The biggest news from yesterday’s elections in the United Kingdom was the achievement of a parliamentary majority by the Conservative Party. The second biggest news was that the entire polling industry predicted a different result.
Nate Silver of FiveThirtyEight had a succinct summary of why this seems to be happening lately, not just in the UK but in the Scottish Independence Referendum, the Israeli elections, the 2014 U.S. midterms and even the 2012 U.S. presidential election:
Voters are becoming harder to contact, especially on landline telephones. Online polls have become commonplace, but some eschew probability sampling, historically the bedrock of polling methodology. And in the U.S., some pollsters have been caught withholding results when they differ from other surveys, “herding” toward a false consensus about a race instead of behaving independently.
But I added this comment at Washington Monthly:
All these are big and legitimate concerns. But probably the bigger problem is that such issues will be seized upon by anti-data zealots and “game-change” journalists–think of them as like the old-fart baseball scouts in Moneyball who knew a good player when they saw one–to seek to discredit any objective measurements of public opinion or any analysis based upon them. After all, polls are “wrong,” right? So let’s just wing it with our instincts, prejudices, snail’s-eye observations from the campaign trail (or bar), insider opinions, and of course, first-person anecdotal takes on the mood of the electorate.
The solution to flawed data is better data, not less data and certainly not data-free reporting and analysis. Keep that in mind next time someone tells you to “ignore the polls.”