Over at LP, Mark Bahnisch asks whether the government’s attack on union power under Rudd will work:
There’s an unexamined premise in commentary about this tactic of the Government – that unions are wildly unpopular. But how true is that? Unfortunately, there is no time series data on union sympathy. But there are three large-scale surveys conducted this decade that reveal some fascinating results.
Actually, there is time series data on union sympathy – I reported on nearly twenty years of fairly consistent questions in this 2005 article. On the the issues of whether unions have too much power and whether there should be stricter regulations of trade unions there is a clear decline in hositility towards unions. All the polls prior to 1990 (including earlier polls with different questions that I did not show in that article) showed between two-thirds and three-quarters of respondents thought that unions had too much power. This century, less than half of respondents have thought that.
This accords with changes in objective conditions. In the worst year of union havoc inflicted on Australian society, 1974, a staggering 6.3 million working days were lost to strikes. In the year to March 2007, there were just 109,500 days lost. It is unsurprising that there has been a change of mood among voters, whose lives are no longer disrupted by unions almost every week – or least that’s how it seemed when I was growing up in the 1970s and early 1980s.
Mark reports on some recent polls finding some positive views about unions, especially their role in the workplace – worth distinguishing, as Mark does, from their potential influence on politics generally. But the government is probably right that there is political mileage left in reminding people of the bad old days of union power.
The best way to assess this is to take out of surveys people who identify with Labor or Liberal, both because they are most likely to vote according to their underlying party ID, and because they hold predictable views on unions – in the 2004 Australian Election Study only 8% of Liberals disagreed that unions have too much power, and only 19% of Labor voters agreed that unions have too much power.
Of those who give no party ID in 2004, a third think unions have too much power, and nearly 40% neither agree nor disagree. Not entirely consistently, slightly more think that there should be stricter laws to regulate trade unions, with 46% neither agreeing nor disagreeing with the proposition. With a clear majority of unaligned voters either already against or undecided on unions, there is scope for a scare campaign. It won’t be as powerful as it would have been 20 years ago, but this old Liberal theme is not exhausted yet. Expect to see Dean Mighell and CFMEU thugs feature in Liberal advertising.
Andrew Norton 
Thanks, Andrew, I couldn’t find the AES data. I only had about an hour or two to research the piece! Strictly speaking the other surveys you cite aren’t time series, because they’re not conducted by the same organisations using the same questions and comparable samples.
LikeLike
Also, again being pedantic, the “too much power” question is really a proxy for union sympathy. It’s not measuring quite the same thing.
My attention has been drawn to a question in Nielsen recently which probably gives us the best info to go on for how things might play out this year:
http://www.theage.com.au/news/opinion/trust-v-economics/2007/06/15/1181414545978.html?page=3
LikeLike
Mark – The 1990 to 2004 data is some pollster same question, the 1987 data is same pollster and the same basic question with slightly different wording. It’s good enough for a graph in which nothing turns, unlike an election, on minor differences. I have similar questions and similar answers from Saulwick as well for the earlier time period, which gives me confidence that the results are robust.
That Rudd result from Nielsen is similar to the AES 2004 in agreeing that unions would have too much power, 38%/41%. The difference occurs on the other side of the result. In 2007, 53% disagree and in 2004 28% disagree. The similar agree result could be rusted-on union haters who will vote Liberal anyway, and the increase in disagreeing could be people confident that Rudd would not take us back to the 1970s or 1980s. On the other hand, it could be that the AES is a mail-out survey with a printed option ‘neither agree nor disagree’ (31% in 2004), while Nielsen is a phone survey which did not give such an option but recorded a ‘don’t know’ from those who were unable to give an answer or refused to commit to one option or the other.
This is why I like multiple pollsters exploring the same issue in slightly different ways, to get a feel for what role the survey itself is playing in shaping the results.
‘Union sympathy’ was just my way of trying to avoid word repetition; the real issues are what the questions ask, about union power and whether respondents favour their regulation. A voter could have personal sympathy for a fading union movement without supporting political measures to revive it.
LikeLike
Start with Combet’s ‘we should run this country’ line, THEN show Mighell, Reynolds etc.
Finish with scary voice: ‘Greg Combet is Labor candidate for the seat of Charlton. Do YOU think unions should run this country?’
It’ll be ugly.
LikeLike
That footage I saw on the 7.30 Report the other night of Reynolds walking along a footpath saying “Faaark ooorf.” to the cameras is priceless, heh.
LikeLike
Andrew, I agree that although there is diminishing hostility to unions, there is still some political mileage for the Howard Government on the double issue – too much power and too much influence on the ALP.
But you are such a fund of knowledge on polling, Andrew, you might be able to answer this question. Have Australians been asked: does business (or big business) have too much power and too much influence on the Liberal Party? If there is such information for a long period, is there a trend to more or less hostility?
I ask this question because, although it is much more dangerous for the ALP to imply sinister business influence, two can play at the same game.
LikeLike
Fred – I am not aware of this question being asked. The only time I am aware of the issue appearing is in the Australian National Political Attitudes surveys of the late 1960s and 1979, which included open-ended questions about likes and dislikes of the major parties. About 5% nominated business influence for a reason for not liking the Liberals. Given the reliance on stereotypes in public opinion, I am sure it would be much higher if a direct question was asked.
There is a regular question on big business power generally. In every survey I am aware of, going back to the 1960s, a majority have thought that big business has too much power, with an upward trend (also recorded in the 2005 article linked to above).
One hypothesis that could be worth exploring is whether there is a counter-cyclical component to opinion on these matters; that when Labor is in office people are inclined to assume that unions have too much power, and when the Coalition is in office people are inclined to assume that business has too much power.
LikeLike
Thanks Andrew.
LikeLike