The argument about Donald Trump’s chances of winning the Republican presidential nomination has gone something like this:
Us: Trump is very unlikely to win the nomination.
Them: He’s likely to win. Just look at the polls.
Us: Those polls don’t mean much. They haven’t been predictive in the past.
Them: But so many people are paying attention to the campaign this year, the polls may be more meaningful.
That last point, for example, was made recently by political forecaster Alan Abramowitz.
The hypothesis is possible, but there’s no evidence to support it. In fact, historically there has been no relationship between how predictive the polls were at this point and how many people said they were paying attention to the campaign in the fall before primary voting.
It’s certainly true that voters say they are paying more attention to this campaign than to previous ones. Two weeks ago, 82 percent of Republicans and Republican-leaning independents said they were following the election “very closely” or “somewhat closely,” according to an ABC News/Washington Post survey.1 These pollsters asked the same question in eight previous primary races with no incumbent president running for a given party’s nomination, and no more than 75 percent of voters in that party ever said they were following the campaign “very closely” or “somewhat closely.”
|CYCLE||PARTY||FALL POLL LEADER||PARTY VOTERS FOLLOWING CLOSELY||WON NOMINATION|
|1988||Republican||George H.W. Bush||51||✔|
|2000||Republican||George W. Bush||50||✔|
But a lot of voters were paying attention during the three most recent primaries without an incumbent president of their party running, and the polls were not all that predictive. The fall polls before primary voting called the winner correctly in 2012, but Mitt Romney barely led the field in late 2011, and he eventually won the nomination fairly comfortably. In fall 2007, Rudy Giuliani was leading in the Republican primary, and John McCain, the eventual 2008 nominee, was in fourth place.
And remember the epic 2008 Democratic primary? It held the previous record for early attention from voters (75 percent among self-identified Democrats and independents who leaned Democratic). Yet in the same poll, Hillary Clinton was crushing Barack Obama by 27 percentage points. Obama, of course, went on to win the nomination.
On the opposite end of the spectrum, the poll leader won the nomination three out of five times when 55 percent or less of a party’s voters said they were paying attention in the fall before the primary.
Although the 2000 election later drew the eyes of the nation, Democratic and Republican voters were not tuning in to the campaign early in the primary campaigns. Just 50 percent of Republicans and Republican-leaning independents and 44 percent of Democrats and Democratic-leaning independents were following the presidential race at least somewhat closely. And even though voters weren’t paying close attention, George W. Bush and Al Gore were both polling above 60 percent at this point, and each went on to win more than 60 percent of his party’s support when all the votes were counted.
Why aren’t the fall primary polls more predictive when more voters are tuned in? I’m not sure, but I have a few theories:
- Highly competitive elections may be drawing more attention, and highly competitive elections may be likelier to change. The 2000 primaries were kind of a snooze fest; Bush and Gore were in dominant positions basically from start to finish. I don’t think anyone would call the 2016 GOP primary uncompetitive.
- The more voters pay attention, the more likely they are to be following campaign events that can shift the race. They are likelier to hear about gaffes, such as Ben Carson defending his claim that the Egyptian pyramids were used to store grain. Statements like those may be contributing to Carson’s decline in the polls.
- National polls are fairly meaningless. Obama defeated Clinton only after beating her in Iowa. It’s possible that there would be a much stronger correlation between the vote and attention paid to the campaign if all states voted on the same day. They don’t. This year, Iowa has consistently been one of Trump’s worst states. He has averaged just 20 percent of the vote in live-interview Iowa caucus polls. It wouldn’t be surprising if he lost Iowa, lots of voters heard about that loss, and his national numbers dropped.
Trump could win the nomination. In this week’s 2016 Slack chat, I laid out a number of reasons why Trump’s current standing could be more meaningful than we think it is. But the fact that voters say they are paying close attention to the campaign is not one of them.