By Jay Jamison
The polls got it wrong once again. This time the unexpected results surprised the pollsters and pundits down under, in Australia. The left-leaning Australia Labor Party, headed by Bill Shorten, was assured victory by everyone in the know, which somehow didn’t include a majority of the electorate. One thing I discovered about Australian elections is that voting is compulsory. All citizens eligible to vote, must vote — or pay a fine. So, the results cannot be blamed on poor turnout, which is often the excuse of losing candidates in America. Since everyone is essentially required to vote, the polling data is not just based on the responses of “likely voters,” which is the usual practice here.
So how to account for another polling fail, to be placed alongside that of the Brexit vote in Britain and the American presidential election of 2016? All kinds of excuses can be made for the British and American vote, but if everyone is essentially required to vote, and the pollsters got it wrong anyway, one distinct possibility could be that a large number of Australians lied to the pollsters. I consider this possibility to be wonderful news, if true. Many who have been reading my stuff through the years know that I am a critic of reporting polls as news. In January of 2011, I wrote, “For the future of the Republic, for the good of the country, please lie to the pollsters.” I doubt if a single Australian voter has ever heard of me, but I’m happy to see that enough of them have arrived at roughly the same conclusion about polling as I have.
Last February, Pat Caddell, the iconic pollster, and in my opinion one of the last voices of sanity in national politics, died in South Carolina. In an interview before he died, he noted that there are at least two distinct types of polls: internal tracking polls used by campaigns to get an accurate picture of the views and attitudes of the electorate; and “push polls,” surveys that are designed to influence the electorate, rather than inform them. Guess which ones are reported on the evening news? Even if the polls are accurate, which recent experience shows is doubtful, they do not give us any good reasons to vote for one candidate over another — they just hint at which direction the herd is going.
In recent elections, a new phenomenon has emerged, which may help explain why the polls have been so fantastically wrong. In the Brexit vote held in Britain in the summer of 2016, there were no candidates to smear. Voters were to either vote to remain in the European Union (EU), or to leave it. Many people who wanted Britain to remain in the EU began to characterize those who might favor leaving in very harsh terms. Brexit voters (those who wanted to leave the EU) were stamped as racists, stupid fools, xenophobic nationalists and other epithets. Demeaning a large block of the electorate, and then somehow expecting them to vote your way, is a bad strategy. I suspect the same thing happened here in the presidential election of 2016 (try Googling “deplorables”). If enough likely voters either lied to the pollsters or simply refused to be surveyed — for fear of being seen as racists or worse — the poll is unlikely to give an accurate picture of the electorate. Many people so abused are likely to keep their views to themselves, and only express them in the privacy of the voting booth.