From the Tuesday Morning Jolt:
New Hampshire Votes… While Most of Us Look On In Envy.
Today New Hampshire holds its presidential primary. As in Iowa, voters in the Granite State will have a chance to effectively end two, three, maybe four presidential campaigns, before those of us in forty-eight other states have a chance to cast a ballot. Then on February 20, South Carolina votes, and in most cycles it effectively picks the eventual Republican nominee. (Last year, Newt Gingrich won the Palmetto State; Mitt Romney recovered in Florida and racked up more wins on Super Tuesday.)
Every four years there’s rumblings of discontent that these small states get to go first, and talk of changing the order. Every four years, the rumbling never amounts to anything. At least Iowa didn’t make its choice based on ethanol support this year.
Secretary of State William Gardner predicts that today 282,000 votes will be cast in the Republican primary, 268,000 votes in the Democratic primary. That’s a 62 percent turnout rate.
By the way, there’s a chance a certain number of registered Democrats may show up today and want to vote for Donald Trump or John Kasich or any one of the Republicans with crossover appeal. They can’t; the last day to change parties was October 30.
Over on Newsday, Andrew E. Smith and David W. Moore have a nice column dispelling some of the conventional wisdom about New Hampshire. For starters, expect at least one significant surprise compared to the polls:
- In 1980, a CBS poll showed Ronald Reagan beating George H.W. Bush by 45 points, though his actual margin of victory was 27 points.
– In 1984, the final Washington Post-ABC News poll had Walter Mondale tied with Gary Hart, and the final CNN poll had Mondale winning by six points. Hart won by nine.
– In 1988, Gallup had Bob Dole beating George H.W. Bush by eight points, and the Post-ABC poll had Dole up by three.Bush won by nine.
– In 1996, CNN-Time showed Dole winning by 15 points. Patrick Buchanan won by one point.
– In 2000, the average of all polls showed John McCain beating George W. Bush by eight points. McCain won by 18 points, more than twice what polls predicted. Though this was a larger average error than in 2008, it was not labeled a “fiasco,” nor did the AAPOR investigate the causes.
Most famously, all of the pollsters predicted Barack Obama would win the Democratic primary in 2008, and Hillary Clinton won.
Why are polls often wrong? It’s not usually because of methodological issues but because of timing. When pollsters conclude their interviews (some by Friday, others as late as Sunday), many voters have not made up their minds. Exit polls show that 30 to 45 percent of voters make their decisions in the final three days of the campaign, and 15 to 20 percent do so on Election Day itself.
Smith and Moore note, “among registered voters, less than 15 percent are actually ‘independent,’ and only about one-third to one-half of them vote in the primary.”