The revelation earlier in September that the seat allocation talks between DAP and PKR have 'hit a snag' should not surprise those who are familiar with the process.

It should certainly not surprise those who were aware of the number of multi-corner fights featuring both parties (as well as DAP and PAS) in the 2004 general election.

The difficulties faced by DAP and PKR in agreeing to the allocation of parliamentary and state seats illustrate the different challenges faced by both parties in terms of defining its own political identity, mapping out its political future and balancing short-term and long-term electoral goals.

This article is divided into two parts. The first examines the underlying attitudes and preconceptions among DAP and PKR leaders as they come to the negotiating table and then examines the validity of these mostly electoral preconceptions in light of the 2004 general election results.

Before examining the process of seat negotiation and allocation in detail, one needs to understand the psyche of both political parties and the baggage they bring before they enter into these talks. Let me play the role of the fly on the wall listening to the thoughts and feelings of both DAP and PKR leaders as they discuss the issue of seat negotiation internally.

Imaginary DAP monologue

First of all, let's listen to the DAP leaders, dramatised and over-exaggerated, of course (or maybe not?):

Who do these guys (PKR) think they are? They expect to be treated as equals or even superior to us on the negotiating table but they are a party with only one parliament and one state seat1. We have 12 parliamentary seats and 21 state seats and we are the leader of the opposition in Parliament and in the state assembly in Pahang, Perak, Selangor, Negri Sembilan and Malacca.

We have been contesting in general elections since 1969. They only started contesting since 1999. And we have to kowtow to them?

They want to talk about us being selfish and not giving them enough seats to contest? We contested in 38 parliamentary seats in 1999 and only 37 in an expanded Parliament in 2004 in Peninsular Malaysia. They contested in 43 parliamentary seats in 1999 and 48 in 2004! How about state seats? We contested in 89 state seats in 1999 and 98 state seats in 2004 in Peninsular Malaysia. They contested in 67 state seats in 1999 and 103 state seats in 2004!

They were a floundering party in 2004. They had no direction, no leadership, no policy positions, very little grassroots support and were seen by many as a single issue party. And yet, they expect us to give them more seats?

To think that they actually thought that they could win the Ipoh Timor and Bukit Gelugor parliamentary seats in 2004! What gall! They had very little support among the Chinese voters and also performed badly in Malay majority seats. What makes them think that they still have popular electoral support, especially among the non-Malays?

They say they want to contest in Chinese majority seats but do they actually have the candidates to stand in these seats? Candidates who are respected and known by the Chinese community? Candidates who have worked the ground in these areas?

Imaginary PKR monologue

How then would a conversation between PKR leaders sound, equally embellished?

What makes them (the DAP) think that they have the right to contest in all the Chinese majority seats? Do they think that they are being 'generous' to us by giving us all the un-winnable mixed seats and keeping the winnable Chinese majority seats for themselves?

They say that we don't have the support of the Chinese community but didn't one of our Chinese candidates win a Chinese-majority state seat in the 2006 Sarawak state election? Did we not win around half of the popular vote in the Chinese-majority polling stations in the recent Ijok by-election?

They say that we don't have policy positions but did Anwar Ibrahim not time and time again reiterate his position that he's in favour of abolishing the NEP in favour of a policy that will benefit all poor and marginalised Malaysians regardless of race? They say we don't have leadership, but since Anwar's release from prison, has he not been an effective advocate for opposition causes? Wasn't he the reason why the Lingam tape scandal actually came to public light?

They say that they are a multi-ethnic party but they don't have any prominent Malay leaders whereas we do have many Chinese leaders including people like Tian Chua, Dr Lee Boon Chye and Cheah Kah Peng and Indian leaders like Sivarasah Rasiah. Only PAS and ourselves have elected candidates who are Malay and Chinese.

How can they expect to win some of the mixed seats they still want like Taiping and Beruas without Malay support? How do they expect to win back Kota Melaka without the support of the 30 percent of voters who are Malay? If they want the seats where the percentage of Chinese voters is over 60 percent, why don't we take some of their seats which are only barely Chinese majority since we can win some of the Chinese votes and more of the Malay votes?

Looking back

The imagined monologue/dialogue above highlights the main arguments and pre-conceived notions which both parties bring to the negotiation table. The DAP thinks that it has conceded enough ground to PKR especially in regard to the number of seats contested by both parties in 2004 compared to 1999. It is determined to hold on to the Chinese majority seats (more than 60 percent Chinese) which they think they have a better chance of winning compared to PKR.

It also thinks that it has staked out policy positions which are more attractive to non-Malay, especially Chinese, voters over the past year or two which will give the party an advantage in these seats in the upcoming general election. Finally, it thinks that it is capable of fielding better and more qualified candidates in these seats compared to PKR.

PKR feels that DAP's insistence on its inalienable right to contest in Chinese majority seats is without grounds, especially when PKR feels that it is a multi-ethnic party with many prominent non-Malay leaders who can contest in these seats.

PKR feels that Anwar's return to active politics has given the party a political and electoral boost and that his statements in regards to the NEP has strengthened it among the non-Malay, especially Chinese, electorate as evidenced by its electoral performance in the Chinese-majority polling stations in Ijok and in the Padungan state seat in the 2006 Sarawak state elections.

These arguments can be 'spun' by either side using selective evidence. So how can we ascertain the validity of some of these claims? One method is to use the 2004 general election results as an indicator of the electoral strength and popularity of the DAP and PKR. While electoral conditions have shifted significantly since the 2004 general election, it is still worthwhile to look at what the results tells us vis--vis their respective electoral strength.

If both DAP and PKR want to contest in Chinese majority seats, we should attempt to answer the following question: Given a choice between a non-Malay DAP and a non-Malay PKR candidate at the parliamentary and state levels, how did non-Malay voters vote? Did they prefer the non-Malay DAP candidate at the state level to the non-Malay PKR candidate at the parliamentary level or vice-versa? In other words, what does split ticket voting tendencies tell us about non-Malay preferences for DAP and PKR?

For this type of analysis, one needs to examine the polling station election returns. For example, one can compare the percentage of the popular vote won by a DAP candidate at the parliamentary level versus a PKR candidate at the state level or a PKR candidate at the parliamentary level versus a DAP candidate at the state level.

Using the 2004 general election polling station data which Dr Bridget Welsh and I have collected, there were three parliament-state configurations which featured a PKR non-Malay candidates at the parliamentary level and a DAP non-Malay candidate at the state level - namely PJ Selatan-Bukit Gasing, Kapar-Sungai Pinang and Gelang Patah-Skudai.

There were four parliament-state configurations which featured a DAP non-Malay candidate at the parliamentary level and a PKR non-Malay candidate at the state level - namely Bukit Mertajam-Machang Bubok, Bukit Bendera-Pulau Tikus, Rasah-Senawang and Kulai-Bukit Batu. (This was only for seats where we had complete polling station data at the parliamentary and state levels.)

Split ticket - parliamentary

Table 1 summarises the level of split ticket in these seven parliamentary-state configurations. What it tells us is that, for the three seats which featured PKR at the parliamentary level and the DAP at the state level, the BN candidate at the parliamentary level received more votes than the BN candidate at the state level. To put this in more concrete terms, we can examine the voting in the PJ Selatan-Bukit Gasing polling stations.

The BN won 72.9 percent of the popular vote in the polling stations in Bukit Gasing at the parliamentary level (excluding postal votes) while the BN only won 57.8 percent of the popular vote in the polling stations at the state level (excluding postal votes), giving us a difference of 15.1 percent in favor of the BN at the parliamentary level. In absolute terms, a 15.1 percent difference translates into roughly 3,300 voters in a state constituency with approximately 22,000 valid votes in the 2004 general election.

For the seats with featured DAP at the parliamentary level and PKR at the state level, the BN won fewer votes at the parliamentary level compared to the state level (hence, the negative values), again showing the preference for the DAP candidate over the PKR candidate.

Interpreting split ticket voting in Chinese majority seats as voters preferring one party over another can be tricky because of the tendency of Chinese voters to prefer parliamentary opposition over state opposition, even if it is the same party contesting at both levels as is the case for Chinese majority seats in Penang. Hence, it might be inaccurate to conclude the Chinese voters preferred the DAP at the parliamentary level compared to PKR at the state level when Chinese voters might have preferred DAP at the parliamentary level compared to DAP at the state level.

While this criticism is accurate, the results from the configurations featuring PKR at the parliamentary level and DAP at the state level does conclusively show that Chinese voters prefer DAP over PKR as shown by their willingness to split votes for the DAP candidate at the state level even though Chinese voters value opposition at the parliamentary level more.

For the seats featuring DAP at the parliamentary level and PKR at the state level, especially those in Penang, we need to compare the level of split ticket voting in the other state seats in Bukit Mertajam and Bukit Bendera, which featured DAP at the parliamentary and state level. If Chinese voters prefer DAP over PKR, we would expect the level of split ticket voting in favor of the opposition to be lower in state seats featuring DAP candidates compared to state seats featuring PKR candidates.

Split-ticket - state

Table 2 shows the level of split ticket voting in the state seats in Bukit Mertajam and Bukit Bendera.

We would expect the figures in brackets to be negative meaning that the voters split their votes in favor of the opposition at the parliamentary level. In Bukit Bendera, the level of split ticket voting was lower in the three state seats featuring DAP candidates (Tanjong Bungah, Air Putih and Kebun Bunga) compared to the state seat featuring the PKR candidate (Pulau Tikus). This means that given the same DAP parliament candidate, more Chinese voters preferred the DAP candidates at the state level compared to the PKR candidate at the state level.

In the Padang Lallang state seat in Bukit Mertajam, the level of split ticket voting was lower than in Machang Bubok, again showing that more Chinese voters in Padang Lallang preferred the DAP candidate compared to the Chinese voters in Machang Bubok who were given the choice of voting for a PKR candidate at the state level.

The level of split ticket voting in Berapit was actually higher than that in Machang Bubok but this is the only state seat where split ticket voting trends were not favorable to DAP. So, by comparing state seats in these two Chinese majority parliamentary seats of Bukit Mertajam and Bukit Bendera, we can show that there is a strong preference among Chinese voters for DAP over PKR in five out of the six state seats in these two areas.

That Chinese voters prefer DAP over PKR should not surprise anyone given DAP's long- standing history of advocating for non-Malay rights and issues compared to PKR's more ambiguous stand, at least during the 2004 general election, towards issues of importance to the non-Malay community and its shaky credibility in the eyes of the non-Malay community.

Equally unsurprising would be the preference of Malay voters for a Malay PKR candidate over a Chinese DAP candidate, given DAP's poor reputation among the Malay community. This is clearly shown in parliamentary seats which feature Chinese DAP candidates and state seats which feature Malay PKR candidates.

Data is available for polling stations in two such seats - polling stations in the Pengkalan Baharu state seat in the Beruas parliamentary constituency and for the polling stations in the Tualang Sekah state seat in the Kampar parliamentary constituency. Both Beruas (55 percent Chinese) and Kampar (64 percent Chinese) are Chinese majority seats but the Pengkalan Baharu (64 percent Malay) and Tualang Sekah (57 percent Malay) state seats within these two parliamentary seats are Malay-majority state seats.

After calculating the results at the parliamentary and state levels in these two state seats, it was found that the BN candidate at the parliamentary level received 8.5 percent and 10.5 percent more votes respectively than the BN candidates at the state seats of Pengkalan Baharu and Tualang Sekah, which means that the Malay voters in these two state seats preferred the Malay PKR candidate at the state level over the Chinese DAP candidate at the parliamentary level.

So far we've seen Chinese voters splitting their votes in favor of Chinese DAP candidates (over Chinese PKR candidates) and Malay voters splitting their votes in favor of Malay PKR candidates (over Chinese DAP candidates). Are there any instances where Malay voters split their votes in favour of Chinese PKR candidates?

This is important since it underlies a possible electoral strategy for PKR in that by fielding non-Malay candidates in mixed seats, it can hope to gain some proportion of non-Malay support which a Malay candidate would not otherwise obtain and at the same time, it can hope to gain some proportion of the Malay support which a non-Malay DAP candidate would not otherwise obtain.

As it turns out, there was such a case in the 2004 general election and it involved three state seats in the parliamentary seat of Kapar. This was a Malay plurality seat (49 percent Malay) in the 2004 general election and it featured an MIC candidate and a Chinese PKR candidate at the parliamentary level.

Three of its state seats, two of which were Malay majority seats Meru (59 percent Malay), and Selat Klang (61 percent Malay) - and one of which was a Malay plurality seat (Sementa, 49 percent Malay) - featured contests between Malay candidates: Meru (Umno vs PAS), Sementa (Umno vs PKR) and Selat Klang (Umno vs PAS).

Again, using polling station results at the parliamentary and state levels for these seats, it was found that split voting in these Malay-majority/plurality seats actually favoured the Chinese PKR candidate at the parliamentary level. The difference between the share of popular support won by the BN candidate at the parliamentary level and at the state level in Meru, Sementa and Selat Klang were 6.5 percent, 13.4 percent and 1.1 percent respectively, all in favour of the opposition candidate at the parliamentary level.

What this tells us it that there were a significant number of voters in these Malay-majority constituencies who voted for Umno at the state level but switched their vote to the Chinese PKR candidate over the MIC candidate at the parliamentary level.

Before we go on, a note of caution in interpreting these results is necessary. The data shown here demonstrates the tendency of different voters to switch their votes between the BN and the opposition at the parliamentary and state levels depending on the parties contesting at each level. The tendency of voters in some seats to split their votes is not a guarantee that the beneficiaries, on the part of the opposition, will be able to win seats as a result of split voting.

For example, an opposition party which receives 30 percent of the popular vote at the state level but 40 percent at the state level will still not be able to win the state seat in question. The only place historically where split voting has enabled opposition candidates to win is in Chinese majority parliament seats in Penang.

Impact on electoral strategy

What do these results tell us with regard to the optimum short term strategy which DAP and PKR should take in terms of allocating seats between these two parties? The evidence presented thus far shows that all things being equal, the DAP stands to win a much larger share of the popular vote in Chinese majority seats compared to PKR. Purely from an electoral standpoint i.e. ability to win seats, it makes sense for the DAP to contest in all the Chinese majority seats.

The evidence also seems to indicate that PKR stands to win a higher share of the popular vote in mixed seats (where no one race comprises of more than 60 percent of the electorate). The question in these mixed seats becomes - how much of the Chinese vote would be lost if a Chinese PKR candidate were to be fielded instead of a Chinese DAP candidate and how much of the Malay vote would be gained from such a move?

Evidence also seem to suggest that any electoral gains made by such a move would not be sufficient for the opposition to win these mixed seats, a traditional stronghold area for the BN.

With this evidence in mind, Part 2 of this article will discusses the short and long term dynamics underlying the seat allocation exercise and also take into consideration opposition co-operation and tensions surrounding issues of political identity and political space, in regard to the DAP and PKR.

Part 2 will appear tomorrow

ONG KIAN MING is currently pursuing his PhD in political science at Duke University.