2022-11-24, 15:00–16:20 (Australia/Melbourne), Zoom Conference Room
In the leadup to the 2022 Australian election, there was considerable speculation about the possible success of minor parties and independents, and the likelihood of a hung parliament.
Just two weeks before election day, YouGov released estimates for all 151 House of Representatives electorates, the first time any organisation had done this during an Australian national election campaign. It indicated hung parliament speculation was wrong, predicting Labor would win 76-85 seats, with the most likely outcome being 80 (they won 77), and that the Coalition would win 58-68 (they won 58).
These estimates came from a model-based approach called Multilevel Regression and Poststratification (MRP), which combined a large survey with information from the Census and other data, allowing for a more granular approach to polling.
YouGov’s Director of Data Science — Public Affairs and Polling, Dr Shaun Ratcliff, will talk about how the model worked, what it said about the election, and lessons from YouGov’s innovations at this election.
Dr Shaun Ratcliff
Director, Data Science - Public Affairs and Polling, YouGov
Shaun is responsible for complex, often high profile projects for clients engaged in political activities and related work across the Asia-Pacific region, and other specialty custom research.
His work is largely focused on understanding how the public thinks about issues, organisations and events, and what influences their beliefs and actions. He helps national political campaigns, NGOs, trade unions and corporate clients collect the data they need to make well-informed decisions.
Shaun also regularly engages in public discussions on voter attitudes and behaviours. His research has appeared in a range of academic publications, and he has authored articles on and been interviewed about these topics for international and Australian media outlets.
He has a PhD in Political Science from Monash University, and he has taught data science and survey research for ACSPRI, the University of Sydney, Monash University and the University of Melbourne.