The number of seats won by the political parties in Kenya's National Assembly would be significantly different if the existing constituencies were to be redrawn into geographical units with equal populations.

Research conducted by a leading American political scientist, Prof Joel Barkan, shows that had this been done for the 1997 polls, the ruling party, Kanu, which received only 41 per cent of the total votes cast but took 51 per cent of the seats in parliament, would have had four seats less.

The official opposition, the Democratic Party, which got 39 seats, would have ended up with 41 seats in parliament, while the defunct National Development Party, which took 21 seats, would have received 24.

Ford Kenya, which took 17 seats, would have ended up with a similar number of seats, as would have the Social Democratic Party, which took 15 seats.

Safina, which won five seats, would have ended up with only three seats in the new scenario. The other minor parliamentary parties, which shared six seats between them, would have ended up with an extra seat.

There have been calls in the past to redraw Kenya's current 210 constituencies to reflect equal population sizes.

Critics say the current system has tended to favour the sparsely populated areas, where Kanu enjoys greater support, over the more densely populated areas like Nairobi, Central and Nyanza provinces, where the opposition enjoys more support.

Both Kanu and the opposition agreed that a new electoral system that would reverse the trend witnessed during both the 1992 and 1997 multiparty polls is needed, although none of these have detailed the precise scenario they would prefer.

Kanu's secretary of legal and constitutional affairs, Mr Otieno Kajwang', said last week that the party was pushing for radical changes in the country's electoral system which if adopted, would introduce proportional representation (PR) in the country's parliament alongside the current constituency-based one.

This, he said, was one of the proposals the party intended to forward to the Constitution of Kenya Review Commission (CKRC).

The CKRC, which has been in existence since September 2000, is mandated under the Constitution of Kenya Review Act to provide a comprehensive review of the current constitution.

"This is one of the proposals in our party's position paper that we intend to forward to the CKRC once it is ratified by our national executive committee," Mr Kajwang' said, adding that the party felt strongly that a "mixed electoral system" would redress some of the limitations of the existing electoral system.

"The constituency-based electoral system," Mr Kajwang' said, "was one of the factors that had, in the first instance, brought about the urge for a constitutional review in the country because of what many felt were inherent deficiencies of that system."

Mr Kajwang' chairs the Kanu sub-committee which was recently appointed to work out the party's position on the current constitution review.

The committee also includes Cabinet ministers Nicholas Biwott, Bonaya Godana, Julius Sunkuli and assistant ministers William Ruto and Joseph Kiangoi. Party secretary general Raila Odinga is an ex-officio member.

Legal and political experts in Kenya have in recent years called for a shift from the current constituency-based electoral system on the grounds that it gave Kanu undue advantage over the other parties.

They say this advantage was institutionalised by the British colonial government, which drew up independent Kenya's first constituencies.

Critics of the constituency-based electoral system say the system tends to ensure that the larger parties tend to get a much larger proportion of the seats in parliament than the proportion of votes obtained at elections.

Prof Njuguna Ng'ethe a political scientist at the University of Nairobi's Institute of Development Studies says this was why Kanu won 107 seats in Kenya's 210-member parliament after the December 1997 elections even though the party only won 41 per cent of the total votes cast.

Prof Ng'ethe says that even the nomination of 12 other members of parliament by various political parties did not quite resolve the problem.

"It simply reproduced the same scenario that existed in parliament as the nominations were based on the number of seats each party had in parliament," he said.

Had the nominations been based on the number of popular votes, the allocations of seats to nominated Members of Parliament would have been different and Kanu with only 41 per cent of the votes would have got five seats instead of the six it took.

The latest move by both Kanu and the opposition is being viewed as a step towards addressing the issue which has not been given much thought in the past.

"The choice of electoral systems has never been systematically discussed in most of the emerging democracies with the exception of South Africa," says University of Iowa's Prof Barkan.

"Yet," he notes, "it is electoral systems that determine election outcomes and affect the composition of the legislatures of various countries." 

"What we have often had, according to him, has been a continuation of systems bequeathed by the former colonial powers.

In English-speaking countries the British system was adopted while the same happened in French-speaking countries.

Prof Barkan who has written widely on the politics of eastern Africa, is the author of Politics and Public Policy in Kenya and Tanzania and Beyond Capitalism versus Socialism in Kenya and Tanzania, a comparative analysis of the transition to multi-partism and the early economic reform efforts in the two countries.

Prof Barkan has done a series of simulations of different alternatives on his website www.uiowa.edu/~electdis which present different scenarios with different electoral systems.