AMSRO Inquiry Finds 2019 Election Polling 'Failure'

Australian research association AMSRO has published the final report of its inquiry into opinion polls during the 2019 Federal Election, finding evidence of a 'polling failure' rather than simply a 'polling miss'.

Last year, some Australian politicians and commentators called for opinion polls to be scrapped, following Prime Minister Scott Morrison's conservative Coalition win. Of the sixteen polls held since the election was called, all were in favour of the Labor party, when respondents were asked to pick from the main two parties. The final result was: Coalition 41% and Labor 34%.

AMSRO and the Statistical Society of Australia then launched the inquiry, which included assessing the accuracy of the polls, and evaluating whether there was any previous pattern of inaccuracies. The inquiry found that from 2007 until 2016, Australian pollsters had, in aggregate, a 96% success rate (that is, 25 out of 26 final week polls called the right result) across four Federal elections. AMSRO described the 2019 May election as a 'notable anomaly', and it then invited pollsters, media organisations and others who commission election and political polling to contribute to the investigation.

The report, which offers ten recommendations for how to improve political polling in Australia, found that the source of errors lay in the polls themselves, rather than resulting from a last-minute shift in preferences among voters. The inquiry also found some evidence that the reporting of the polls failed to consistently meet the basic disclosure guidelines for editors and journalists, as set out by the Australian Press Council.

Inquiry Chair Darren Pennay, founder and former CEO of The Social Research Centre, comments: 'The first preference votes were either underestimated (LNP) or overestimated (ALP) because of inadequately adjusted, unrepresentative samples. It is likely the polls were skewed towards over-representing more politically engaged and better educated voters, and this bias was not corrected. As a result, the polls over-represented Labor voters. Pollsters should seek to better understand the biases in their samples and to develop more effective sample balancing and/or weighting strategies to improve representativeness, by looking at education or other variables'.

The full report can be found at: www.amsro.com.au/amsro-polling-inquiry .

All articles 2006-23 written and edited by Mel Crowther and/or Nick Thomas, 2024- by Nick Thomas, unless otherwise stated.