Are punters smarter than pollsters in predicting election outcomes?

By James Lane and Louise Evans

The Statement

"The betting markets are often cited as a better indicator. They were spectacularly wrong-footed in this election too. The theory that punters are smarter than pollsters is now a dead letter."

Ipsos pollster Jess Elgood. May 28, 2019.

The Analysis

The fortunes of prime ministers and political parties have risen and crashed on the back of opinion polls. But the polls for the May 18 federal election were consistently wrong, pointing to a Labor victory. The pollsters are now under fire as they conduct post-mortem examinations on their data and methodology. [1]

AAP FactCheck examined the claim by Ipsos pollster Jess Elgood that while betting markets are often cited as a better indicator of an election outcome they too were “spectacularly wrong-footed” by the May 18 result and as a result the theory that punters are smarter than pollsters is now dead.

On May 16, 2019 two days before polling day, Sportsbet issued a media statement stating “punters have declared Saturday’s federal election run and won, backing Labor into Winx-like odds of $1.16 with 70 per cent of all money wagered on the election going on Bill Shorten’s team”. [2]

Sportsbet paid out more than $1.3 million to punters who backed a Labor victory before the May 18 poll result. One punter collected $128,000 before a single vote had been counted. Sportsbet’s communications manager Rich Hummerston said: “Our punters have spoken through their bets! 7 of every 10 bets on the election are on Labor. They’re supremely confident we will be paying out on Saturday so we have decided to pay them early. Punters rarely get it wrong on elections.” [2] [3] [4]

On election day Sportsbet.com.au said the betting trend over the previous 48 hours had favoured Bill Shorten and Labor to win and offered odds of $1.10. The coalition was at $7. Late on election night, Sportsbet admitted its error of paying out early with a combined dig at United Australia Party leader Clive Palmer in a tweet: “Yep, we blew $1.3 million. Could have been $80 million though eh Clive?" [3]

TAB issued a media release on May 17 with the headline ‘$100k bet for Labor as betting surge continues’. It had Labor priced at $1.15 for victory and the coalition at $5.50. TAB’s communication advisor Gerard Daffy admitted afterwards that the election result was another example of bookmakers misreading the political landscape. “This result joins the 2016 election of Donald Trump and the Brexit referendum in the same year as recent political events where the betting market has not predicted the winner," he said. [5]

Online gambling firm Betfair had Labor priced at $1.08 and the coalition at $11.50 just before polling closed in eastern Australia. [6]

AAP FactCheck found betting markets produced mixed results in predicting the outcome of the previous three federal elections in 2010, 2013 and 2016.

In 2016 the punters got it right betting on the coalition to be returned to government. Prime Minister Malcolm Turnbull won with a reduced majority of 76 seats while Labor claimed 69 seats. Two days before the July 2 poll, Sportsbet had Labor priced at $8 and the coalition at $1.08 while Centrebet had the coalition at $1.10 and Labor at $7. [7] [8]

In a betting preview for the 2016 poll on June 30, Ladbrokes said betting on the coalition had been “very steady at the $1.10 mark for a number of weeks now” adding the surprise Brexit result on June 23 had cast some doubt the coalition “are as much of a sure thing as the betting suggests”. [9]

Punters also correctly called the outcome of the 2013 election. On September 6, a day before the 2013 election, Sportsbet declared the Rudd Labor government was unbackable and shifted the focus to the number of seats the coalition was likely to win. The coalition led by Tony Abbott was at odds of $1.85 to win 91 to 100 seats. The coalition won 90 seats to Labor’s 55. [10]

In 2010 the punters installed the Gillard Labor government as favourite but there was a late surge for opposition leader Tony Abbott. BetEasy reported an “unprecedented election eve plunge” on Mr Abbott. William Hill Australia shortened its odds for a coalition victory over the Gillard Labor government from $3.60 to $2.60 in the space of eight hours, however the bookmakers maintained Labor as the favourite at $1.48. Similarly, Sportingbet and Centrebet also reported late betting surges for Mr Abbott and the coalition. [11] [12]

The 2010 result saw Labor and the coalition claim 72 seats each. Protracted negotiations with independents and Greens MPs resulted in a minority Labor government for Prime Minister Julia Gillard. [13]

Based on the evidence, Ms Elgood’s claim that betting markets were “spectacularly wrong-footed” for the 2019 election is true. However, there is an error in her claim that “punters are smarter than pollsters is now a dead letter” because betting correctly reflected the results for 2016 and 2013 elections and punters had an each-way bet in 2010 which produced an each-way result.

The Verdict

  • Somewhat True - Mostly accurate, but there is more than one error or problem.

The References

1.‘Nation's most influential pollster can't explain election disaster’, by Peter Hartcher. Sydney Morning Herald. May 28, 2019: https://www.smh.com.au/federal-election-2019/nation-s-most-influential-pollster-can-t-explain-election-disaster-20190527-p51rhc.html

2. We’ve paid out early on Labor to win this election’, by Patrick Garschagen. Sportsbet News. May 16, 2019: https://news.sportsbet.com.au/novelty/sportsbet-pays-out-early-on-labor-to-win-election/

3. ‘Betting agency makes monumental and costly error ahead of election’, by James McKern. News.com.au. May 19, 2019: https://www.news.com.au/sport/sports-life/betting-agency-makes-monumental-and-costly-error-ahead-of-election/news-story/a9a49b6aef6d2b03e9a8d0841a376d81

4. ‘Punter puts $850K on Labor to win federal election’. Yahoo News Australia. May 18, 2019: https://au.news.yahoo.com/australian-federal-election-2019-punter-places-850000-bet-on-labor-to-win-election-053423987.html

5. ‘Sportsbet expected to lose at least $5.2 million after election bungle’. Nine News. May 20, 2019:https://www.9news.com.au/national/federal-election-2019-sportsbet-expected-to-lose-at-least-52-million-after-election-bungle/358fad0d-ed93-47c3-bddf-6aa175e6ca64

6. ‘Coalition wins election but Abbott loses Warringah, plus how the polls got it so wrong’, by Adrian Beaumont. The Conversation. May 19, 2019: https://theconversation.com/coalition-wins-election-but-abbott-loses-warringah-plus-how-the-polls-got-it-so-wrong-116804

7.’The 2016 federal election’, by Dr Damon Muller. Parliament of Australia: https://www.aph.gov.au/About_Parliament/Parliamentary_Departments/Parliamentary_Library/pubs/BriefingBook45p/FederalElection2016

8. ‘Election 2016: Labor's odds blow out - but why?’, by Michael Koziol and Nicole Hasham. The Sydney Morning Herald. June 30, 2016: https://www.smh.com.au/politics/federal/election-2016-labors-odds-blow-out--but-why-20160630-gpv45x.html

9. ‘2016 Australian Federal Election Betting Preview’, by Thomas Hackett. Ladbrokes. June 30, 2016: https://www.ladbrokes.com.au/blog/2016/06/30/2016-australian-federal-election-betting-preview/

10. ‘Election betting points to a landslide’, by Paul Sheehan. The Sydney Morning Herald. September 6, 2013: https://www.smh.com.au/opinion/election-betting-odds-point-to-landslide-20130906-2t9zl.html

11. ‘Huge betting plunge on coalition’. BetEasy. August 20, 2010: https://beteasy.com.au/blog/politics/huge-betting-plunge-on-coalition/

12. ‘Betting odds shortened in final moments of campaign’. AM. ABC. August 21, 2010: https://www.abc.net.au/am/content/2010/s2989501.htm

13. ‘2010 Federal Election: a brief history’. Parliament of Australia. March 6, 2012: https://www.aph.gov.au/About_Parliament/Parliamentary_Departments/Parliamentary_Library/pubs/rp/rp1112/12rp08

  • First published May 29, 2019 17:10 AEST