Britain’s faulty polling

I have been present for every British general election since 1987. This time, I could not understand how the incumbent Conservative coalition could have such positive policy results – low unemployment and high economic confidence – yet remain tied or even behind their Labour opponents in the months leading up to the vote. On election night, I finally got it: The polls were wrong.

And that is the first of three important lessons Thursday’s British election can teach us in America. Even as the science of polling has achieved great precision, political preferences on both sides of the Atlantic are more difficult to measure and the public more fickle than ever. A warning for Americans in 2016: What people say even 24 hours before the vote can be quite different from what they do at the ballot box.

In my recent travels up and down the British countryside, I was shocked at how many people didn’t know who the party leaders were or what was in their platforms yet still claimed they cared about the outcome. Voters told me they couldn’t be bothered to follow the daily machinations of the candidates – “they really aren’t saying much of anything” – but they asserted they were “absolutely” going to vote. I didn’t believe them. I was wrong. Turnout was up.

No doubt the reason for this – and this is lesson two – is voters increasingly are so turned off by the political process that they pay attention only on the eve of the vote. We saw it in the United States in the 2014 midterms, thanks to a barrage of negative ads. I expect to see it again next year.

Lesson three, and this part I did get right, is the absolute necessity of “trust” in voter intentions. In just about every pre-election poll, Conservatives and Labour were locked in a dead heat, yet voters gave Conservative leader David Cameron a large advantage over Ed Miliband, his Labour challenger, on the question of whom they would trust more to run the country.

In the end, trust is a far more accurate predictor of the election outcome than any other metric.

This is a huge lesson, in particular, for Hillary Rodham Clinton, and she should bear it in mind as she prepares for her upcoming appearance before Congress on Benghazi and her email usage. While it is very difficult to appear presidential under oath at a witness table, any hint of dishonesty or misstep that undermines her credibility could fatally damage her already embattled campaign.

Luntz is a Republican pollster.