By the end of 2007, it is possible that the Liberal Party will be out of government throughout Australia. As I noted in part 1 of this post, this has led some people to forecast its demise as an organisation. Like political parties around the world, the long-term trend in Liberal membership is down. But unlike political parties in many other parts of the world, the major Australian political parties have retained large support bases, people who tell pollsters that they ‘generally think of themselves as’ Liberal, Labor, or whatever.
Quite surprisingly, given all that has occurred since, a larger proportion of the electorate generally thought of themselves as Liberals in 2004 than they had in 1967, the first time the question was asked and a year after the Coalition’s biggest ever share of the two-party preferred vote, at the 1966 federal election. 40% were Liberals in 1967, 41.5% in 2004. Because Country Party/National Party identifiers have shrunk from 7% to 3% of the electorate in that time the proportion of Coalition identifiers is down slightly, but the Liberal Party itself has held up very well (though fewer voters overall class themselves as ‘strongly’ preferring their party, with 29% of Liberals strong supporters in 1967 and 21% in 2004). Labor is down 5 percentage points, from 37% to 32%. There was a lot of talk in the 1990s about the rise of minor parties, but in reality the two major parties have proven to be highly durable.
Though Liberal-leaning voters cannot be taken for granted – more on that below – they are a good foundation on which to build towards a majority. While Liberal infighting may damage the party’s electoral prospects, the large Liberal support base shows why it is still an institution worth fighting over. Norman Abjorensen may be right that in theory there is scope for realignment in Australian politics, but in practice the voters aren’t likely to pay enough attention to make it work. The Liberal brand has value, independently of who its key figures at any given time are or what it stands for at any given election. Most people who have voted Liberal at recent state elections will vote Liberal no matter how unimpressive the party’s performance. The trick is in getting enough non-aligned voters and weak supporters of other parties to vote Liberal to secure victory.
There is little doubt that this task is getting harder. The problem for the Liberals, as I argued in a pessimistic post after the 2004 election, is that while large their support base is ageing. Looking at the 2004 Australian Elelection Survey by generations, among those born up to the end of WWII 48% identify with the Liberals, among the baby boomers (to 1963) 40% identify with the Liberals, among Generation X (to 1979) 37% identify with the Liberals, and among Generation Y (1980-) 34% identify with the Liberals (though Y was a small sample). Labor has a similar problem, but it can get support via Green preferences in the younger groups.
This doesn’t mean that the Liberals can’t win, but it does mean that they find it more difficult to win over younger than older voters. In the 2004 AES, the proportion of people in the WWII and before generations who voted Liberal was nearly the same as the proportion who identified with the Liberals. But for the baby boomers, 3% more voted Liberal than identified with the Liberals, among Gen X 9% more voted Liberal than identified with the Liberals, and among Gen Y 6% more voted Liberal than identified with the Liberals. But the Newspoll two-party preferred results suggest that in 2004 among the 18-34 year olds the Coalition was getting low to mid 40s, mid 40s among the 35-49 year olds, and mid 50s in the 50 years and over group. They are crucial to Liberal victories. No wonder the Howard government is spending so much on health.
The Liberals main electoral targets are the 16% of so of people who say they do not identify with any party and soft supporters of the other parties. The Howard government has, so far, been successful in attracting their support, but state Liberal oppositions have failed to do so. Partly this is probably due to mummy party/daddy party considerations, with the federal sphere playing better to Liberal strengths in economic management, defence and immigration. But incumbency is important in providing opportunities to communicate with the people who aren’t that interested in politics and, if things are going ok, becoming the ‘safer’ alternative. But once an election is lost it is hard to regain those people’s attention and restore their confidence.
So an election defeat in 2007 would be very bad news for Liberals, compounding the problems caused by demographic trends. Yet even if the current relatively weak Liberal support in younger groups persists as they age, on current trends they will still have the largest support base of any party, with left-leaning voters split between the ALP and the Greens and a quarter of voters unaligned. Their base is probably enough to keep the Liberals in the game if they can make themselves look like reasonably competent alternatives to Labor governments that have stuffed up. In the 1980s and early 1990s people were writing the Liberals off, but in the mid-1990s Keating’s failures and Howard’s political discipline brought them back for at least eleven years. It’s happened before, and it can happen again.
Andrew Norton 
The figures about generational identification with the Liberal Party is interesting (and in according with what people generally expect), and a point pertinent to your analysis is whether there is an age dependency in the identification as well as, or instead of, a generational identification, ie, are people more likely to identify with the Liberal Party as they get older?
If so, then it will be easier for the Liberal Party to win elections. My guess is that there would be some age component, but I couldn’t guess the extent of it.
In any event, either of the major parties can win elections provided they win enough of the non-committeds.
LikeLike
Sacha – In the original Catallaxy post I tried to track the same generations over time using old surveys, and concluded that there was evidence of generational effects, ie particular age cohorts tending to vote one way or the other. In this post, I looked at identification with a party, which I think would be more stable than the vote, which will always include people who have cast their ballot for ephemeral reasons.
There may be some ‘life cycle’ effects; abandonment of youthful radicalism being a common one. But even if the Libs gain 5 or 6 percentage points on Gen Y they would still be well behind their base support in the current 50+ group.
LikeLike
Andrew, I agree with all of your last comment. Older people are a plentiful source of Liberal votes at the moment.
Imagine 20 years in the future, when there are relatively more older people (the oft-quoted ageing of the population). If more voters identify with the Liberal and National Parties and the parties that generally preference LP/NP than identify with the ALP and the parties that generally preference the ALP, then the LP/NP will probably win elections more easily than the ALP, and vice-versa. I’d suggest that it’s the relative identifications (coalition and coalition preferencing parties versus ALP and ALP preferencing parties) that matter.
LikeLike
Andrew, do you have the generational ALP and Green identification figures to compare with the coalition figures?
LikeLike
Sacha – I have them at home, I will post them tonight.
LikeLike
I don’t think there is ever any long-term demographic trend away or towards any party simply because such trends are towards policies rather than parties. Parties, in the longterm, shape their policies to fit such trends.
The interesting questions are where demographic trends are sending policy. I’m not keen on having the median voter being an old fogy.
LikeLike
DD – I’m not sure of this; I think there is significant socialisation into party loyalties that has only a vague connection to changing policies. There is evidence that policy views are driven by party loyalties, rather than the other way around.
LikeLike
I recall reading in my undergrad studies that there are studies showing party loyalty socialisation with family a huge influence (not surprisingly).
Hence across each of the major parties you have a wide range of policy positions – the policy positions aren’t neatly divided between the parties – eg you don’t have a free market party and a non-free market party, a socially liberal and a socially non-liberal party, although the “concentrations” of each tendency differ in different parties.
LikeLike
Sacha – The other generational party ID figures:
WW2 and before: ALP 34, Green 2 (clearly a change there, since the Greens did not exist when they were young)
Boomers: ALP 36, Green 4.5
Gen X: ALP 29, Green 7
Gen Y: ALP 22, Green 13.5
On the vote in 2004
WW2: ALP 35, Green 4
Boomers: ALP 39, Green 8
Gen X: ALP 34, Green 10
Gen Y: ALP 32, Green 17
LikeLike
DD, Sacha
On party socialisation: in the 2004 AES, about 30% of the sample did not know how their parents voted. Taking out those who did not know, of people with Liberal ID 55% had a father with Liberal ID. If we count fathers who were National (and predecessor party) supporters, 62% of the Coalition vote is inherited.
The effect is stronger on the Labor side: 75% of people with Labor ID had fathers who also had Labor ID.
What these statistics show is that Liberal voters (26%) are more likely to come from Labor families than Labor voters are from Liberal families (14%).
This is possibly some upward social mobility, but also Labor’s weakening grip on the working class. In the 2004 AES, the two major parties were almost even in their share of votes from self-described working class people: Lib 41.6%, Lab 42.3%.
LikeLike
I may be reaching, but it strikes me that the high number of Liberal identifiers in older and low in younger Australians may have something to do with the current PM. And that his replacement with a Costello or a Turnbull might lead to some re-alignment.
Possibly over-interpreting and giving too much weight to ‘leadership’ effects. Just a thought.
LikeLike
Thanks Andrew.
LikeLike
Andrew, you may or may not be right about party identification rather than policy preference, being socialised. But those figures don’t demonstrate it because they don’t rule out the possibility of policy preferences being what is socialised and party identification merely following from those preferences. You need an instrument to let you work out how good a proxy that is.
LikeLike