. A revised computer analysis was done and this gave a slightly more accurate result as shown below:

The above results show that 54.4% of the Malays, 39.5% of the Chinese and and 10.4% of the Indian voters voted for Barisan Nasional in the 2008 General Election.

DAH IKHWAN

Note: While I do not claim full accuracy, I am confident that these are close estimates of the actual results.
Copyrights (c) by DAH Ikhwan. This article or any part of it may be reproduced or quoted, but due credit must be given to the original author DAH Ikhwan.

Reminder:[Comments are welcome, but please refrain from making libelous, seditious or defamatory comments]