[Ed. – No real surprise here, but cautionary information nonetheless. Think about this: when do polls ever seem to overstate or overrepresent the “expected” impact of conservative voting? Never, that’s when. Wrongly weighted polls always seem to favor the left. And that’s what happened in Britain last year.]
So now it’s official. We know what the pollsters did, and we know how they did it. The University of Southampton’s report into the great 2015 general election polling debacle stops short of telling us why they did it, but it doesn’t matter. We can fill that bit in for ourselves.
First, it’s been confirmed that the pollsters “herded”. Of course we knew the polls had mysteriously aligned in the final hours of the campaign – we could all see that for ourselves. But we know now – as some of us argued at the time – that this was not the basis of some ordinary statistical anomaly. As Professor Patrick Sturgiss, head of the inquiry, confirms: “A surprising feature of the 2015 election was the lack of variability across the polls in estimates of the difference in the Labour and Conservative vote shares…”
“The primary cause of the failure of the 2015 pre-election opinion polls was unrepresentativeness in the composition of the poll samples. The methods of sample recruitment used by the polling organisations resulted in systematic over-representation of Labour voters and under-representation of Conservative voters”. In other words, the pollsters inserted too many Labour voters into their polls, and not enough Conservatives.