One of the key factors to consider when looking at the history of Australia’s constitutional referendums is which government submitted the redistribution proposal. Labor and their Coalition opponents (and their predecessors) have held differing attitudes to the constitution, all the way back to the original formation of the Constitution. Labor was barely represented at that Convention, and the party was very critical of key parts of the federation settlement from the very beginning.
Over the first century of federation, Labor proposed more referendums than the Coalition (despite being in power for much less time), but had much less success. Only one out of 25 Labor proposals were successful (4%) while seven out of 19 proposals from non-Labor governments were successful (37%).
While these facts are true, the story is a bit more complex. The story is partly about how parties operate in opposition and whether they are willing to support their opponents’ proposals, and it also ignores quite a lot of close-run races. A slightly higher Yes vote in a number of Labor-backed referendums would have changed the picture dramatically.
Another factor we have to consider is the trend over time. My blog post last Thursday focused on a variety of trends over time. There were a lot more close contests in the early 20th century – later in the century, we saw more referendums defeated by wide margins.
For this first chart, I have shown the average national Yes percentage for Labor and non-Labor proposals over four time periods.
Through the first three time periods, Labor and non-Labor generally proposed about the same number of referendums. But since 1984, the Coalition has only proposed two: the Republic and Preamble referendums of 1999, which don’t fit easily into previous categories.
Over time, support levels for referendums have generally declined for Labor. For the Coalition, there was a spike in the 1970s (when three referendums proposed by Coalition governments were successful) which bucked the trend.
The first impulse when looking at this chart is to say that Labor proposals have done worse. That is technically true, but there’s a lot more to the story. The big difference is whether the proposal is also supported by the opposition.
The Coalition has generally been far less willing to endorse proposals from Labor governments than vice versa.
This may reflect the types of questions that Labor governments put up – certainly they have in the past put up a number of proposals to increase Commonwealth powers that Coalition governments have not supported. But there are also clear examples of the same issue coming up on a number of occasions, with the Coalition supporting the proposal in government but not in opposition.
The issue of aligning the terms of the House and Senate was raised in slightly different proposals in 1974, 1977, 1984 and 1988. The Coalition opposition opposed the Whitlam government’s proposal in 1974, and then proposed something very similar in 1977, indeed with the same name (Simultaneous Elections). The 1977 proposal was bipartisan, with Labor opposition support, and managed 62% of the vote nationally. This was the biggest Yes vote for an unsuccessful referendum, only managing to win in three states.
The Hawke government then returned to the topic in 1984 and also rolled it into a broader proposal to align both Houses on a four-year term in 1988, and both were also defeated in the face of Coalition opposition. The 1984 referendum managed a narrow national majority but only won in two states, while the 1988 referendum lost badly.
There have also been cases where a Labor opposition came out against proposals similar to those they had proposed earlier. The Labor government had made proposals in 1911 and 1913 to expand Commonwealth powers, but Labor went on to oppose a similar referendum in 1919 under a Nationalist government. The one common factor was Billy Hughes, who had been the Labor attorney-general in 1911 and 1913 and was the Nationalist prime minister in 1919, so this could explain Labor’s change in attitude.
I have taken a chart I used last week to compare the national Yes vote and the Yes vote in the pivotal fourth-best state for each referendum, but tweaked it. Each dot is coloured according to the government at the time of the proposal, and proposals with bipartisan support are marked with a star.
(Thanks to Peter Brent’s 2021 article on this topic where he listed which referendums had bipartisan support. It’s surprisingly difficult to determine this convincingly. I differ from Peter by listing the 1919 referendums as not bipartisan (he had them as unclear) and also listed the Republic as not bipartisan (you can’t really have a bipartisan referendum when the Prime Minister is opposed). I remain unclear on 1926 and 1937, but have grouped them in with non-bipartisan referendums on the chart.)
The results are stark – every successful referendum, along with the two most popular unsuccessful referendums, were all bipartisan. The only clearcut bipartisan referendum not to at least manage 50% of the national vote was the 1967 referendum on breaking the nexus between the number of seats in the House of Representatives and the Senate. That referendum barely managed 40% of the vote, and one state.
Once you take out the eleven bipartisan referendums, the gap between support for Labor and Coalition-sponsored referendums almost entirely disappeared. Of the nine non-Labor referendums without bipartisan support, the average Yes vote was 44.9%. Of the 24 Labor referendums without bipartisan support, the average Yes vote was 44.3%.
One other point that should be considered when assessing Labor’s historical record in terms of referendum results: there was a lot of close calls.
If you look at the chart above, you can see a tight cluster of red dots just outside of the victory zone.
Of the five referendums that won a national Yes majority but failed to achieve the double majority, three of them were Labor-initiated: Industrial Employment and Marketing in 1946, and Terms of Senators in 1984. If about 7,000 No votes in South Australia switched to Yes, that would have flipped both of those 1946 referendums. If under 28,000 No votes in Western Australia (3.5%) and a few hundred No votes in South Australia switched to Yes, the 1984 referendum would have been won.
The other referendum worth mentioning was the 1913 referendum. After failing to pass two referendums in 1911, the Labor government split up the questions into six separate referendums that was put to the 1913 election. Labor lost the election, but came very close in all six referendums. The national Yes vote was over 49% in all six referendums, but didn’t reach a majority in any of them. All six referendums won three states, and all narrowly fell short of the requisite fourth state in Victoria. A swing of about 1% nationally and in Victoria would win at least five of those six referendums, and a sixth wasn’t much further away.
If a few votes had gone a different way, Labor could have won another nine referendums. While I’m at it, I should acknowledge that the Coalition likewise came close to winning the 1951 referendum on the Communist Party.
Ultimately the statistic that we all remember is the wins and losses – eight out of 44. But that can really flatten the history and you can miss a lot of the details.
