The Michigan Democratic and Republican primaries took place Tuesday. The Democratic contest is interesting not because any candidate had a chance of touching President Joe Biden, but for what it might say about the reach of the Israel-Hamas war into Democratic Party politics.
The Republican primary drew interest because it could shed more light on the lopsided race between former President Donald Trump and former South Carolina Gov. Nikki Haley. Well, at this point it's not really a race, but Michigan will show how long the cycle of Trump-trouncing-Haley and Haley-vows-to-continue will go on. It will also shed more light, after elections in Iowa, New Hampshire, Nevada and South Carolina, on the number of voters, Republican and Democrat, who will vote against Trump in a GOP primary.
There has been very little polling on the race. In fact, there is just one recent poll, done by Emerson College and The Hill, which was taken between Feb. 20 and Feb. 24, the day of the South Carolina primary. The poll showed Trump with a massive lead over Haley in Michigan, 69% to 20% -- a 49-percentage-point margin. An additional 11% said they were undecided. When those people were asked which candidate they lean toward, the Trump lead got even bigger, 76% to 24% -- a 52-point margin.
The numbers seem crazy, like a lot of polls in the GOP race. But it's worth noting that Emerson and The Hill did one of the final polls in the South Carolina primary. It showed Trump defeating Haley by 22 points. The final margin was 20 points, which meant the poll was pretty accurate.
Indeed, as the New York Times' Nate Cohn remarked recently, primary polling has been "more accurate than average." But at the same time, polls in Iowa, New Hampshire and South Carolina have overstated Trump's support. Looking at the RealClearPolitics average of polls, in Iowa, the final pre-election average had Trump over Haley by 34 points. He actually beat her by 32 points. In New Hampshire, the final pre-election average had Trump over Haley by 19 points. He actually won by 11 points. In South Carolina, the average showed Trump up by 23 points. He won by 20 points.
Not a bad performance from the pollsters. But still, each average overstated Trump's lead. Cohn looked into various explanations. The most persuasive is that pollsters have been unable to fully and accurately identify the electorate in the Republican contests so far. This is the problem: When a pollster seeks to measure the support of "likely GOP primary voters," he looks for voters who have either participated in a Republican primary in the last eight years or have registered recently as a Republican. That's fine for finding Republican voters, but what about those Democrats who chose to vote in the GOP primaries and caucuses? Screening poll respondents for Republican activity would not find them. So when they showed up at the polls to vote for Haley, they were a surprise to the pollsters.
"The decision to survey prior Republican primary voters is understandable," writes Cohn. "It makes the poll much cheaper and homes in on the respondents likeliest to vote, but it will obviously miss any previous Democratic voters who hadn't voted in a Republican primary and now choose to do so."
Cohn suggests that has been an important theme in polling the Trump-Haley race so far. The number of Democrats who chose to vote in a Republican primary in order to support Haley and oppose Trump has been larger than previously estimated. That is one reason we've seen post-election data showing that many Haley voters characterized their vote not as a show of support for her but as a show of opposition to Trump.
Obviously, such voting has not been decisive anywhere, but it has been bigger than expected. It could happen again in Michigan, and then all around the country on Super Tuesday on March 5. For those voters, Nikki Haley is just a vessel -- their current means of opposing Trump. When the general election comes around, they'll be back, voting for whatever candidate is not named Donald Trump.
Join the conversation as a VIP Member