Voice referendum betting odds: what the 2023 prices showed

What was the Voice referendum?

The referendum asked Australians whether to alter the Constitution to recognise Aboriginal and Torres Strait Islander peoples by establishing an Aboriginal and Torres Strait Islander Voice. The proposed body is also often called the Indigenous Voice to Parliament.

Under the proposed constitutional text, the Voice could make representations to the Commonwealth Parliament and Executive Government on matters relating to Aboriginal and Torres Strait Islander peoples. Parliament would have power to make laws about the Voice, including its composition, functions, powers and procedures. The proposed wording did not itself require that its members be elected.

A constitutional amendment needed a double majority to pass:

• A national majority: more than half of formal votes across Australia, including the territories, had to be Yes.
• A state majority: more than half of formal votes had to be Yes in at least four of the six states.

The official result

The Australian Electoral Commission’s final national count was 39.94% Yes and 60.06% No. No state recorded a Yes majority, so the proposal failed both parts of the double-majority test.

The ACT recorded a Yes majority. Its votes contributed to the national total, but the ACT did not count as a state for the separate state-majority requirement. The AEC returned the writ on 6 November 2023, completing the formal process.

That official count establishes the outcome of the vote. It does not, by itself, establish the exact terms on which a bookmaker settled an individual wager.

Voice referendum betting odds: the documented snapshots

Contemporary reporting provides several snapshots of BlueBet’s Yes and No betting odds. Decimal prices below identify what was reported at particular points, not a complete record of the market.

The September article was stamped 11:09 am, and the October page carried an update time, but neither establishes the exact time or time zone at which its quoted prices were captured. An article’s publication or update time should not automatically be treated as the time of the bookmaker’s quote.

• June 2023 opening, reported by The Guardian on 4 September: BlueBet opened with Yes as the favourite at $1.70. The report did not give the corresponding No price or the precise June opening date.
• 19 September 2023, reported by BettingSite.com.au: BlueBet had Yes at $4.75 and No at $1.15. The precise capture time and time zone were not stated.
• 13 October 2023, reported by Bets.com.au: BlueBet’s market, labelled “Will Australia Vote Yes To The Voice Referendum?”, listed Yes at $5.75 and No at $1.09. The source said the prices were correct that date, but did not establish their capture time or whether they were closing prices.

How the reported prices changed

The Guardian reported that Yes was BlueBet’s opening favourite in June. By the September and October snapshots, No was the shorter-priced selection. A shorter decimal price indicates the favourite in that particular pair of prices.

Between the two later reports, No shortened from $1.15 to $1.09, while Yes lengthened from $4.75 to $5.75. That comparison does not show every movement between the snapshots, and the sources do not establish precisely why the prices changed.

Nor do the BlueBet figures establish what every operator offered. On 4 September, The Guardian reported that Sportsbet, Betfair, Neds, Ladbrokes and Betr had said they would not offer a Voice referendum market. That is evidence about those named operators at that time, not proof that BlueBet was the only provider throughout the campaign.

The October report did not reproduce the market-specific settlement wording. In the actual vote, No won nationally and in every state. The available report does not show how the market would have treated a hypothetical national Yes majority without the required state majorities.

How to read the decimal odds

Decimal odds show the total return per dollar staked on a winning bet, including the original stake, subject to the applicable terms. A lower price means a smaller potential return for the same stake and a higher raw implied probability.

To calculate that raw implied probability, divide one by the decimal odds and multiply by 100. Using the reported 13 October BlueBet prices gives:

• No at $1.09: approximately 91.74% raw implied probability.
• Yes at $5.75: approximately 17.39% raw implied probability.
• Combined: approximately 109.13%, or 9.13 percentage points above 100%.

What the implied percentages do and do not mean

The amount above 100% is the overround in the two offered prices. It means the raw percentages are not independent, margin-free forecasts. The overround also does not guarantee the bookmaker a profit on every possible outcome.

The 91.74% figure relates to a priced No outcome. It does not mean that 91.74% of voters were expected to vote No. A favourite can also lose: a short price is not a guarantee.

For a simple return example, a hypothetical $10 winning fixed-odds bet at $1.09 would return $10.90 in total, including the $10 stake, before applicable terms. That calculation is an illustration, not evidence that a particular customer was paid.

Odds, polls and the final result measure different things

The No price in BettingSite.com.au’s 19 September report was $1.15. Dividing one by that price gives a raw implied probability of approximately 86.96%, before allowing for the bookmaker’s margin. It was a price on the outcome, not a forecast that nearly 87% of formal votes would be No.

A Roy Morgan online survey conducted from 18 to 24 September asked 1,511 Australian electors about their voting intention. It reported 44% No, 39% Yes and 17% undecided. When undecided respondents were excluded, the reported split was 53% No and 47% Yes.

The survey fieldwork overlapped the date of the odds report, but it ran over a week; the price was a snapshot whose exact capture time is not verified. Both pointed towards No in their respective ways. Their percentages are not interchangeable, however: the poll measured respondents’ stated intentions, while the price concerned the referendum outcome.

The final AEC result of 60.06% No was different again: it was the No share of formal votes counted. It was not the realisation of the price’s implied percentage. The dated favourite and the final outcome can be compared, but the evidence does not show that the price caused the result.

• Price on an outcome: the reported $1.15 No price implied approximately 86.96% before accounting for margin.
• Sampled intention: 44% of Roy Morgan respondents said No, while 17% were undecided.
• Formal vote share: the AEC counted 60.06% No nationally.

Can Australians still bet on the 2023 outcome?

Readers should not expect to place a new bet on the Voice referendum of 2023: the vote is complete and the AEC result is final. The research did not identify a verifiable current market on that completed outcome, although that is not a claim that every operator worldwide was checked.

For other online betting, the Australian Communications and Media Authority has a register for checking whether an interactive gambling provider is licensed in Australia. A provider’s licence alone does not establish that a particular market is permitted or available, and rules and customer eligibility can depend on jurisdiction.

Gambling Help Online offers free, confidential support on 1800 858 858, 24 hours a day, seven days a week.

What is known about settlement?

The official result was No, but the Voice-specific settlement terms, actual settlement notices and payout records were not located. The evidence therefore does not establish a settlement date or prove how every individual bet was treated.

BlueBet’s archived sport rules dated 4 April 2023 included general rules for political markets. They said markets would generally be resulted according to the initial officially declared election result, and allowed settlement to be withheld, including where challenges or recounts were possible.

Those general rules are useful context, but they do not prove that the same version remained unchanged on voting day or that it contained every term applicable to the Voice market. A straightforward No-outcome bet would logically correspond to the official result if the market terms treated No as the winning selection and the bet stood. That is a conditional explanation, not a verified payout.

For an individual historical wager, the relevant evidence would be the market terms accepted when the bet was placed, any operator settlement notice and the customer’s own account record.

FAQ

What were the final Voice referendum betting odds?

The closing odds have not been verified. The latest documented snapshot identified in the research was reported for 13 October 2023, when BlueBet had No at $1.09 and Yes at $5.75. The exact time those prices were captured is unknown.

Did the betting favourite win?

Yes. In the named 13 October BlueBet snapshot, No was the shorter-priced selection, and No was the official outcome. That does not establish who was favourite at every bookmaker or at every point in the campaign; The Guardian reported that Yes was BlueBet’s opening favourite in June.

Were Voice referendum odds the same as opinion polls?

No. Odds are prices on an outcome, while polls record voting intentions from a sample of people. The September No price of $1.15 implied approximately 86.96% before accounting for margin; Roy Morgan’s overlapping survey found 44% of respondents said they intended to vote No. Those percentages describe different things.

How were Voice referendum bets settled?

The official outcome was No, but the Voice-specific settlement wording and actual payout records were not found. BlueBet’s archived general political rules described an approach based on official results. A No bet would correspond to the outcome only if the relevant terms treated No as the winning selection and the bet stood.

Can I still bet on the 2023 Voice referendum?

The vote is over and the result is final, so readers should not expect to place a new bet on that outcome. The research did not identify a verifiable current market, but did not check every operator worldwide. The AEC’s final results are available for anyone researching the vote.

Would a national Yes majority alone have passed the referendum?

No. The proposal also needed a Yes majority in at least four of the six states. Territory votes counted towards the national total but did not count as a state majority. The ACT recorded a Yes majority, while no state did.