Explaining What Happened in New Hampshire

I've been looking at the Campaigns & Elections web site, where an article by Shane D'Aprile seeks to explain what went wrong in the pre-primary polls in New Hampshire.

As you're bound to have heard by now, Hillary Clinton defeated Barack Obama in New Hampshire. Polls released the day before were unanimously predicting Obama would win. Some polls anticipated margins in double digits.

Pollster Scott Rasmussen told D'Aprile that he has some "theories" about what happened, "but I don’t think there’s just one explanation.”

* One of the factors was a "late trend" in Clinton's favor. Presumably, this occurred after Clinton got misty on camera.

* Another factor, Rasmussen says, was "there were a lot of people who showed up, motivated by the historic nature of [Clinton's] campaign. They were people who wouldn’t normally vote in a primary, and probably didn’t make it through our screen.”

It is true that turnout was historically high in New Hampshire Tuesday. More than half a milliion people cast ballots in the primaries -- a ridiculously high figure in that state.

* D'Aprile also points out that participation by women went up 3-4%, which was more of an increase than Rasmussen expected.

"Pollster John Zogby also points to this high turnout among women, particularly older women," D'Aprile writes, "and the large number of voters that didn’t settle on a candidate until primary day ..."

* "Another potential explanation," writes D'Aprile, "is that more independent voters chose to cast their ballot in the Republican primary than anticipated."

* In addition to the independent voters, the role of race has been mentioned as "a factor in overstating Obama’s support -- the theory being that respondents are reluctant to tell pollsters they don’t support the African-American candidate." D'Aprile admits, "It’s something that likely holds less sway in a primary election, though."

Irregular voters -- as would be the case with a higher-than-expected turnout of women -- and choices made in open primaries -- as with higher numbers of independents voting in the Republican primary -- often do not get through pollster's filters or receive representation in the findings. Thus, their eventual impact may not be fully appreciated -- until after it happens.

But the public expects pollsters to anticipate these things before they happen.

Which leads to the question -- Are pollsters really fortune tellers?

My feeling is that pollsters may have a little more insight into these things than ordinary people. But I think they tend to find out about the future the same way the rest of us do -- when it happens.

And conventional wisdom still has its role to play in elections, for good or ill.

For some reason, I've been thinking about the Civil War lately, and I was recalling that, in the election campaign of 1864, the general feeling was that Abraham Lincoln would not be re-elected, the Emancipation Proclamation would be repealed, and the North and the South would seek a truce. There were no public opinion polls in those days, of course, but that was the way many people felt things would play out.

But they were wrong. The North began to turn the tide militarily, Lincoln was re-elected, and the future was far different from what many anticipated 144 years ago.

These days, when the unexpected occurs, the tendency is to look for someone or something to blame.

In this case, D'Aprile says, everything was right -- except for the winner on the Democratic side. "In pre-primary polls, [John Edwards'] support hovered between 17 and 20 percent -- he finished with 17 percent ..."

And on the Republican side, D'Aprile says, "pollsters had the race pegged just right with Sen. John McCain holding a four-point lead on average before the primary –- he bested [Mitt] Romney by six points. "

I might point out, in the interest of full disclosure, that I have referred to the findings from both the Rasmussen and Zogby surveys in this blog on many occasions.

And I see nothing wrong with reading the polls to gain some insight into what people are thinking.

But it's important to remember that polls and elections are two different things.

Just ask President Dewey!