The 1967 Referendum: What a 90.77 Per Cent Vote Can and Cannot Prove
In May 1967, Australian voters approved two constitutional amendments by a national majority of about 90.77 per cent, with majorities in all six states. The amendments allowed the Commonwealth to legislate for Aboriginal people and required Aboriginal people to be counted in national population statistics. The campaign had been conducted for years by organisations including the Federal Council for the Advancement of Aborigines and Torres Strait Islanders, and it received support from all major political parties and the press. Some historians argue the result shows overwhelming public support for Aboriginal rights; others argue it shows that the question was framed so narrowly, and with so little organised opposition, that it measured something less than that.
1. Explain why a 90.77 per cent result is weak evidence for the claim that the public supported Aboriginal rights in 1967, and what would be stronger evidence.
Read the worked answer
A referendum measures approval of a specific constitutional change, not approval of a broad programme. The two amendments were technical: they removed references that prevented the Commonwealth counting Aboriginal people and legislating for them. A voter could approve both while opposing land rights, equal wages, or the repeal of protection laws, because none of those was on the ballot. The question was framed narrowly, so the answer is narrow.
The absence of organised opposition also weakens the inference. When all major parties and the press support a measure, a large majority may express the absence of a campaign against it rather than the strength of conviction. Support under no pressure is not the same as support under contest. The result also tells a historian nothing about intensity: a voter who gave the matter no thought and a voter who had campaigned for years both appear as one yes.
Stronger evidence would be contemporaneous survey data asking about specific policies, records of public meetings where land rights or wage equality were debated, the size and distribution of donations to Aboriginal advancement organisations, and the behaviour of governments after 1967 when specific legislation was proposed. If the same electorate that voted yes then resisted specific measures, the referendum result cannot be read as endorsement of those measures. The 90.77 per cent figure is best treated as evidence of what was not politically contested in 1967, which is itself historically interesting.
2. Majorities in all six states are sometimes cited as proof of national consensus. Assess that use of the state results.
Read the worked answer
The six-state majority is genuine evidence of breadth, because it shows the result was not produced by one populous state outweighing the rest. That matters in a federal system where a national majority alone can conceal regional division, and it distinguishes 1967 from referendums that passed on the votes of New South Wales and Victoria. So the state results do support a claim about geographical spread.
They do not support a claim about consensus, for three reasons. First, a state majority can be narrow or overwhelming, and the setup gives only the national figure, so a historian cannot tell whether some states passed by slender margins. Second, breadth is not depth: a measure can be approved in every state because it is uncontroversial, which is exactly the alternative explanation the setup raises. Third, the absence of a No campaign means the state results record the distribution of a largely unopposed proposition, not the resolution of a contest.
The phrase 'national consensus' also implies that Aboriginal people were part of the agreement, but the setup gives no figures for Aboriginal voters and no evidence about how Aboriginal organisations understood the result. Many campaigners saw the vote as a means to further claims, not as an end. A careful historian would say the state results establish geographical breadth of approval for a narrow amendment, and would avoid the word consensus, which claims more than the data can carry.
3. The campaign was conducted for years by Aboriginal-led organisations and supported by all major parties. Explain how these two facts complicate the idea that 1967 was a victory won by either Aboriginal activism or establishment politics alone.
Read the worked answer
The two facts describe different contributions that are easily collapsed into a single credit claim. The years of campaigning by organisations such as FCAATSI produced the pressure, the educational work, and the public demand that made the issue salient. Without that, there is no reason to think a government would have put the question. But the organisations could not put it themselves: a constitutional referendum requires parliamentary action, and the support of all major parties is what converted an activist demand into an official proposal with the resources of the state behind it.
The complication is that each contribution was necessary and neither was sufficient. Aboriginal-led campaigning without party support would have remained a demand; party support without campaigning would have had no public constituency pressing for it. This means a narrative that credits only activism understates the institutional mechanism, and a narrative that credits only the political parties erases the years of Aboriginal organising that set the terms.
The framing also conceals a tension. The parties supported a narrow amendment, and the campaigners had broader aims. The victory therefore belonged to a coalition whose members wanted different things from it, which is why the years after 1967 saw conflict over what the mandate meant. A historian should describe 1967 as a convergence of distinct projects, not as the triumph of one.