Research Report 3 - Analysis of Declaration Voting

Updated: 30 May 2013

Trends in declaration voting

At the 2001 Commonwealth election, of the 12 098 490 votes counted, 1 925 873 or 15.92 percent were declaration votes. This represents a 39.4 percent increase over the number of declaration votes counted in the 1993 election. By way of comparison there was only a 6.4 percent increase in ordinary votes over the same period. Despite accounting for only a small proportion of the total votes counted, declaration votes contributed 47.2 percent of the increase in total votes counted between the 1993 and 2001 elections. Table 1 shows a summary of declaration votes counted at the last four Commonwealth elections.

Table 1: Votes Counted, Senate Elections 1993 to 2001
  1993 1996 1998 2001
  No % No % No % No %
Declaration votes:
Absent votes 653435 5.97 657539 5.82 776859 6.70 780961 6.46
Provisional votes 60592 0.55 105091 0.93 116158 1.00 107396 0.89
Pre-poll votes 357301 3.26 434841 3.85 692377 5.98 585616 4.84
Postal votes 310579 2.84 359604 3.18 488671 4.22 451900 3.74
Total declaration votes 1381907 12.62 1557075 13.79 2074065 17.90 1925873 15.92
Ordinary votes 9572351 87.38 9737404 86.21 9513300 82.10 10172617 84.08
Total votes 10954258 100.00 11294479 100.00 11587365 100.00 12098490 100.00

The table shows a steady increase in declaration voting between 1993 and 1996 and 2001. There was a dramatic increase in declaration voting at the 1998 election. This increase may have been due to the timing of the 1998 election. The 1998 election was held on 3 October, during school holidays and on a long weekend in some states and territories.

Over the period 1993 to 2001 all categories of declaration votes have increased their proportion of the total votes counted. The most dramatic increases occurred with Provisional votes and Pre-Poll votes which increased by 77.2 percent and 63.9 percent respectively over the period. This compares to increases of 19.5 percent for Absent votes and 45.5 percent for Postal votes.

The rate of increase in Provisional votes appears to have slowed recently, with an increase of only 2.2 percent over the period 1996 and 2001, a rate of increase only half that of ordinary votes in the same period. On the other hand the rate of increase in Pre-poll votes and Postal votes has accelerated, with increases of 34.7 percent and 25.7 percent respectively.

The rate of declaration voting at the 2001 election was the second highest ever recorded in Commonwealth electoral history. The last three elections have recorded the three highest rates of declaration voting. Until the 1970s rates of declaration voting were generally below ten percent. However since 1974 the rate has not fallen below ten percent. Since the 1984 electoral reforms the rate of declaration voting has averaged nearly fourteen percent. A summary of votes counted by type of vote at each Senate election since 1901 is provided in Appendix 3.

Table 2 shows votes counted by type at the 2001 Senate election for the States and Territories and for geographic regions. Generally the rate of declaration voting is higher in the more populous States and lower in the less populous States, although the Australian Capital Territory has the highest incidence of any jurisdiction. The inner metropolitan region has the highest rate while the outer metropolitan region has the lowest rate.

Table 2: Votes Counted by Type, 2001 Senate Election
  Ordinary Votes Absent Votes Provisional Votes Pre-poll Votes Postal Votes Total Declaration Votes
New South Wales 83.95 6.62 0.92 5.24 3.28 16.05
Victoria 84.09 6.31 0.87 4.75 3.98 15.91
Queensland 83.45 6.28 0.92 4.49 4.87 16.55
Western Australia 84.22 8.25 0.82 3.88 2.83 15.78
South Australia 85.94 6.32 0.83 3.34 3.57 14.06
Tasmania 86.43 4.82 1.00 3.71 4.03 13.57
Australian Capital Territory 79.53 2.54 0.68 14.43 2.83 20.47
Northern Territory 85.35 2.17 0.79 9.28 2.41 14.65
Australia 84.08 6.46 0.89 4.84 3.74 15.92
Inner Metropolitan 82.80 7.17 0.95 5.30 3.78 17.20
Outer Metropolitan 85.19 6.61 0.82 3.93 3.45 14.81
Provincial 83.54 5.72 0.98 5.47 4.29 16.46
Rural 84.31 5.87 0.87 5.15 3.80 15.69

Detailed statistics on declaration votes at the 2001 Senate election for individual electoral divisions are at Appendix 4. The following table shows divisions with the highest and lowest proportions of declaration votes.

Table 3: Declaration Votes, 2001 Senate Election, Selected Electoral Divisions
Division Region percent Division Region percent
Melbourne Ports Inner metropolitan 23.30 Braddon Rural 11.49
Wentworth Inner metropolitan 22.79 Prospect Outer metropolitan 11.60
Sydney Inner metropolitan 22.23 Scullin Outer metropolitan 11.98
Maranoa Rural 21.91 Wakefield Rural 12.03
Melbourne Inner metropolitan 21.00 Fowler Outer metropolitan 12.08
Fraser Inner metropolitan 20.64 Bonython Outer metropolitan 12.14
Brisbane Inner metropolitan 20.56 Throsby Provincial 12.63
Canberra Inner metropolitan 20.29 Barker Rural 13.09
North Sydney Inner metropolitan 19.75 Calwell Outer metropolitan 13.13
Higgins Inner metropolitan 19.75 Reid Inner metropolitan 13.29

Given the requirements that have to be satisfied in order to cast a declaration vote (religious beliefs, travelling, illness, etc.) it is not surprising that electoral divisions that contain substantial numbers of electors with these characteristics are at the top of the list of declaration voters. Inner metropolitan divisions generally have a higher incidence of declaration voting while rural and outer metropolitan divisions generally have a lower rate.

Inner metropolitan divisions generally have higher proportions of high income earners who may be more likely to be travelling at election time while the reverse is true for some outer metropolitan and rural divisions.[4] Religious beliefs may also play a part in explaining some variations; as Commonwealth elections are held on Saturday (the Jewish Sabbath) it is probably no coincidence that the two divisions with the largest number of persons of the Jewish religion also have the highest proportion of declaration voters.[5] Ease of access to polling places in other divisions may also be a factor in explaining higher incidences of absent voting in inner city divisions.

The two Australian Capital Territory divisions of Canberra and Fraser have pre-poll voting rates substantially higher than any other divisions. As the ACT electorate has higher incomes than average[6] and is probably more mobile it is not unreasonable to expect considerable numbers of ACT electors to cast pre-poll votes before travelling interstate on the weekend.

The Queensland rural division of Maranoa, however, is one division that does not fit the above scenario. Maranoa has the highest incidence of postal voting in Australia, 10.4 percent, nearly twice the rate of the next highest division. Maranoa is a large rural division without any substantial urban centres and with pattern of settlement that makes remote mobile polling facilities impractical. As such a considerable number of electors in Maranoa do not have ready access to a polling place. As a consequence there are more general postal voters registered in Maranoa than any other division, thus contributing to the high incidence of postal voting in the division.

4 At the 2001 Census, 35.4% of families in Inner Metropolitan electoral divisions had a family income of $1500 or more compared to 26.8% of families in Outer Metropolitan divisions and 19.1% in Provincial divisions and 16.2% in Rural divisions. See A. Kopras, 'Electorate Rankings: Census 2001', Research Paper , No. 2 2002–03, Department of the Parliamentary Library, Canberra, March 2003

5 At the 2001 Census, 13.4% of the population of Wentworth and 13.0% of the population of Melbourne Ports stated their religion as Jewish. Australian Bureau of Statistics, 2001 Census of Population and Housing.

6 Kopras 2003