As the smoke clears after the Malaysian election battle it has become clear that under the current electoral system defeat of the ruling Barisan Nasional (BN) was never quite on the cards, even without the electoral roll and election day cheating and vote buying claimed by the opposition.

Indeed all other factors being equal it would probably take another 4 percent swing against the BN for the opposition Pakatan Rakyat (PR) coalition to win the majority of seats.

As it was this was a remarkable victory for the PR which won 51.7 percent of the popular vote and 53 percent in under-represented peninsula Malaysia. Yet it took only 89 seats compared to the BN's 133 seats. Those numbers tell the tale of just how rigged the system has become.

Each BN seat cost an average of 39,400 votes while each PR one cost 63,200. Those figures show the success of years of outrageous BN gerrymandering which has made nonsense of democratic, one-man one-vote principles. The extent of this has gone almost unnoticed by the foreign media - and foreign government reaction, treating the result as though it were the outcome of a relatively normal democratic process. Gerrymandering on Malaysia's epic scale is just as undemocratic as the ballot-counting frauds which President Ferdinand Marcos used to turn defeat into claimed victory during his years as president of the Philippines.

Given the BN's control of most media and the machinery of government, the result was a remarkable victory for the PR. So although there is much disappointment and some youthful anger among the ranks of the PR that is because expectations of what could be achieved were naturally over-optimistic. Assuming the PR holds together till the next general elections it will probably need to begin to reverse the gerrymander - or break the Barisan - if it is to break UMNO's stranglehold on power.

More immediately the opposition and its component, Anwar Ibrahim's Parti Keadilan Rakyat in particular, is focused on the cheating alleged to have taken place in a significant number of constituencies whether through giving ballots to non-nationals, voting more than once, manipulating the electoral roll or simply offering cash to those who prove they voted for the BN. The PKR says that as 27 seats were won by the BN with a majority of less than 5 percent of the vote, cheating could have made a critical difference. Certainly in some of the more closely fought contests, the arithmetic somehow seemed to favor the BN. Thus it was declared the winner in no less than 11 of the 15 seats where the margin was less than 1,000 votes and in 25 of the 35 seats where the margin was under 2,000 votes. This may or not have been chance.

Nonetheless given the other factors favoring the BN, cheating would have had to be on a more substantial scale than has been shown to be the case so far, and to have been well-targeted - some alleged instances occurred in constituencies where the BN was never under threat.

So looking ahead the PR must find some way to raise not just awareness but real anger at how the democratic process has been undermined to protect the BN's politics of corruption and patronage. Some of the numbers are quite startling thanks to Dr Mahathir's getting rid of the rules which once governed the relative size of constituencies. Thus today the largest constituency PKR held, Kapar in Selangor, had an electorate of 144,159, nine times that of the smallest, fittingly Putra Jaya, the seat of government with just 15,791.

In many countries rural voters are somewhat favored over urban ones but Malaysia has taken this to remarkable extremes. This was originally primarily aimed at limiting the predominantly Chinese DAP which once dominated all the inner cities and remain strong there. But now it is the mixed, mostly Malay majority new urban and suburban districts which are increasingly disenfranchised primarily to the disadvantage of the PR, and PKR in particular.

This disenfranchisement can only get worse unless new constituencies are created and boundaries re-drawn for the simple reason that this is where most population growth is taking place. Selangor with its urban centers surrounding the federal territory now has 10 constituencies with electorates of over 100,000. By contrast, Johor, an UMNO stronghold has only one such constituency, that held by DAP leader Lim Kit Siang. Other relatively large constituencies in Johor are also mostly DAP or PKR. The same phenomenon is found in other major cities such as Seremban, where the DAP held seat has over 100,00 voters, the cities of Melaka, Ipoh, Taiping, Kuala Kedah etc etc. The extreme bias combines ethnic with political discrimination on a grand scale.

The most egregious of all is found in Sarawak, whose PBB is now the second largest component of the BN, with 14 seats acquired at average cost of 16,600 voters. Needless to say the urban constituencies of Kuching, Sibu and Miri with their large Chinese populations have vastly larger electorates. The rural Sarawak seats are not only mostly tiny but display the power of local leaders over the electorate. In one the PBB candidate polled 10 times that of his only opponent.

The over-representation of Sarawak - and to a lesser degree Sabah - owes something to the special consideration they were given to persuade them to join Malaysia. But it has been augmented by UMNO's need for the seats that the PBB and its mega-rich leader Abdul Taib Mahmud could deliver to the BN. In both the Borneo states there appears a constituency size and boundary bias not just against the Chinese and urban people generally but also non-Muslim bumiputras.

It is ironic then that these states, which often appear less starkly divided by race and religion than the mainland ones, and also harbor some semi-separatist sentiments, should now be the main prop of UMNO at the federal level. No wonder Taib is left alone despite the many money scandals for which the state has become especially notorious.

So is there any prospect of the reform of the gerrymander? This year a national boundary delineation is due. But changing boundaries will do relatively little. New seats need to be created (or old ones merged). That requires a two-thirds majority in parliament. In the past the BN two-thirds majority enabled it to gerrymander by adding seats. Now any new seats will require agreement with the opposition. In theory the status quo suits the BN. But it also increasingly exposes it to ridicule and resentment by the new urban voters and makes it ever the less likely that Chinese support can be won back. So can it do a deal which grants some electoral reform?

There seems scant prospect of anything remotely constructive while the PR is still seething with frustration and Najib is under pressure from UMNO critics headed by Mahathir and must face UMNO elections later this year. But if UMNO is to do more than circle the wagons of racial discord around its bloated patronage system it must take aboard some lessons from its electoral defeat. The main one is that class and income distribution issues are now more important than race for increasing numbers of voters. With urbanization of a youthful population, that can trend only grow, especially in the currently disenfranchised new urban areas. The semi-feudal UMNO is failing to adapt to the modern urban Malaysia that it has actually, and to its credit, created.