The election of Donald Trump in 2016 came as a hammer-blow not just for Hillary Clinton and the Democratic party but also for the opinion polling industry. While they had correctly predicted her majority in the popular vote, they had catastrophically missed the bigger story: Trump had won in the key swing states to clinch the presidency via the electoral college.
The Guardian’s US data editor, Mona Chalabi, tells Anushka Asthana that following 2016, the pollsters tweaked their models as they ate humble pie, but this year they are back to projecting the results with similar confidence: and once again they are predicting a defeat for Donald Trump. But have those trends they missed in 2016 been properly accounted for? And should parts of the media be far more sceptical than they currently are about the scientific-sounding claims of the major polling sites?