Thursday, July 31, 2014

Why bother with delineation?

12:00PM Jul 31, 2014

Electoral delineation or delimitation in Malaysia is a highly politicised exercise as the end result is about which parties get their political position strengthened or weakened.

One stark example is the result of a 2002 delineation exercise that contributed to Barisan Nasional (BN)’s thumping majority of 199 seats out of 219 seats (90.9 percent) and a popular vote at 63.9 percent. This means BN got 142 percent of their rightful share.

Violation of the ‘one person, one vote, one value’ principle

This happened because electoral delineation created un-leveling the playing field in favour of the ruling coalition.

It has created incredible seat-vote disproportionality and undermined the fundamental principle of ‘one person, one vote, one value’ for representation democracy, where voters are supposedly able to elect their representatives in a fair manner to represent their interests. This principle is paramount because it underlines political equality of every voting citizen, and without political equality, there can be no democracy.

But, how do we identify, measure and prevent the violation of this cardinal principle? There are two ways.

Malpportionment - disparity across constituencies

Most people cast their eyes on the disparity across constituencies, technically known as ‘malapportionment’. The logic is simple: if Constituency A elects a parliamentarian with less voters than Constituency B, then voters in Constituency A have greater power than voters in Constituency B. The larger the difference, the greater the unfairness.

Malaysians are not unfamiliar with ‘malapportionment’, and until recently it is often called wrongly as ‘gerrymandering’. Most can cite the extreme examples of Kapar having more than 140,000 voters and Putrajaya having only nearly 16,000 voters, nearly 10 times the difference in the last election. The more savvy ones may point to the ridiculous disparity between mountainous Baling (72,387 voters) and state capital Alor Setar (56,007) after the 2003 redelineation.

In an ideal world, every constituency should have exactly the same size, but this may not be possible in reality as constituency boundary may be constrained by administrative, physical or human boundaries. Absolute equal apportionment may result in very odd-shaped and possibly meaningless constituencies which impair representation, similar to the outcome of ‘gerrymandering’.

In other words, ‘malapportionment’ may not necessarily be deliberate and malicious but it is nevertheless bad for democracy.

The greater the degree of malapportionment, the worse it is. To strike a balance between equal apportionment and representative boundaries, the international best practices on redelineation propose not one single electorate size for all constituencies, but a reasonable range for deviation from the average, or what is technically called ‘electoral quotient’. For example, Europe’s Venice Commission proposes a 10 percent band for normal circumstances and a 15 percent in the extreme.

Seat-vote disproportionality - disparity across parties

What most people may have overlooked is the disparity across parties, what is called seat-vote disproportionality. While malapportionment can be known before election, seat-vote disproportionality can only be known after an election.

How bad is it then in the last general election?

From Table 1, we can tell that BN won 1.26 times of its rightful seat share, while Pakatan Rakyat won only 0.79 times - that’s the very reason why Pakatan loses power despite winning the mandate. In particular, PKR and PAS won only 0.66 and 0.64 times of their rightful seat share, greatly disadvantaged. In contrast, DAP was slightly over-represented at 1.09 times as its votes were more concentrated.

At Bersih Delineation Action Research Team (DART) training, we use the analogy of money. If a vote is equivalent to a bank note of RM1, and RM 1 can buy you a kg of sugar; then BN walked into a shop with RM1 and walked away with 1.26kg of sugar, while PAS got just barely half a kilogram.

Divide the vote value of BN by each of the opposition parties, we get the relative value of a BN vote to the vote of the latters. A BN vote is 1.19 times of a DAP vote, 1.91 times of a PKR vote, a 1.98 votes of a PAS votes and 1.59 times of a Pakatan Rakyat vote.

Was this bad? Quite the contrary, the GE13 delivered the fairest result in history.

Table 2 was calculated in a similar manner to Table 1. Look at the 2004 election which we discuss in the beginning of this article, a vote for BN was 26 times in weight to a vote for Keadilan. Why? The opposition parties won some 8 percent of votes but only one seat in the 219-seat Parliament.

Was that the worst case? No. In 1986, PAS won some 15 percent of votes but only one seat in the Parliament with 177 seats then. So, one vote for BN was equivalent to 40 votes for PAS!

The underestimated danger - gerrymandering

What caused the stark seat-vote disproportionality which practically made our elections a scam? One culprit is of course ‘malapportionment’ which has been creating generally smaller constituencies to be won by BN. The other greatly under-understood and under-estimated cause is ‘gerrymandering’.

Gerrymandering means drawing constituency boundaries to deliberately advantage some parties and disadvantage some others. And this can be done independently of malapportionment.
Graphic 1 shows a hypothetical country of 100 voters divided into two equal-sized constituencies. Voters are divided between the Red camp and the White camp, with a 56:44 edge to the reds.

Before the redelineation (left), the Reds will carry Constituency X and the white Constituency Y. Using the same way we calculate the GE13, the vote value of Red is 50/56 or 0.89 while that of White is 50/44 or 1.13, which is the fairest one can get.

Now with a skillful gerrymandering (right), the voters’ pattern change drastically. Both Constituencies X and Y now have 28 reds and 22 whites. So, Red will now win both seats, resulting in a vote value of 1.79 for Red and 0 for White.

This is the danger of focusing on only malapportionment and ignoring gerrymandering in redelineation.

Are there alternatives?

Can the obsene degree of seat-vote disproportionality of 1986 or 2004 replay in the next election after another round of redelineation? Of course, unless we find the way to stop the rape of elections by none other than the Election Commission.

But how? The People’s Tribunal, carried out in September 2013, recommended the establishment an independent and objective boundaries commission to review and make recommendations to draw the boundaries of electoral constituencies rather than leave this task to the Election Commission.

Members of the boundaries commission will have to be independent and multi-partisan which includes civil societies, political parties, professionals and academicians who have expertise in boundaries delineation or electoral processes.

But realistically to have such an independent Boundary Commission will almost be impossible under BN.

Hope is however not lost. The federal constitution provides a process for public intervention into the redelineation process, but made too technical for understanding.

To break the Election Commission’s technical and informational edge, Bersih 2.0 and Engage, a citizen group based in Johor Baru, have launched a bottom-up initiative called DART. It provides digital tools for easy usage.

This will be a flagship project of Bersih 2.0 in preparation for the 14th general election. We strongly believe, ordinary people, like you and me, must be allowed to take part in re-drawing our constituency boundaries to reflect our local ties, rights, and interests.

MARIA CHIN ABDULLAH is the chairperson for the Coalition for Clean and Fair Elections (Bersih 2.0) and the executive director of Empower. She believes politicians are bad masters if not made good servants through free, fair and competitive elections.

- See more at:

No comments: