Article content continued
This is not a new phenomenon
So this is not a new phenomenon. It has, in fact, become pretty much a ritual, regularly leaving news editors scratching their heads and wondering how the pollsters managed to bugger it up so badly, and why they’d been so naive as to trust in their predictions.
Donald Trump’s surprise victory in 2016 was supposed to be such a shocking event that the polling industry delved deeply into its practices and procedures to ensure nothing so embarrassing could happen again. And yet, four years later, there was Trump within inches of an unexpected victory over Joe Biden, after we’d all been told the Democratic candidate was well ahead in plenty of crucial swing states, and should have a relatively smooth run, perhaps even a blowout.
So, OK, how did they get it wrong? Well, let’s see, all sorts of reasons. Turns out Cubans in Miami still haven’t gotten over Castro, and Hispanics aren’t one big homogenous population that all votes for the same party. Who knew? And according to Pew Research, a big U.S. pollster, response rates to survey inquiries have fallen to six per cent, meaning 94 per cent of people can’t be bothered. Add in those who lie, aren’t planning to vote at all, or don’t want to admit they’re supporting someone like Trump, and you’ve got an industry based on the word of two or three people out of a hundred, with nothing better to do.
Even industry veterans wonder what’s wrong. “You cannot be this wrong, this consistently wrong, and actually pretend like you know what you’re doing,” Democratic pollster Chris Kofinis told the CBC.
