models?
Tom Lodge
A SECOND LOOK
Before considering any changes to the existing electoral system it may be useful to recall its virtues. National list proportional representation is just about the simplest way to translate voters’ preferences into equitable representation.
It benefits small parties, even very small ones, by allowing them a voice in law-making without harming larger parties.
In the last election 27 257 votes were
sufficient for the Azanian People’s Organisation to win a seat in Parliament.
As a system for bringing radical ends of the political spectrum into the mainstream it is difficult to beat. Even its supposed blemishes may have merit.
Firstly, the prohibition on floor- crossing stops the opposition from being weakened by defections to the ruling party. As the (quite unconstitutional) floor-crossing in the run-up to the 1999 election indicated, very few people defect from the ruling party. Most defections are motivated by careerist considerations, not principles.
A more liberal dispensation concerning parliamentarians changing sides would virtually guarantee the disappearance of most opposition parties between one election and another.
Secondly, the absence of constituencies may help to curb the development of patronage and corruption, as well as giving MPs more time to concentrate on legislative review, their main function.
Thirdly, the discretion exercised by party leadership, ensured through national list proportional representation, can benefit groups who otherwise might be excluded if public choice prevailed: in South Africa women are beneficiaries of the power of leaders to decide who should appear on party lists.
Advocates of electoral reform believe that a system which incorporated constituencies, that is, geographically defined areas which parliamentarians represented as individuals, would make legislators more accountable to voters.
There may be some truth in this, though the experience in South Africa over the past three years with ward-based local government is not encouraging.
Local councillors clearly do not believe that their prospects for re-election depend upon satisfying their constituents.
It is likely, though, that a constituency system which gave local political organisations a decisive role in candidate selection might well produce more attentive parliamentarians as well as arousing more public interest.
Given the three-million fall in voter participation between 1994 and 1999, a strong case can be made for any reform that might improve voter participation.
In a constituency system candidates have an incentive to construct the kind of electoral organisation that can expand turnout. Whether it generates better politicians is quite a different story, though.
There is no need to choose between proportional representation and constituency-based winner-take-all elections – a number of electoral systems combine the equity of the one with the accountability of the other.
For example, a mixed member representation system allocates a proportion of seats to legislators who are elected through “first past the post” polls in single member districts.
Parties which lose out through this allocation are compensated from the remaining proportion of seats.
A party which, for instance, won 20% of the votes but no seats in the constituencies would be rewarded with an appropriate number of national seats to bring its representation up to 20%.
In effect, parties field two sorts of candidates, one group for constituencies and another group who appear on a national list. This system functions in New Zealand, Germany and Bolivia with no serious difficulties, though list parliamentarians complain that compared to constituency representatives they are more subordinate to the will of party leadership.
Systems which are structured around multi-member constituencies in which voters have a single vote but several seats need to be filled are another relatively simple way to achieve an outcome which will produce a fair degree of proportionality. For example, in four member districts candidates who receive about 20% of the vote will almost certainly obtain a seat.
Such an arrangement is simpler than mixed member representation and does not create two classes of parliamentarians. It also excludes the smaller minorities and this exclusion might help to produce a more cohesive opposition.
Only two countries use this arrangement at present, Jordan and Vanuatu, though Japan employed it until 1995.
Altogether, within nine main types of electoral systems, there are more than 100 ways of electing parliaments which differ in detail from each other.
South Africa is obliged by its Constitution to select a method which results in proportional representation, but there are plenty of ways of achieving this.
The most popular electoral system world- wide is the “first past the post” system which continues to ensure minority under- representation in most parts of the British Commonwealth.
Its admirers suggest that it promotes strong parties and hence political stability.
Last year’s events in Lesotho, however, represent a telling example of what can go wrong when opposition groups fail to win seats despite obtaining substantial numbers of votes.
It may be too easy to blame the system for any shortcomings in the South African parliamentary representation.
Any method of election can deliver lazy, incompetent time-servers who simply rubber-stamp legislation prepared by the executive and whose prime motive is self- interest.
Nor is one form of election more likely than another to reward the conscientious, diligent and capable.
South Africa’s present Parliament has a fairly typical mixture of the two kinds of membership.
In a political system in which party loyalties play the main role in influencing the way people vote, the most important determinant of who represents voters is the procedure of candidate selection.
This lies outside the scope of any constitutional regulation.
Professor Tom Lodge works for the Electoral Institute of South Africa