The modest republican constitutional change proposed in the November 6, 1999, referendum was hardly the most significant political question facing Australians in recent times. Nevertheless the results provide a very useful snapshot of a changing Australia.
The results were actually much better for the republic than most of the media would admit. A 46.5 per cent Yes vote for a republic, first time up, is a very good result when you consider that British-Australia was still celebrating Empire Day about 30 years ago, and when you remember the enormous grip all the hype about the British royal family still had in the 1940s, 1950s and 1960s.
Most older Australians can remember being bussed as schoolchildren to showgrounds during royal visits to stand in the hot sun waiting for the Queen to pass by. For most of the period since white settlement, the Australian establishment has energetically promoted the monarchical British connection as an invaluable support for the hegemony of the ruling class in Australia.
To better understand the results, I have studied the detailed figures, booth by booth, for all the seats in NSW, and the national results for five categories of votes. The following analysis is based on my examination of these results, supplemented by some useful figures supplied by Mick Armstrong in the magazine, Socialist Alternative, published by the group of the same name, which was one of the socialist groups with sufficient understanding of the class forces at work in the referendum to very sensibly advocate a Yes vote. Mick Armstrong's article is very useful and a lengthy quote from it is worthwhile here:
Indeed it has much in common with the Hanson phenomenon. Significantly, the No vote in the referendum was highest in those rural areas where One Nation polled well in the last federal elections. The three seats with the highest No vote were the seats with the highest Hanson vote in the last Federal election - the Queensland rural seats of Maranoa (No: 77 per cent, Hanson: 22 per cent), Hanson's own seat of Blair (No: 75 per cent, Hanson: 36 per cent) and Wide Bay (No: 75 per cent, Hanson: 26 per cent). This pattern was replicated outside Queensland.
In NSW, Victoria and Western Australia the seat that topped the state No vote also had the top One Nation vote: Gwydir, NSW (No: 75 per cent, Hanson: 21 per cent), Mallee, Vic (No: 72 per cent, Hanson: 13 per cent), O'Connor, WA (No: 72 per cent, Hanson: 14 per cent).
Similarly, the outer suburban areas with the highest No votes had above-average support for Hanson: Canning in Perth (No: 68 per cent, Hanson: 14 per cent), Bonython in Adelaide (No: 67 per cent, Hanson 15 per cent), Oxley in Brisbane (No: 66 per cent, Hanson: 18 per cent), Werriwa in Sydney (No: 58 per cent, Hanson: 12 per cent).
By contrast the Yes vote was strongest in areas most resistant to the appeal of Hansonism: the core working class suburbs of Melbourne and the inner suburbs of Sydney, Brisbane and Canberra.
One misconception propagated by the media is that the Yes vote was strongest in better-off Liberal electorates. It is true that Sydney's wealthy North Shore voted Yes and that the republic was narrowly defeated in some Labor seats in Sydney's outer west. However, in NSW two-thirds of the seats that voted Yes were in the working class areas of Sydney, Newcastle and Wollongong.
Nationally, the five seats with an overwhelming Yes vote were safe Labor seats, headed by Melbourne with 71.5 per cent, Sydney 68 per cent, Melbourne Ports 66 per cent, Fraser (ACT) 65 per cent and Grayndler in Sydney 65 per cent. And the Yes vote in these seats was well above that in super rich Toorak.
Nearly two-thirds of the 30 seats with the highest Yes vote were Labor seats. These were not simply inner-city "chardonnay socialist" areas but included the core working class areas in Melbourne's western and northern suburbs.
The working class Yes vote was strongest amongst non-English-speaking migrants and slightly better-off workers but lower in the poorest, most depressed sections of the Anglo working class. In Melbourne it tended to be the marginal outer suburban seats with fewer non-English-speaking migrants and a larger churchgoing Protestant middle class that voted No. So while there was not a totally clear-cut working class Yes vote, the No vote was concentrated amongst the sections of the population most easily swayed by populist appeals: the rural population, the outer-suburban middle class, the less unionised and class conscious workers, older people and traditional Anglo-Australians.
In addition to the points that emerge from Mick Armstrong's analysis, a number of other points emerge from my own investigations. The Australian Electoral Commission has five special categories in each electorate, in addition to each booth, of which there are usually between 30 and 50. The five special categories are:
· absentee votes (cast on election day in other electorates).
· pre-poll votes (cast by arrangement before election day, usually because of travel commitments on election day).
· postal votes, routinely made available by the Electoral Commission to elderly or housebound people.
· special hospital mobile teams (votes cast in hospitals, nursing homes and aged care facilities on election day, again in practice, with a heavy predominance of elderly people)
· votes cast at the capital city Town Hall polling booth in each state where a number of votes are cast on election day for each electorate.
The result for these five categories is very illuminating. Both the postal votes and the special hospital votes show a much higher No vote than the general vote in each electorate, even in the electorates that strongly voted Yes. This No vote is most pronounced in the hospitals.
On the other hand, absentee voters, people who voted outside their electorate on voting day, showed a significantly higher Yes percentage than the general vote for their area. Pre-poll votes, the votes cast by arrangement before the election, averaged nationally about the same Yes vote as the national average, being a little less than the average in the Northern Territory and Victoria, roughly the same in NSW and the ACT and dramatically higher in WA, Queensland, SA and Tasmania.
The last grouping, the small but significant sample of people voting outside their own electorate at capital city Town Hall polling booths, shows by far the highest Yes vote of all. An obvious inference is the existence, in the referendum result, of a strong tendency for younger cohorts of voters voting Yes and older cohorts voting No.
Those voting at the capital city Town Hall polling booth for other electorates are obviously people who get around, and they are probably younger people. Anyway, they clearly show a pronounced Yes bias. The overwhelming result for No in the special hospital team votes and the postal votes clearly suggests a very heavy vote against the republic in older age groups. (A significant group amongst the postals, in addition to the aged, are people who are housebound for other reasons, such as disability. Probably some features of the situation of being housebound, such as being exposed to a steady diet of talkback radio, has a conservatising effect on voting patterns on an issue like the republic.)
The tendency for the young to vote Yes and the old to vote No is confirmed indirectly in another way. For the Yes vote to have done as well as it did, it emerges clearly that, to counterbalance the No vote among the old, younger age groups must have voted solidly Yes, including the "young fogies" of Generation X and younger, who conservative pundits desperately hoped were in a deeply conservative frame of mind.
This alleged conservative mood among the young didn't show up in the republic referendum results at all. The story that a large number of the young voted No is a conservative invention, not backed up in any way by the actual results. The electorate-by-electorate pattern confirms the general observation made by most electoral observers that people with tertiary education voted Yes very heavily.
This question of levels of formal education is somewhat intertwined with the age factor. As the steep rise in the number of Australians with tertiary education has taken place progressively over the last 30 years, the cohort of Australians in the age group, say of 55 and above, is the same cohort where the proportion of people with tertiary education is far less than in younger cohorts.
It also ought to be said that the older cohort are also the cohort whose whole lives were moulded in the rabidly royalist British-Australia of the 1940s, the 1950s and the 1960s, many of whom have an entirely natural human nostalgia for the period of their youth, which seems to translate electorally into a certain reluctance to vote for a republic. Natural demographic evolution will inevitably reduce the electoral impact of this cohort over time.
In parliamentary elections, postal votes and hospital votes are an area of fierce contest between the campaigning political machines of candidates for different parties, whose interests are directly involved. All the anecdotal evidence suggests that on the occasion of the republic referendum this very sharp intervention by the various political machines was minimal because their electoral interests were not directly affected.
Consequently, the overall result in these two categories can be taken as a reasonable indicator of the viewpoint of these categories of voters on the referendum questions. The votes in these five categories, examined above, total about 1.8 million votes nationally, or about a fifth of the votes cast in the referendum, so the variations revealed are quite significant.
The variations in the voting pattern between the five special categories are of considerable interest. Among other things they clearly highlight the age factor in the results. An even more fascinating inference is what one might call the mobility factor. Greater republican inclination appears to be associated with greater mobility.
The absentee voters, many of whom voted in electorates quite close to their own electorate, show a higher Yes vote than the vote in their electorate. Voters who are even further away on voting day, visiting the state capital, show the highest Yes vote of the lot. (Possibly people who work outside their own electorate on Saturdays, and therefore vote absentee, also have a greater republican bias.) So, on the face of it, the further you travel, the more likely you are to vote for a republic, which is a new and rather novel concept in political science.
There is no question that there was a strong Yes vote from most non-British migrant communities, including most second and third generation people of migrant background. In Sydney this was particularly apparent, with all the Labor seats having a large ethnic component, even seats like Lowe and St George, where the ethnic component is mainly older, more established and affluent people of second and third generation Italian and Greek background, voting solidly Yes.
This is also one of the major explanations for the extraordinarily high Yes vote in metropolitan Melbourne, where recent migrants and second-generation ethnics are fairly evenly distributed in almost all areas and are not concentrated so strongly in particular regions as they are in Sydney.
There is also a very high component of first, second and third generation Greek and Italian Australians scattered all over Melbourne, which has a very high proportion of migrants. Of the 20 Melbourne electorates, 17 voted comfortably Yes, with very high Yes votes in working class areas. The only three Melbourne electorates that voted No were outer-suburban electorates with fewer migrants and ethnic Australians.
All of this suggests that the widely distributed cultural weight of migrant ethnicity was a major factor in the very strong Melbourne Yes vote.
In NSW the contrast in the results between the Newcastle and the Illawarra-Wollongong areas was very informative. Newcastle, a working-class area, with a number of Labor seats but proportionately a much lower number of migrants and people of migrant background, showed a very bad result for Yes. The only electorate that voted Yes in this region was the Newcastle electorate itself, by a very narrow margin.
Newcastle is the Hunter Valley electorate in which tertiary educated people are most heavily concentrated.
On the other hand, the story was dramatically different in the Illawarra region, an area where there is a very high migrant and ethnic population, perhaps the highest proportionally in the whole of Australia. The electorate of Cunningham, the main Illawarra electorate, showed an overwhelming Yes vote, both in the more affluent suburbs north of Wollongong, where there are more tertiary educated people, and in the strongly ethnic working-class suburbs south of Wollongong.
In Cunningham, it is clear, both major social layers: blue-collar workers of whom, these days, a very high proportion are migrant workers; and tertiary educated people, voted Yes. In the next electorate south, Throsby, there was a No majority, but it was derived mainly from a strong No vote in the Southern Highlands area, where there are few migrants, and where a generally affluent Anglo middle-class and rural mood prevails.
A number of the booths in the northern part of Throsby, which are in outer-suburban working class suburbs of Wollongong, with a large migrant component, voted Yes. The different and contrasting results in Newcastle and the Illawarra underline the significance of migrant ethnicity in the results.
Even in metropolitan Brisbane, the capital of conservative Queensland, there was a strong Yes vote, and here again there is a clear association between a Yes vote and two elements: firstly, migrant ethnicity, and secondly, tertiary education. At this point it is worth saying that by my reading of the results, there was a majority Yes vote in descending order of magnitude, in Canberra, Melbourne, Sydney, Wollongong, Brisbane and Hobart, with a majority No vote in Perth, Adelaide, Newcastle, Geelong and Launceston. The more heavily urbanised, cosmopolitan cities were the centre of the Yes vote.
In Sydney there was a striking geographical divide, starting at Bobbin Head, going down to Baulkham Hills then through the middle of Parramatta, down to the northern outskirts of Liverpool and across from Liverpool, past the affluent Anglo suburbs north of the Georges River, and hitting the Georges River at about Tom Ugly's Bridge, then out to sea. The electorates, Liberal or Labor, to the east and north of this divide voted Yes, and the electorates south and west of it voted No, although there were strong pockets that voted the other way in all these areas.
There were some striking but significant local idiosyncracies. Often distinctively individual, slightly isolated communities, with a strong local identity and a larger old, established Anglo component, seemed to vote heavily No. Two examples that jumped out at me were Kurnell in Sutherland Shire, which voted almost two thirds No in fairly sharp contrast with the rest of that electorate, where the No vote was lower. Another striking example was Riverstone-Schofields, an old working-class, largely Anglo community, where the meat works was closed some years ago, which showed a No vote approaching 70 per cent, much higher than the No vote in the rest of that electorate, a Labor electorate, where Yes did quite well in the other areas.
These kinds of results suggest strongly that there is some truth in the proposition that pockets of traditionally Labor-voting people who exercised a strong No vote were often expressing a fairly sharp social protest against the political class, against economic and political elites and against the fact that not much has been done for them lately.
On the Tory side of the usual electoral divide, the break from the suburban North Shore to semi-rural kind of activity at Baulkham Hills is the sharp divide in the republic referendum result. Semi-rural areas and, once again small distinctive Anglo communities such as Richmond, Windsor, Castle Hill, etc, were strong No areas, whereas the dormitory North Shore voted fairly solidly Yes.
The Yes vote on the normally Liberal-voting North Shore was quite high, but not as high as the Yes vote in the Labor electorates, where the Yes vote had the majority. Once again, education obviously has a bearing. The North Shore electorate with the lowest Yes vote was Bronwyn Bishop's electorate of McKellar, which stood out from the rest of the North Shore, with an almost 50:50 split between Yes and No.
When you look at the Bureau of Statistics breakdown of Sydney, the Northern Beaches area, which comprises Bronwyn Bishop's electorate, has a high concentration of self-employed tradespeople and contractors. Another Anglo area where there is a strong concentration of self-employed tradespeople and contractors, intertwined, however, with people with tertiary education, are the three subdivisions in Daryl Melham's Labor electorate of Banks, just north of the Georges River.
In federal and state elections Labor wins these subdivisions with a lowish margin, much smaller than the margin in the rest of Banks. In the republic referendum, in which Banks as a whole voted No by a significant margin, these very affluent Anglo subdivisions showed a very substantial No majority. (On the other hand, in Melham's electorate, subdivisions such as Penshurst, with a large Asian community, voted solidly Yes.) This patchwork of voting patterns suggests strongly that people such as self-employed tradespeople, contractors and Anglo small-business people very largely voted No.