How could professional pollsters go wrong so substantially when it came to calling New Hampshire?

On Primary day, the four most up-to-the-minute state-wide polls showed Obama ahead of Hillary by an average of 8.5%. The same polls showed a very tight Republican race with McCain winning by an average of 3.2%.

As it turned out, McCain beat the expectations of the pollsters by 2%, on average; one of the respected polls on primary day (Suffolk/.WHDH) actually showed a Romney victory by 4%, so the 71-year-old Arizona Senator (“The Ancient of Change”) beat that one by 9%. Hillary bested the expectations of the four most recent polls by an average of  11% -- an embarrassing blown-call for the so-called professionals.

So, what went wrong?

I’d like to suggest a public opinion equivalent of the “Heisenberg Uncertainty Principle” – the idea in physics that suggests that the very act of observing certain phenomena can alter those phenomena. 

Okay, it’s stretching it a bit to compare Hillary and McCain to sub-atomic particles, but still, I believe that the polling results announced all day Tuesday and most of the day Monday helped to make those polls inaccurate.

The message in all media was very clear about the state of the two races: we were told repeatedly the GOP contest was very, very close (though it ultimately produced a surprisingly comfortable win for McCain) while the Democratic battle would amount to a one-sided blow-out for Obama (though Hillary won, very narrowly).

Here’s my theory about what happened:

I think pollsters and experts were right that most of the independents in New Hampshire (45% of all voters) liked two candidates: Obama and McCain. In the forty-eight hours before the polls closed, they got a consistent message about their two favorites: Obama had his victory in the bag, but McCain was potentially in trouble. Therefore, sophisticated independent voters (who could choose to participate in either the Republican or Democratic contest) reasoned that McCain needed their help but Obama didn’t. Therefore, those who wanted, above all, to make a difference, switched at the last moment to the GOP side, abandoning their previous intention to vote Democratic. That’s why the split of independent voters between those who went with the GOP and those who went with the Democrats wasn’t nearly as one-sidedly Democratic as expected.

This shift of some independents who had originally intended to vote for Obama to the Republican primary to vote for McCain, explains the fact that McCain did better than predicted, and Obama did vastly worse.

According to exit polls, McCain slaughtered Romney among the self-described independents (who were an unexpectedly high 37% of total GOP voters) by a margin of 40% to 27% (Ron Paul got 13%). Among self-described Republicans, the Mittster came much closer to the Mac (33% to 35%) but McCain still won.

By the way, the Romniacs can’t try to blame “anti-Mormon prejudice” among Evangelicals for their guy’s sorry showing this time: Mitt actually ran closer to McCain among self-described “white Evangelical Christians” (27% for Mac, 27% for  Mitt, 28% for Huck) than he did among the non-Evangelicals (38% for Mac, 34% for Mitt). Evangelicals represented only 23% of the New Hampshire Republican voters (as opposed to 60% of GOP caucus goers in Iowa) but Mitt scarcely improved his disappointing performance, once again investing millions while falling well behind (nine points and five points, respectively) an underfunded rival. .

As he did last week, Romney displayed special weakness among three key groups in the electorate among which McCain-- just like Huckabee in Iowa-- beat the Mitt-man soundly. Romney performed disproportionately poorly among women (they just don’t like him, well-tailored good looks and all), voters below 30, and voters who earn below $50,000 a year.

Unless team Romney somehow manages to attract female, younger and working-class voters, he’ll continue to lose – just as surely as he’d get creamed in the general election.

 Meanwhile, he now claims an unwanted distinction – as only the second prominent Massachusetts politician in the history of the New Hampshire Primary to fail to win his contest in the Granite State. He joins Ted Kennedy in 1980 (who lost to incumbent president Jimmy Carter) as a singular Bay Stater who crashed and burned in New Hamphire—managing to avoid the near-automatic victories achieved by the likes of John Kennedy, Henry Cabot Lodge, Michael Dukakis, Paul Tsongas, and even John Kerry.

If Romney bases his whole campaign on business competence and management expertise, how does he justify investing an estimated $120 million so far (thirty times more than Huckabee, by the way, including at least $17 million of his own money) and earning in return only two feeble defeats in Iowa and New Hampshire?

That question may be impertinent, but it’s certainly appropriate.

And as to those who want to discount McCain’s victory because it was based on the support of independents, isn’t it blindingly obvious that no Republican can even come close to winning the White House without drawing similar support from independents in November? It’s difficult to discount the proposition that the guy who attracts independents in January will likely stand the best chance of getting their backing again some ten months later.

Good luck to all our candidates – and it’s on to Michigan, and new clarity, next week!