THE best way of understanding PR-STV (Proportional Representation by means of the Single Transferable Vote) is to start by seeing it as a sophisticated and economical way of doing what the French electoral system does rather crudely and at considerable expense.
The key problem for all electoral systems is how to translate votes into seats. The simplest and crudest solution is the first-past-the-post formula, i.e. give a seat to the candidate who is ahead of everybody else when all the votes have been counted.
The kind of problems which arise with such a system could be seen at the outset of the Thatcher era in British politics. In the 1979 general election in Britain, the Liberals won 14 per cent of the vote, but got 2 per cent of the seats at Westminster, whereas, with 44 per cent of the vote, the Conservatives got 53 per cent of the seats. All very well if you were a Thatcherite, but not very satisfactory if you were a Liberal and far from satisfactory from the point of view of fairness.
The system used in French presidential elections introduces a refinement of the above procedure. In order to win a seat a candidate must cross a certain threshold, in this case, 50 per cent plus one, or an absolute majority. A major problem is that, with more than two candidates, no one may reach the threshold. When this problem arises, all but the top two candidates are eliminated and the voters troop back a second time to choose between them. This amounts to asking those who voted for eliminated candidates to register their second preference.
These two features are integral aspects of PR-STV. In PR-STV, however, elimination is carried out in a much more refined way and the process by which the voter's lower-order preferences are ascertained is both more sophisticated and far less costly.
These points can be seen most clearly by considering how PR-STV would operate in a single-seat constituency. It should be noted that it would not then be a proportional representation system. On the other hand, it is not a wholly unrealistic example since it is the situation which obtains in by-elections.
PR-STV imposes a threshold, called a quota, which a candidate must achieve in order to win a seat. The formula for calculating the quota is:
This may look somewhat obscure until we consider it in the single-seat situation. Suppose, for simplicity's sake, that there are exactly 100 valid votes cast in this single-seat contest. Application of the formula leads to the following calculation:
So, in the single-seat situation, the quota is identical to the threshold used in the French system, i.e. 50 per cent +.1 or the absolute majority principle.
But the two electoral systems begin to diverge when it comes to dealing with the situation in which no candidate reaches the quota. PR-STV does not waste the voters' time by asking them to come back later and register their second preference. Instead, it collects both this information and information on third, fourth, fifth etc choices, all in one economical operation.
Then, rather than disposing of all but the two leading candidates in one fell swoop. PR-STV eliminates them one by one. The advantage from the voter's point of view is to preserve a wide range of choice of candidate for as long as possible, while the process moves by successive eliminations towards the goal of one candidate reaching the vital threshold.
PR-STV is not, however, merely a refined version of the majority-rule procedure. This is because, with the exceptions of by-elections and presidential elections, it has the all-important additional feature of multi-seat constituencies. This feature is essential in achieving the objective of reducing the disproportionality between votes and seats for any group or party (e.g. the British Liberal party's problem in 1979 of getting 14 per cent of the vote but just 2 per cent of the seats).
Multi-seat constituencies introduce two complications. The first is that the quota is changed.
In the single-seat situation the quota was half the votes plus one. A quick look at the formula shows that this principle can be easily extended as follows: in a two-seat constituency, the quota would be one-third plus one, in a three-seater one-quarter plus one, in a four-seater one-fifth plus one, etc.
The second complication is the need to transfer not just the votes of eliminated candidates, but the surplus votes of elected candidates, i.e. the number of votes by which an elected candidate exceeds the quota. If this were not done, those who voted for such a candidate would not get the full share of representation to which they are entitled.
For example, suppose, in a three-seat constituency, 50 per cent of the voters' vote for one particularly popular candidate. Since the quota is 25 per cent and since this is sufficient to elect the candidate, the second 25 per cent achieves no representation and is, as it were, wasted.
This problem is solved in PR-STV by transferring the surplus votes to continuing candidates in proportion to each such candidate's share of second preferences in the elected candidate's set of votes.
In summary, PR-STV is a highly refined and highly sophisticated instrument for the expression of political preferences. There is, however, evidence to suggest that many Irish voters do not exploit the full potential of the system. For example, traditionally some 60 per cent of Fianna Fail voters plumped for that party. But plumping does not confer any extra benefit on the voter's preferred party and deprives the voter of the possibility of further influencing the outcome of the election. An important question in this election will be whether Fianna Fail voters have broken the habits of a lifetime to maximise their party's chance of being in government.