I think I see what you are getting at FG but I am not sure I agree - the AEC figures are suggesting a decline even now that most of the vote is counted.
Also, given that the AEC made a massive effort to ensure people were enrolled (including a door knock and several millions spent in advertising) as well as enrolled in time there is the question of why so many people were still not on the electoral roll and still failed to vote. Is there a growing concern about the electoral roll itself perhaps? Is privacy a concern? Is our population now too mobile to support such a roll?
There is also the question of why some of those who did vote failed to do so in a manner which meant their vote was valid. There will always be people who are incapable of understanding the process but are there others protesting as well? Is it democratic to permit the intellectually impaired to vote when they do not understand the process and vote because a candidate 'looks nice' or to pressure a new citizen who is fearful and suspicious of the entire process to make a decision when they do not trust any candidate?
Both major parties have shied away from any discussion about 'compulsory' voting. We are told it is democratic and the assumption is that it is democratic and a 'good thing' but is it really? It is an issue which needs robust public debate but one even the media seems reluctant to tackle. Perhaps we need to start a new thread?

Communicat,

After some fresh research since my last post, I emphasise that voter turnout can not have declined in relation to levels of former years if the AEC VTR progressive count tabulations for 2007 contain correct ballot issue figures. On the contrary, turnout, when calculated in the manner I have outlined in my last post, is very comparable with the 2004 turnout for the sample of seven Divisions listed earlier in the thread.

My previously estimated approximate turnout for Dobell in 2007 of 94.3% compares favourably with the final figure for Dobell turnout in 2004 of 95.07%. When figures used from VTR pages updated at 3:12:02 PM on Thu 29 Nov 2007 are used, it is theoretically possible for turnout to reach 96.47% in Dobell.

The total valid ordinary vote of 64,499 obtainable from the 'Two Candidate Preferred by Polling Place' page by adding totals from all polling places, plus the 3,176 informals, comes to 67,675 votes. When this is added to the total declaration vote envelopes received less preliminary rejections, 19,464 votes, you get a total of 87,139 as the maximum possible vote in Dobell. When this theoretical maximum is divided by the 90,328 enrolments at roll close, you get 96.47% maximum apparent turnout. This will decline somewhat if there are rejections of provisional vote claims, postal votes received too late, or other declaration vote rejections.

The informal vote in Dobell in 2004 was 7.41% of the total vote cast. It looks like the Dobell informal vote in 2007 will be around 4% when finalised.

To update the picture for the other six Divisions I spot checked, voter turnout in 2004 for Charlton was 95.73%, Lyne 95.62%, Newcastle 94.72%, Paterson 96.08%, Robertson 95.38%, and Shortland 95.54%. The theoretical maxima this time are 96.82%, 96.08%, 96.39%, 99.09%, 88.42%, and 95.47% respectively.

Only Robertson appears anomalous, due to seemingly incomplete reporting of declaration vote issues.

No way has there been a decline in turnout this election!

Why did I say what I did in the 26th post in this thread, the one beginning with "(continued)"?

I made the now seemingly stupid claim "For the first time anywhere in Australia, to my knowledge, more vote claims have been made than there are names enrolled in an electoral Division!"

I made this claim because of what I saw posted to the AEC VTR when I started looking on Tuesday 27 November 2007. When I first looked, the 'Declaration Vote Scrutiny Progress' pages were not showing any count figures. There is provision for count figures on the page, but only zeros were showing. The corresponding 'First Preference Two Candidate Preferred' pages I took to at that point and in that circumstance to be a record of the ordinary vote, the vote cast in polling places with its accountancy finalised on election night in the presiding officers' returns.

Both page types showed the same update time in the case of each Division spot checked. Despite this, it appears that there had been some counting of declaration vote included in the totals showing at the foot of the pages, if not the inclusion, without separate identification, of informal votes. At this early stage I was not taking screenshots, but I noted down the following totals then showing of what I took to be the valid ordinary vote, and declaration vote issues for the seven Divisions spotchecked:

Charlton 75,354 12,430

Dobell 69,644 20,522

Lyne 72,009 16,724

Newcastle 75,216 12,578

Paterson 71,609 21,089

Robertson 75,458 8,137

Shortland 75,184 18,637

The cached pages of progressive postings to the VTR will reveal the accuracy of these transcriptions, in the event there is inquiry into the basis upon which such a presumably disturbing claim could have been made.

Informal votes are part of the ordinary vote, and are known on election night. They should have been identified in all updates to the VTR made after the Saturday night. I don't recall seeing informals identified when I started looking on Tuesday.

dibs' post (the 17th in this thread) deserves some comment.

dibs is bang-on with respect to the methodology for determining just what proportion of Australians failed to vote.

I suspect the figure used by dibs, of near to 75% of the population as being eligible to enroll electorally, may be a little high. I think the right figure would be nearer 65%. The reason I say this is because a figure of this order is indicated in a study of Australian electoral enrolment levels covering the years 1947 to 1987 submitted to the Parliamentary inquiry into the conduct of the 2004 Federal elections. Whilst this study stops at 1987, the age distribution and citizenship status of the population, the two determinants of eligibility for enrolment, would probably only have moved slightly upward since the 1987 figure of 62.74% given in that study.

It is common knowledge that there was an extensive (and expensive) advertising campaign undertaken by the AEC in the year before the elections to encourage people to enroll. There exist studies that show significant failure to take up electoral enrolment among the newly age-qualified part of the population. It is accepted that there was a significant level of failure (or refusal) to enroll in the community at large. That's why they ran the advertisments.

dibs' maths, when applied to the eligible proportion of the population indicated in the study referred to above, reveals a conundrum. 65% of the Australian population of 21.1 million comes to 13.72 million. dibs tells us (correctly, from official figures) that 13.65 million names are carried on the electoral rolls. Near as dammit to 100% enrolment, if the 65% of population being eligible is near the correct figure. If you believe it.

AEC studies over the last 20 or so years have indicated only around 85% of the eligible actually enroll.

So how do the extra names get to be on the rolls?

The tune "Colonel Bogey" echos through my head. It has another name in Australian military musical circles. Not "Bridge on the River Kwai".

I should have given the link to the enrolment level study I referred to in my previous post. The word limit beat me.

http://www.aph.gov.au/house/committee/em/elect04/subs.htm

This link is to a list of submissions to the 2004 Inquiry into the Conduct of the 2004 Federal Elections. Click on submission No. 123. This will give you a PDF document which includes a table titled "Australia - Aggregate Enrolment Levels 1947-1987". The relevant figures are in Column J, headed "E%P", which is explained as being ".... the ratio of total number of persons estimated as being eligible to enroll to the total population, normally expressed as a percentage of the latter, but in Column J expressed as a decimal fraction for calculation purposes.". Line No. 122 of the table gives E%P for 30 June 1987 as 0.6274, which, being converted, is 62.74%.

There is no line in the AEC VTR tabulation of "Two Candidate Preferred by Polling Place" for informal votes counted in the ordinary vote, a count that is finalised on election night. Why not?

Informals instead are shown in the tabulation of the progressive counting results titled "First Preferences and Two Candidate Preferred", but this total may include informals from the declaration vote as well as those from the ordinary vote. The sum of the valid ordinary vote and the informals subsequently posted does not equate to what I transcribed from the progressive count pages on Tuesday.

With respect to Paterson, the valid ordinary vote of 65,528, plus 2,793 informals as at 3:16:30 PM Thu 29 Nov 2007, does not equate to the 71,609 total I transcribed on Tuesday when I commenced my spot check of seven Divisions. It falls 3,288 short, but at the time of transcription, to my recollection, the "Declaration Vote Scrutiny Progress" table showed no declaration vote counted!

Like discrepancies exist with respect to all seven Divisions' Tuesday transcriptions.

Was the VTR website cracked? Were these transient discrepancies 'messages' as to what extra vote could safely yet appear?

The voter turnout at the Federal elections, according to the AEC Virtual Tally Room website, continues to rise.

From an average turnout of 79.36% for NSW as at 9:06:23 PM on Tue 27 Nov 2007 (then shown on the 'Turnout by State' page linked to), the seemingly very low figure that sparked this discussion, the turnout as at 10:22:47 PM on Mon 3 Dec 2007 stands at 86.30% for NSW.

The State average lags significantly behind the turnout percentages currently posted for many individual Divisions within this State. Whether this is because some Divisions have experienced a relatively low turnout, or because in some Divisions the counting of the declaration vote has been slower than in others, remains to be seen.

One can only wonder what this curious presentation of claimedly comparative turnout figures was meant to achieve. Could it have been intended to convey the impression to the public that there had been a relatively low turnout, when in fact the turnout is in truth (at least in some Divisions) perhaps greater than previously experienced?

What would have been the point of conveying such an impression? Could it be that if there was an impression of relatively low turnout, less attention might be paid to the progress of the count, or to apparent anomalies that might reside therein? Or could it be something else?

The turnout in Dobell, for example, although showing as being 87.95% as at 5:00:49 PM on Mon 3 Dec 2007, on figures displayed at that time in the 'Declaration Vote Scrutiny Progress' page taken in conjunction with the information on other pages, indicates a ceiling turnout of 97.31% based upon declaration vote received. If the turnout ceiling was to be based upon declaration vote ballot paper issues, then the ceiling would be 98.05%.

Clearly, some of the declaration vote issues may never be returned, and some of it may well be properly rejected at the scrutiny yet to be completed. It would nevertheless appear that turnout in Dobell will be in excess of 95%.