[EM] Voting chaos

[James Gilmour]

> Brian Olson > Sent: 10 May 2007 02:24
> Too many more stories like this and I'll have to write up my own software 
> and go into business selling election counting systems.

The "chaos" was nothing to do with the software.  Be aware that many media reports have exaggerated
and twisted the facts, including the BBC.  I know because I was an Accredited Observer at the
Edinburgh Counting Centre.

The "chaos" was the unexpectedly large numbers of rejected ballot papers, especially in Glasgow and
Edinburgh, and predominantly ballot papers for the Scottish Parliament AMS elections (= regional
MMP).  This time we used a combined ballot paper, because prior research showed that 83% of ordinary
electors would prefer a combined ballot paper instead of two separate ballot papers as used in 1999
and 2003.  The combined ballot paper had been recommended by the independent Arbuthnott Commission.

The most common error on the MMP ballot paper was to put two Xs in the regional vote column and no
mark at all in the constituency vote column.  Germany and New Zealand both use combined ballot
papers for their MMP elections (one paper, two Xs in different columns) and their rejected paper
rates are very much lower than we saw last Thursday/Friday.  So the real question is why did
Scottish voters react so differently?  Some analysis in progress  -  maybe more on that later.

The delays in getting the results out were not caused by the counting system, but by the
unexpectedly large numbers of ballot papers that were queued for adjudication by the Returning
Officers.  There were thousands and thousands of them.  And representatives of the political parties
registered dissent again and again (recorded in the "Notes" that accompany the ballot paper image)
on a partisan basis when one of the two Xs was for their party.  These "pointless" challenges, that
by law must each be recorded, added to the delays.

Much less publicity has been given to the fact that, on the same day, we used STV-PR for elect 1,222
councillors to our 32 Local Government Authorities (councils).  In comparison with the Scottish
Parliament ballot, this all went very smoothly, caused voters few problems and the proportion of
rejected ballot papers was very much lower, mostly between 1% and 2%.  We should have liked more
figures below or around 1%, but given that the voters were marking another ballot paper with two Xs
at the same time and this was the first time the current generation of Scottish electors has used
STV for public elections, the rejection rate was well within expectation.

James Gilmour