The big news of the weekend before the election was the release of the final Des Moines Register poll conducted by the West Des Moines polling firm Selzer & Co., widely regarded as the most reliable pollster in Iowa and maybe the most reliable state pollster in the nation. Or, as Nate Silver’s FiveThirtyEight headlined an article in 2016, “Ann Selzer is the Best Pollster in Politics.” Her polls are not always matched by the final results, but they have been often enough that a Selzer poll can make headlines, as in our Washington Examiner story that ran after the Register released its story on Selzer’s most recent poll early Saturday evening.
“Donald Trump takes over lead in Iowa as Joe Biden fades,” read the headline on the Register story. September’s Register poll showed a 47% to 47% tie between President Trump and Biden. The October poll showed Trump ahead, 48% to 41%. That looks a lot like Trump’s actual victory margin in Iowa over Hillary Clinton in 2016: 51% to 42 percent. The result also showed, in a state and a nation where straight-ticket voting has become the rule rather than the exception, Republican Sen. Joni Ernst leading her Democratic challenger Theresa Greenfield by 46% to 42%. That’s a change from the Register’s September Selzer poll, in which Greenfield lead by 45% to 42%. The Senate race numbers are in line with three of the last four public polls, which showed Greenfield leading in September and up to mid-October, but Ernst narrowly leading now.
The Register also reports the party preferences for Congress in each of the state’s four congressional districts. Despite small sample sizes (about 200 interviews each), they show Republicans 15 points ahead in the 1st Congressional District in northeast Iowa and 17% ahead in the fourth district in the west. Both districts consist of many rural and small-town counties and a few small cities in blue-collar historically Democratic counties. They show the 2nd District, which includes the state’s second-largest city Cedar Rapids and the university town of Iowa City (in the state’s most Democratic county these days), evenly split and the 3rd District, which includes Des Moines and its affluent suburbs in Polk and Dallas Counties, leaning Democratic. Given the small sample size, not too much should be taken from these results, but they’re in line with the trend toward Republicans in rural and small-town counties and small blue-collar cities and the trend toward Democrats in larger metropolitan areas with their relatively large number of college graduates. Those trends were apparent both in the 2016 presidential election and in the 2018 gubernatorial race.
Currently, Democrats have a 3-1 edge in the Iowa House delegation, which means that if the Electoral College were deadlocked 269-269 and the election went to the House, and if the current parties held their seats, Iowa would cast its vote for Joe Biden. But the Selzer poll’s numbers suggest that it’s at least possible that Iowa would end up with a 2-2 or even a 3-1 Republican delegation. Since Democrats currently hold only 23 of the House’s state delegations, to 26 for Republicans, such a change could have national reverberations.
But that’s getting ahead of the game. And it should be kept in mind that the Register poll could be wrong: That is, it could be one of the 1 in 20 polls in which the results differ from the opinions of the total population by more than the margin of error.
But if it’s right, it could be pretty significant. The big national story in 2016 was the move toward Donald Trump among white non-college voters and outside million-plus metro areas, and Iowa has a large majority of white non-college voters and is the largest state without a single county in a million-plus metro area. The big Iowa story in 2016 was that the state, more Democratic than the national average in the Reagan years, and a very closely divided state in 2000 and 2004, was moving solidly toward Trump and the Republicans. In other words, Iowa was a distant early warning of trends that in areas outside million-plus metro areas in states such as Florida, Pennsylvania, Ohio, Michigan and Wisconsin elected Trump president.
Is this Iowa poll an indication of a late trend toward Trump in such areas this year? He’s been running behind his 2016 showings among the white non-college subgroup in polls in such states for much of the year. But subgroups have a larger margin of error than that for an entire state like Iowa, and late changes in opinion are always possible — and arguably swung the 1948 and 1980 elections to Harry Truman and Ronald Reagan.
Republican consultant Patrick Ruffini, who has been bearish about Trump’s chances of late, seems open to this idea. “Selzer, Susquehanna poll of WI, and Muhlenberg poll of PA are evidence of some late Midwest/WWC tightening a la ’16. Basically any Michigan poll and the CNN polls are evidence against,” he tweeted Saturday night.
And Republican analyst John Ellis in his Substack column, chimed in on Sunday morning, “Forty-one percent. I have been covering American politics for a long time and I can’t remember a number that so dramatically altered the political community’s perception of a presidential campaign as that number did, last night, at 7:30 p.m. Eastern Time.”
“What made it doubly disconcerting,” he went on, “was the way the Des Moines Register (accurately) described the poll results: ‘Republican President Donald Trump has taken over the lead in Iowa as Democratic former Vice President Joe Biden has faded. . .’Faded! Could there be a more terrible word in the last week of a presidential campaign? . . . For Democrats, last night’s Iowa Poll was the worst possible news at the worst possible moment. It foretold close results in Wisconsin and Minnesota. It undermined the Biden campaign’s momentum and morale. And it fracked [sic] Democrats’self-confidence. What had seemed reasonably certain no longer seemed certain at all.”
Of course it could be, just as so many liberal tweeters are insisting, just an oulier. But maybe not.
