The counting of votes, known as the scrutiny, began in each polling place as soon as they closed their doors at 6pm on polling day.
Only ordinary votes were counted on referendum night – just over 83 per cent of votes at the 1999 referendum. Absent and provisional votes cast at each polling place were put aside as checks had to be made at the divisional office to ensure that these ballot papers were eligible to be included in the count.
After the doors to the polling place closed the polling officials opened and emptied the ballot boxes. The buff ballot papers for the republic question were unfolded and the 'yes', 'no' and informal votes were put into separate piles and counted. This procedure was repeated for the mauve ballot papers for the preamble question.
The totals of 'yes' votes, 'no' votes and informal votes for both questions were phoned through to the DRO. The DRO then entered the figures received from each polling place in their division into the AEC's computerised Referendum Night Results System. The figures entered into the computer system were then electronically transmitted to the AEC's web site which served as the main method of disseminating referendum results.
On completion of counting, the polling officials placed all the ordinary ballot papers into parcels and along with the declaration vote envelopes securely delivered them to the DRO.
Ballot papers correctly marked according to the rules for voting are called formal votes and only formal votes contribute to determining the results of a referendum. Ballot papers that do not satisfy these rules are regarded as informal and after their total has been tallied they are excluded from any counting.
A ballot paper in the 1999 referendum was informal if:
As it had been 11 years since the last federal referendum and electors were more familiar with numbering ballot papers, the AEC devoted considerable effort to emphasise to electors the importance of the words 'yes' and 'no' in marking their referendum ballot papers. The referendum advertising campaign was based around these two words and the other public information and education activities reinforced their importance.
These activities, aimed at minimising the number of informal votes, contributed to a very low level of informal voting being achieved at the referendum. For the republic question only 0.86 per cent of votes were informal and only 0.95 per cent of votes were informal for the preamble question.
The national and State and Territory summary of informal voting for both proposed constitutional changes is on page 74 of the statistics section and the divisional breakdown is on pages 74 through 76.
Scrutineers were able to observe the counting of the votes at all polling places on referendum night. They were also able to attend the scrutiny of declaration votes carried out in the following weeks.
Scrutineers had the right to observe all stages of the scrutiny and could challenge the formality of ballot papers but they could not handle, separate or arrange ballot boxes, ballot papers, declaration envelopes or other referendum materials.
The listed officials were also able to appoint one scrutineer for each counting centre on referendum night.
The 1999 referendum did not have a National Tally Room as it was determined that the internet would be the most timely and cost effective way of providing referendum results to the media and other interested people on referendum night.
A specially created website at referendum.aec.gov.au was established to house the virtual tally room. The virtual tally room was the official source of 1999 referendum results on referendum night and in the post-referendum period.
On referendum night, the virtual tally room received a direct feed from the AEC's computerised Referendum Night Results System to display a number of live tables containing a range of referendum results. A national graph screen also showed a graph illustrating the results on a State/Territory and national basis. The site was continually updated on referendum night as results were entered at divisional offices.
The results of the count were published simultaneously to three web centres in Canberra, Sydney and Melbourne. The three centres shared the large load of users who logged on to the site on referendum night so as to avoid delays in accessing information. On referendum night alone, over 154 000 users downloaded over 1.3 million pageviews on the virtual tally room, with the average user spending 45 minutes surfing the site. Peak use of the site occurred at 8pm with 5 000 users per second accessing the site. In the three weeks following polling day, more than 55 300 people accessed over 597 264 page views.
Based on these figures, the 1999 referendum site eclipsed the previous record for the 'largest live internet event in Australia!', with the previous record being the virtual tally room for the 1998 federal election.
The software and technology necessary to host the virtual tally room had been developed especially for the AEC for the 1998 federal election. It was further developed for the 1999 referendum with results screens appropriate to the referendum. The virtual tally room website for the 1999 referendum cost just over $214 000 to develop and run.
Utilising the internet to transmit electoral results is a key innovation which has been used successfully by the AEC at both the 1998 federal election and the 1999 referendum. It has greatly enhanced the access that the media, political consultants and other interested people have to timely progressive electoral results.