The curious Lurgan electorate

The fairly irrelevant little by-election in Lurgan is now past history, and will only ever again be remembered by the small band of Northern Irish psephologists.

One miniscule issue remains unanswered, for this blog in any case. The reported turnout was low – the official result states that it was 23.62%. However, the turnout percentage is a function of two other figures: the votes polled (valid and spoiled) and the electorate. The votes polled (3,936) divided by the reported electorate (16,663) gives the turnout (23.62%).

So far so clear. The number of votes polled may or may not be 100% accurate – the counters can make mistakes, and the recounts that occur in many elections always give different figures. But this blog has no reason to question them in this case.

The figure given for the electorate, though, is a different story. Where did that figure of 16,663 come from? The Guide for candidates and agents, published by the Electoral Office in advance of the election, states that:

"Only those whose name is on the electoral register on 2 January 2010 will be able to vote." (Guide, p. 3)

"[Candidates] will be entitled to receive copies of the register for the District Electoral Area published on 1 December 2009, which you must use when completing nomination papers, and 4 January 2010, which is the register that will be used on polling day." (Guide, p. 18)

So that is clear then. The relevant electoral register is that for 4 January 2010.

But the electoral register for 4 January 2010 gives the following numbers of the electorates in the seven wards that make up Lurgan DEA:

Magheralin: 3,563
Donaghcloney: 2,660
Waringstown: 2,970
Parklake: 2,256
Mourneview: 1,623
Church: 1,385
Knocknashane: 2,473

This gives a total of 16,930. Not the figure of 16,663 that the Electoral Office used in their publication of the results.

So why is there a difference of 267 between the electorate that the Electoral Office said was the correct one, and the electorate that was published on the official result sheet, and that was used to calculate the turnout percentage?

If the 'correct' (4 January 2010) electorate is used, the actual turnout on Wednesday would be 23.25%. A tiny difference, of course, but a puzzling one.