Richard Sinnott's article in yesterday's Irish Times highlights a number of interesting aspects of the Assembly election results but, as he remarks, a full analysis will take some time. A number of facets are, however, worth identifying at this stage.
First of all, it appears to me that three candidates owe their success to substantial cross-community transfers. These are the two successful Women's Coalition candidates, Monica McWilliams and Jane Morrice, and the SDLP's Danny O'Connor in East Antrim.
In Belfast South, Monica McWilliams received almost 20 per cent of second preferences from the eliminated Sinn Fein candidate and almost 15 per cent of preferences from the PUP candidate in what proved to be something of a roller-coast affair. Firstly, the Sinn Fein preferences enabled her to pass, and then stay ahead of, the Alliance candidate, Steve MacBride, while, at the same time, losing her temporarily the lead of fewer than 200 votes she had built up over her SDLP rival.
Then, in the very next count, 400 preferences from the eliminated PUP candidate, Ernie Purvis, restored her lead over Carmel Hanna. She then remained ahead of both her rivals until the final count, in which the additional 920 votes she had received from the two extremes enabled her to defeat both the Alliance and SDLP candidates by margins of 350 and 400 votes.
In North Down, from the fourth count onwards, Jane Morrice of the Women's Coalition was running 700 to 1,000 votes behind both the UUP candidate, Peter Weir, and the DUP candidate, Alan Graham. But, in the ninth count, the gap between her and Alan Graham was narrowed by transfers from the Alliance Party. And then, in the final count, the transfer to her of three-quarters of the 2,458 votes of the eliminated SDLP candidate shot her ahead of both her UUP and DUP rivals - in the former case by a narrow margin of 137 votes.
But by far the most striking case of cross-community voting was in East Antrim, where the transfer of one-quarter of the surplus of UUP candidate Ken Robinson elected Danny O'Connor of the SDLP, thus defeating the second DUP candidate and creating a majority of two for the pro-Agreement unionist parties.
This remarkable result was facilitated by the fact that, in an earlier count, barely 40 per cent of the DUP lead candidate's surplus in this constituency had transferred to that party's second contender.
But if these seem to be the only three cases where cross-community transfers proved decisive, there were two others with very close results at the final stage: North Belfast (289 votes between the last two candidates) and Strangford (159 votes between the last two) as well as several other counts with wider margins but where the outcome could, nevertheless, in certain circumstances been different.
In North Belfast an 80 per cent transfer from Alban Maginness to his SDLP partner Martin Morgan still left the latter short of the votes needed to win the seat from the anti-Agreement Independent Unionist William Agnew. But that second seat would have been secured had Martin Morgan initially secured even 300 of the 6,196 first preference votes cast for his Lord Mayor colleague. Such a result would, incidentally, have eased David Trimble's position by reducing from 28 to 27 the number of anti-Agreement unionists in the Assembly.
In Strangford, the margin by which the SDLP candidate failed to gain a seat from the UKUP was only 159 votes but it would have required an unrealistic 95 per cent transfer from his party colleague to have bridged this small gap - or else an equally unrealistic 90 per cent transfer from both his own party colleague and from the previously eliminated Sinn Fein candidate.
One is forced to conclude that in this instance there simply weren't quite enough nationalist votes to win a seat.
In West Belfast the failure of unionist parties to secure any seat reflected not merely a poor turnout by their voters and highly disciplined transferring of votes within Sinn Fein, but also plumping by more than 900 DUP and PUP voters. Despite the low unionist turn-out, their transfers could have elected Chris McGimpsey.
In Upper Bann, as in North Belfast, a more even distribution of SDLP votes between Brid Rodgers and her party colleague Mel Byrne could have won a second seat.
But, despite criticism of SDLP tactics in Newry and Armagh, it is not in fact clear that even if Seamus Mallon's massive vote had been better spread, a third seat could have been won by the SDLP at the expense of Sinn Fein. The gap of more than 1,300 votes on the last count would have been difficult to bridge by any combination of tactics and strategy.
It should be added that the fact that Sinn Fein candidates secured more than 900 Seamus Mallon transfers represented not so much a leakage of SDLP votes to Sinn Fein as evidence of Mallon's ability to pull first preferences from Sinn Fein on a purely personal basis.
An interesting case where exceptional circumstances led to an unusual preference voting pattern was in Fermanagh and South Tyrone. There the distribution of the surplus of the second Sinn Fein candidate yielded 441 additional votes for the surviving UUP candidate, Joan Carson. This was only 46 votes short of what would have been needed to enable her to take the last seat from Maurice Morrow of the DUP.
The UUP would have won this seat had the 7 per cent of Sinn Fein voters who passed preferences to the DUP transferred these votes instead to the UUP candidate!
More generally, it is clear from the details of the counts that the great majority of the electorate voted down the ballot paper: the proportion of non-transferable votes was quite small.
Even where at the time of elimination there remained no other member of the party of their choice to whom to transfer their vote, most voters chose to pass on their preferences elsewhere so long as there remained a reasonable range of parties from which they could choose.
In such cases the proportion of non-transferable votes was as low as 8 per cent for the UKUP (two constituencies); 13 per cent for the Women's Coalition (three); 15 per cent for Alliance (five); 18 per cent for the UDP (two); 25 per cent for the PUP (five); and 28 per cent for Sinn Fein (four).
In five constituencies where Alliance voters had a choice of pro- and anti-Agreement unionist parties, SDLP and Sinn Fein, their preferences went as follows: UUP 36 per cent; SDLP 33 per cent; DUP plus UKUP 3 per cent; Sinn Fein 2 per cent. (Independent unionists in three of these constituencies, and a Women's Coalition candidate in another, accounted for a further 9 per cent.)
It may be noted that in North Antrim, where there remained no SDLP candidate when the Alliance candidate was eliminated, the consequence was that 42 per cent of the Alliance votes were non-transferable.
In the case of the Women's Coalition there was an exceptional transfer of 28 per cent to Sinn Fein in Fermanagh and South Tyrone where, incidentally, they had no Alliance candidate to whom to transfer.
But in the other two constituencies where Women's Coalition voters had a choice of a full range of other parties for whom to vote, the transfer pattern was: Alliance 37 per cent; SDLP 18 per cent; UUP 17.5 per cent; DUP/UKUP 6.5 per cent and Sinn Fein 2.5 per cent.
The significance of these transfer figures is that they suggest that some 45 per cent of both Alliance and Women's Coalition voters were Catholics - and it looks as if the same was true of the other smaller cross-community parties. As a total of 9.4 per cent of the electorate voted for Alliance, the Women's Coalition and these smaller groups, the Catholic element of these votes, when added to the figure of just under 40 per cent of the electorate who voted for the SDLP and Sinn Fein, yields a figure of 44 per cent for the Catholic share of voters in this election.
Now, given that the Catholic share of the total electorate is known to be about 41 per cent, this suggests that the turn-out of Catholics may have been almost 10 percentage points higher than that of Protestants.
This is confirmed by the differences in turn-out in constituencies that are predominantly inhabited by members of one or other section of the community.
In the most overwhelmingly Protestant areas outside Belfast the turn-out was only 6061 per cent. Where the Protestant majority was smaller, the turn-out ranged from 64 per cent to 69 per cent. And where, as in Upper Bann and Fermanagh and Tyrone, it was smallest, the turn-out exceeded 70 per cent.
Finally where there was a Catholic majority, the turn-out ranged from 74 per cent to 85 per cent.
Nevertheless, in this election the difference in turn-out between the Catholic and Protestant electorates may have been somewhat less than the exceptional differential in the May referendum, where the Catholic turn-out seems to have been as much as 15 percentage points higher than that of Protestants.