In Balmoral, a stage 6 Alliance (Ekin) transfer went 42% to Sinn Féin (Long) and 42% to the UUP (Birnie and Stoker); normally, one would expect the UUP to get rather more than SF. Then, in stage 7, the UUP (Birnie) was pipped by less than 1 vote by the DUP (Patterson), and in stage 9, the latter was elected.
There is no mystery here; Ekin’s surplus came from the votes transferred to him from an SDLP candidate. Of the 1673 votes in his quote, 1508 were from Alliance first preferences, 16 from an independent candidate and the rest from elected and excluded SDLP candidates. Ekin had a surplus because he got too many votes from the SDLP; it’s entirely defensible that the votes that arrived with Ekin last were also the first out. Sad for Esmond Birnie, but wntirely within the rules.
At stage 6 in West Tyrone, the UUP (Bleakley) on 918.74 and SF (Kelly) on 913.08 were competing for the last seat under a quota of 1078. An SDLP (McDonnell) surplus of 31.19 was then transferred, stage 7, and even though his total of 1109.19 included 3.29 votes from the DUP, nothing went to the UUP, pretty well everything went to SF, and just 0.44 was non-transferable. Kelly went on to 932.58 and was elected (by default), while Bleakley, still on 918.74, was unelected.
This really is a completely inadequate description of what happened. Barry McElduff (SF) had a surplus of 393 (quota was 1078, he had 1471 first preferences and so was elected on the first count). His surplus was distributed on the second count. There were 1427 ballot papers with a first preference for McElduff and a valid next preference. They were therefore each transferred with a value of 0.27 (since 393/1427 = 0.275402943). 170 of these papers went to McDonnell at a value of 45.90; he already had 1060 first preferences and had picked up 3.29 votes (7 papers at 0.47) from a DUP first count surplus, so it was enough to put McDonnell over the quota by 31.19.
When it finally came time for McDonnell’s surplus to be distributed, the 170 votes that put him over the quote were the ones from which the surplus was taken. There were 123 of them, so they were transferred at a value of 0.25 (since 31.19/123 = 0.253577236). Not at all surprising that since they came from SF they went back to SF, since the only other remaining candidate was a Unionist.
To say that “just 0.44 was non-transferable” is simply wrong. In point of fact, 47 of the transferred votes in the package of 170 were non-transferable. The 0.44 added to the “Non-transferable” tally is a book-keeping factor, with only an indirect relation to the pieces of paper.
But even if we took the position that the 3.29 votes from the DUP should have counted in the 31.19 surplus, each of the seven ballot papers in question would have had its value rounded down to 0.01; it would not have done much for Bleakley.
Other anomalies, if not errors, occurred. In Abbey, for example, with two other DUP candidates and one UUP already elected and with 3 other candidates eliminated, a stage 8 surplus from the DUP apparently involved no non-transferables at all!
Nothing at all incredible here. The surplus was 8.96; 32 votes happened to be transferable; so each of them went at a rate of 0.28, with nothing left over.
In Moyle, a stage 2 transfer showed a non-transferable total of 0.22 which equates with an impossibility: just half a voter!
No idea where this “half a voter” comes from. This count is such a simple one that I may as well put the entire thing here (slightly rearranged for clarity):
David McAllister DUP 481 -198.00 283.00 George Hartin DUP 470 – 470.00 William Graham UUP 326 – 326.00 Price McConaghy Ind 222 59.74 281.74 Robert McIlroy UUP 172 123.54 295.54 Thomas Palmer Ind 23 14.50 37.50
It’s fairly straightforward to calculate that 341 of McAllister’s 481 votes were transferable, at a rate of 0.58 (since 198/341 = 0.580645161); and 213 went to McIlroy, 103 to McConaghy and 25 to Palmer. That leaves a remaider of 0.22.