The Electoral Commission’s report into the running of the three elections earlier this year in May does not read well for the Election Office for Northern Ireland (EONI).
Three elections in one day was always going to be a challenge, but the report takes issue with EONI’s planning and processes at the counts.
In her foreword to the report, outgoing Electoral Commissioner for NI Henrietta Campbell addresses the length of time it took to complete the count:
One of the major issues that arose at these elections was the length of time it took to complete the count. Although the count for the Assembly election in fact took no longer than in 2007 or 2003 it was perceived as being slow with little information being provided to candidates and to the media. It is a matter of regret that the Northern Ireland totals for the UK-wide referendum were not declared until 2am on Saturday 7 May, four hours after the rest of the UK, and in the absence of any media presence.
In the absence of any mainstream media presence, but you can relive the moment in video courtesy of new media!
While praising much of the activity on polling day itself, and appreciating the positive benefit of the early issue of polling cards (that encouraged 17,000 extra voters to register in time for polling day), the Electoral Commission noted that key EONI staff were “overstretched” and “fatigued by the time it came to managing the counts”.
There was no evidence of an overall plan on how the counts should be delivered, and instructions on how to conduct the counts were very late in being sent to AEOs [Area Electoral Officers] from the EONI head office.
The report’s commentary on EONI’s referendum preparations and execution explained that neither Chief Electoral Officer nor his Area Electoral Officers participated in the national referendum rehearsal. No separate NI rehearsal was organised. Instead the EONI’s Information Officer stood in.
3.55 It was clear that some AEOs or the person delegated by them to manage the referendum count were unaware that they had to send their verification totals to the King’s Hall by email or fax. Most were under the impression that a verbal communication was sufficient. There was also no understanding that the failure to send through verification totals meant that the announcement of the overall turnout figure for the UK referendum could not be made at the central count hub in London. When the turnout figure was announced on the night the percentage figure given had to be for Great Britain only.
3.56 It was not until 2am on Saturday 7 May that the CO was able to confirm the Northern Ireland referendum total. This was around four hours after the rest of the UK had provided their totals to the CCO. By that time the King’s Hall was deserted and there was no media present to record the total being announced.
3.68 On the first evening of the count there was evidence of poor communication channels in operation between the CEO and each of the count venues. Although the CEO used his mobile phone to contact AEOs this often proved unreliable because of poor signal coverage at some count venues. All count venues had access to a landline, internet and fax. However, in some cases no arrangements were made with the venue management to access this technology.
Training of polling station presiding officers used a different layout of ballot paper accounts form from the version issued on polling day. In the end, the threatened financial sanctions against presiding officers who failed to complete four essential tasks correctly were dropped.
While the Assembly count process (verification and counting) lasted two days – in line with previous Assembly elections – a longer verification process (three sets of ballot papers to process and many discrepancies in the paperwork completed by presiding officers) led to “significant delays in announcing turnout and first preference totals”.
Poor communication didn’t help the process. While broadcasters had been warned not to expect the first results until early evening, mid-afternoon results programmes focussed on delays, and inconsistent information from the media liaison officers at each count centre exacerbated frustrations.
Inconsistent use of PA systems, result boards and plasma screens to convey how the various stages of the Assembly STV count added to the confusion. In contrast, Belfast City Council was praised for the transparency it brought to its local government count by posting the results of each stage online in real time.
At a cost of £5,132 the Royal Mail undertook a sweep of the Mallusk mail centre to identify any postal ballots that had not yet been delivered so these could be retrieved and added to the count. 78 postal votes were rescued in this way … at a cost of £65.79 each! In hindsight, was that good value for money democracy?
As an electoral observer, I’m pleased to see that some of the issues reported back to the Electoral Commission are discussed in the report and recommendations.
Training the staff responsible for classifying doubtful ballot papers might reduce the variation of process I witnessed at the counts in May.
Steps to “conduct user-testing of ballot papers used at combined elections to alleviate the potential for voter confusion” may reduce the chance of using three shades of paper (white, buff, light grey) that are hard to distinguish under the poor light of many polling stations.
Clarifying what parts of the building/grounds hosting a polling station in which parties can canvass on polling day will bring consistency. Given the number of smaller parties standing at elections, any voluntary code of practice (short of legislation) may be limited in its effect.
Electoral matters are ‘excepted’, so not devolved to the NI Assembly. Whether the UK Government will take on board the need to “permit the name of the election being contested” on ballot papers when there is more than one election happening remains to be seen. But this would reduce the chance of voters writing “1, 2, 3” on one ballot and “4, 5, 6” on another.
A record number of 1,210,009 registered voters (since individual registration was introduced) is to be celebrated. However, the ever-decreasing turnout delivers poorer and poorer mandates for elected representatives. The report notes that 69% of people the Electoral Commission surveyed in NI gave a ‘civic’ reason for voting (compared to 61% in GB). Only 10% voted to ‘create change’. Of those surveyed in NI who did not vote, 67% gave reason of either ‘being too busy’ or ‘not being bothered’.