It has been widely assumed that weak cross-community vote transfers in the Northern Ireland Assembly election are responsible for the fact that pro-agreement forces are not in a more powerful position within the Assembly and, especially, within the unionist camp. Does the evidence of transfer patterns bear out this assumption, and what does it tell us about that other vital aspect of transfer behaviour, namely party loyalty?
The Proportional Representation-Single Transferable Vote rules as used in Northern Ireland differ somewhat from those applied in the Republic. In particular, surplus votes are transferred at the appropriate fraction of their value and the process also takes account of non-transferable votes in determining the destination of the surplus. Both features make the Northern results more reliable as a guide to voting behaviour.
For example, the distribution of the surplus votes of Dr Joe Hendron, the leading SDLP candidate in West Belfast, revealed a rather poor rate of party loyalty (60 per cent) and a high rate of plumping for Dr Hendron (29 per cent). If these surplus votes had been redistributed according to the rules of the count in the Republic, the non-transferable votes would have been set aside in calculating the transfer proportions, the reported rate of non-transferability would have been zero, and the apparent SDLP solidarity would have been 85 per cent.
However, there is another analytical pitfall that both systems share. This arises from the rule that, in distributing a surplus that has accrued on a count other than the first, the returning officer only examines the votes that the elected candidate received at the count on which he or she was elected. Provided the votes in question have come from a party running mate, this does not affect the interpretation of the transfer patterns. If, however, the decisive votes have come from some other party, very peculiar transfer patterns can appear.
Thus, on the ninth and last count in East Derry, the results show that, in a contest between the UUP, the DUP and an Independent Unionist (George Robinson), only 7 per cent of SDLP terminal transfers went to the UUP and 90 per cent became non transferable. This would seem to be an egregious example of SDLP voters failing to cross the political divide to help a UUP candidate in a contest with the DUP.
In fact it is no such thing. In accordance with the rule mentioned in the previous paragraph, the destination of the SDLP surplus of 2,102 votes was determined by examining the votes received by the elected candidate (Arthur Doherty) on the eighth count. These votes came from the elimination of a Sinn Fein candidate. In short they tell us a lot about the attitudes of Sinn Fein voters and little or nothing about the willingness of SDLP voters to "cross the divide".
In analysing transfer patterns, it is essential to take anomalies of this kind into account. It is also essential to take into account the different transfer situations, i.e. which parties are still in contention when a particular set of votes is being distributed. A full analysis that takes all this into account will take some time. In the meantime, with the aid of the RTE results system, it is possible to provide preliminary estimates of the key transfer variables mentioned above.
Party loyalty in Northern Ireland, as measured by the rate of internal party transfers, ranges from a high of 87 per cent for Sinn Fein to a low of 57 per cent for the UKUP. The Sinn Fein figure is truly impressive; this is a disciplined political machine capable of delivering votes with remarkable consistency. It is significant that the only other party in Northern Ireland that is in the same league is the PUP. The other parties, the DUP included, showed levels of party loyalty that are significantly below those regularly achieved by Fianna Fail until quite recently.
The commitment to explicit or implicit alliances within each political bloc also varies considerably. Sinn Fein was quite supportive of the SDLP with a terminal transfer rate to that party of 68 per cent. SDLP voters did not reciprocate: only 45 per cent of SDLP terminal transfers went to Sinn Fein in situations in which that party was still in the running. On the other side of the political spectrum, terminal transfers from UUP to the UDP or PUP simply did not arise. Transfers from PUP to UUP, which can be measured, were a modest 43 per cent. This brings us to the crucial cross-community transfers. The key tests here are UUP-to-SDLP transfers when the Alliance Party was not available and SDLP-to-UUP transfers when Sinn Fein was not available. In the former case 36 per cent of UUP terminal transfers went to the SDLP.
Unfortunately, there is no direct measure of SDLP-to-UUP transfers. This is because the transfer situation in question only arose on two counts. One of these (the ninth count in East Derry) has already been discussed above. What one is really measuring in that situation is Sinn Fein-to-UUP transfers and, for obvious reasons, that gulf will take a long time to bridge.
The second situation involving SDLP terminal transfers and a potential UUP but no SF recipient included one of the NIWC candidates (Jane Morrice on the tenth count in North Down). Given the centrist role of the NIWC, this is not a relevant test of SDLP willingness to engage with the other side.
There is, however, an indirect measure of SDLP-to-UUP transfers. On the tenth count in Fermanagh and South Tyrone, 36 per cent of a Sinn Fein surplus went to the UUP. However, before leaping to the conclusion that one-third of Sinn Fein supporters in Fermanagh and South Tyrone were cosying up to David Trimble, one should note that the surplus in question had come from an SDLP elimination. Consequently, the 36 per cent rate of transfer is an indirect indicator of SDLP to UUP transfers, which, as it happens, exactly matches the rate of transfer in the opposite direction noted above. This reciprocal transfer rate indicated by these straws in the wind may not match the expectations of those who are impatient to see a radical breakthrough in the politics of Northern Ireland but, in reality, it is another modest but significant sign of the growth of centrist politics.
Richard Sinnott is director of CEEPA (Centre for European Economic and Public Affairs) at UCD and author of Irish Voters Decide: Voting Behaviour in Elections and Referendums since 1918 (Manchester University Press, 1995)