TV watchers early on election night – Tuesday, Nov. 8 – should pay the most attention to the results from North Carolina, Georgia and Florida, according to Prof. Geoff Peterson, chair of the UW-Eau Claire political science department.
Peterson told his lunchtime audience Wednesday that if Donald Trump loses all three of those states to Hillary Clinton, suspense about the result is as good as over. If there is a split of those states, the next key to the outcome will be in the Central time zone, particularly the results in Iowa and Michigan, he said.
If Trump doesn’t emerge from the Central time zone states with a sizeable lead, he’s almost certain to lose the election, Peterson said. That’s because he is running far behind in California, Oregon, Washington and Hawaii, with a total of 78 electoral votes among them.
That means that Clinton must pick up only 192 electoral votes from other states to reach the 270 electoral votes needed for a majority. Splitting the electoral votes of states in the Eastern and Central time zones will give her very close to that number, he said.
Peterson spoke, and fielded questions, at the UW-EC Alumni Association’s “Let’s Do Lunch” series at the Eau Claire Golf & Country Club. He said there are still “a lot of states up for grabs” (including Texas, to his surprise) but, based on polling data and past election patterns, he expects Clinton to win the popular vote by 6 to 8%.
Clinton’s Electoral Vote Edge
However, he said she is likely to pile up a much larger electoral vote edge, winning perhaps as many as 350 such votes. That’s because Trump is likely to win a sizeable number of the states he carries by “blowout” margins (60 to 70% of the popular vote) while he expects Clinton’s victories in most states to be much closer. He cautioned that although the national polling data seems to be giving Clinton an increasing edge, presidential elections are really 51 separate state elections and each one is different.
Among the other points that Peterson made in the course of an hour were:
- The Wisconsin senatorial race is likely to be much closer than many national polling organizations are predicting, with Republican Sen. Ron Johnson and his challenger, former Sen. Russ Feingold, very likely in a statistical dead heat at the moment.
- The most accurate national polls this year, according to analyst Nate Silver (www.FiveThirtyEight.com), are the ones conducted by or for ABC, Reuters and the New Orleans Times-Picayune. Silver, formerly oriented toward statistical analyses of baseball, has developed a reputation for accurate predications since shifting his focus to elections.
- State polls are better predictors than national surveys of how the Electoral College tally will come out.
- Because “the American electoral system is remarkably decentralized,” it would be almost impossible to “rig” a presidential election, regardless of Trump’s claims to the contrary, because it would have to be done at the county level and there are some 3,500 counties in the country.
In regard to the Wisconsin senatorial race, Peterson acknowledged that several national polls had Feingold ahead by 6 to 8%. He noted, however, that the Marquette University poll – for which he has the greatest respect – now puts Feingold’s margin at only 2.5% and that is within the poll’s margin of error.
The only potentially close Wisconsin contest for the House of Representatives is in the Eighth District, which includes Green Bay and Appleton, Peterson said. There, Rep. Reid Ribble (R-Appleton) is stepping down after serving for three terms and Republican Mike Gallagher and Democrat Tom Nelson are vying for the open seat.
Peterson said that Clinton and Trump “are the two most unpopular candidates in the history of polling.” Nevertheless, he said he expects voter turnout to be in the 56-57% range and that there will not be a large number of people who will ignore the presidential vote while still voting for offices lower on the ballot.
“I suspect that, in the end, most people will vote at the top of the ballot,” he said.
Peterson acknowledged that there are problems with the Electoral College model and noted that other democracies have ignored this option in setting up their own election systems.
“Nobody on the planet looked at it like, ‘we should do that,’” he said. “We’re stuck with it. It’s not going away.”
Polling Difficulties
He also discussed some of the difficulties and shortcoming of election polling. These frequently come down to erroneous “assumptions based on the partisan makeup of the population” and the difficulty of weighting samples to reflect the actual makeups of both the electorate as a whole and of the “likely voters,” he said.
Peterson said that the increasing trend toward using only a cell phone and giving up landlines entirely was once thought to be the death knell for polling. But polling organizations have found ways to obtain cell phone numbers and are increasingly able to include cell-only voters in their samples, he said.
That’s important, he noted, because in 2014, 22% of adults and 64% of people under 25 no longer had land lines and by 2016, the percentage for the under-25 group had climbed to about 80%. In addition, research done in the 2008 and 2012 elections indicated that people who use only cell phones tend to be younger, more urban and more likely to be Democrats than those who retain landlines. Undercounting the cell-only voters in the sampling process could skew the results, he said.
Peterson said this year’s campaign is so unusual that anything could happen in the last two weeks before the election, “from space aliens. . . to (Russian President Vladimir) Putin jumping in as a third party candidate.”