Statement from Ipsos

Regarding political polls following the Federal Election

The Federal election was very close in terms of overall party support but Ipsos acknowledges that its final Ipsos Poll numbers clearly did not reflect the eventual Federal election results last weekend and it needs to do better.

Ipsos will be conducting an assessment of its polling over coming weeks with the local team and the global Ipsos polling experts in order to understand what the research company could have done to achieve a more accurate outcome. No doubt other polling companies who covered the election will be doing the same.

Ipsos’ initial review of the results has shown that its pre-election polling (Labor 50.7% vs Liberal/National Coalition 49.3%) was the most accurate of all polls when it came to the two-party preferred numbers, and from a statistical perspective the actual election result (Labor 49.1% vs Liberal/National 50.9% - based on approximately 76% of the vote counted to date) was within the 2.3% error margin of these results. Clearly, however, the final poll Ipsos conducted did not show the Coalition being ahead of Labor.

Ipsos' primary vote for Labor was (32.9% vs 33.9% so far) and the Coalition (39.0% vs 41.4% so far). Ipsos also notes that the overall trends in its numbers did match the results. Ipsos’ most recent poll, published on Friday 17th May, showed some late momentum toward the Coalition. The two-party preferred (Labor vs Coalition) vote tightened from 52.3% vs 47.7% in the Ipsos Poll published on 6th May to 50.7% vs 49.3%. This was a direct result of an observed spike in support for the coalition increasing from 36.3% on May 6th to 39% on May 17th.

Ipsos upholds the very highest standards of polling using interviewer conducted telephone surveying with sampling based on randomly generated mobile phone and landline numbers, coupled with demographic quotas and weighting to Australian Bureau of Statistics population numbers. Ipsos has maintained its methodology and approach in all aspects of its polling since before the 2016 Federal election where its numbers did reflect the election outcomes. While Ipsos strongly believes its approach is robust, the company recognises there is work to be done to understand why its results differed from the final election outcomes and how to best address this in future polling.

A number of broad trends present challenges to polling in Australia, and to polling organisations globally, which Ipsos is actively reviewing. Areas which will be reviewed are currently being identified and to date include:

  1. In Australia the voting trend away from the two major parties appears to have increased again in this election with both parties showing a reduction in vote share since 2016 in the vicinity of 0.5%-1.0%. This makes predicting the preference flows from an increasingly fragmented base of minor parties and independent candidates difficult.
  2. Politics is inherently local, challenging a national poll in the heat of the last throes of an election campaign that is fought on local issues.
  3. Globally pollsters are challenged by more highly educated people being more politically engaged, and therefore more likely to complete a polling survey. With randomly generated mobile and landline phone numbers this is less likely to be an issue of coverage, but more one of likelihood of different types of people to respond and is not easy to address through weighting.
  4. 7% of Ipsos’ final poll respondents said they “didn’t know” who they would support at the polling booth. The reporting of primary and two-party preferred voting intention effectively assumes these people will vote in the same way as those who have provided an intention. It is unclear whether this is a reasonable assumption
  5. Any poll is only as accurate as the stated intention people give on what they will do when they vote. Stated intention may change between the date polled and when a voting ballot is completed. For most of the people responding to the Ipsos pre-election poll who had not postal voted or pre-voted already this gap was between 3 and 6 days.

Finally, Ipsos notes that political polling has a range of goals, most beyond the measurement of the overall two-party preferred vote. Ipsos’ polls look to understand key issues, leadership qualities and personalities, the primary voting intentions of voters and many other issues. Predicting outcomes is always difficult when the race is tight, and Ipsos remains committed to finessing its approach moving forward.