IE 11 is not supported. For an optimal experience visit our site on another browser.

How one Republican nearly blew a lead bigger than Trump’s — and what it means for 2024

Bob Dole’s 1996 near-collapse in Iowa after racking up a huge polling lead offers some hope for Trump challengers, but the underlying numbers show key differences.
Former President Donald Trump, then-Sen. Bob Dole, R-Kan.
Former President Donald Trump; then-Sen. Bob Dole, R-Kan.Getty Images file

For Republicans hoping to avoid a third consecutive Donald Trump nomination, the main finding of the new NBC News/Des Moines Register/Mediacom poll was not encouraging, with the former president swamping Florida Gov. Ron DeSantis, his nearest rival, 42% to 19% in Iowa — a 23-point margin. 

But the history of the Register’s polling in Iowa also seemed to offer these same Republicans a glimmer of hope. As the graphic below shows, Trump’s sizable lead is not the largest ever enjoyed by a Republican at this early point in a caucus campaign cycle. And, crucially, we have seen one previous GOP candidate — Bob Dole, in the 1996 election cycle — lose more than 40 points of support between roughly this point and the actual caucuses. 

The polling leaders at this point ahead of past Iowa GOP caucuses.
The polling leaders at this point ahead of past Iowa GOP caucuses.

The implication for 2024 is clear: If someone has bled that much support before, surely it can happen again. And it wouldn’t take nearly as dramatic a plunge for Trump to lose his lead, which is only half of what Dole’s was at the start of the ’96 campaign.

But a closer look at the numbers reveals some key differences between then and now, and those differences suggest Trump’s 2024 edge may be more durable — and his challengers’ task much tougher.

First, some quick context. When Dole began his campaign, he was the Senate majority leader and the clear national front-runner, having finished second to George H.W. Bush in the previous open GOP primary race, in 1988. Dole had romped to victory in Iowa by double digits in the ’88 campaign, leading him to later dub himself the “president of Iowa.”

So when Dole entered the ’96 race — his third stab at the White House — it was considered a foregone conclusion that he’d win Iowa again.

That was the backdrop for the release of the Register’s first Iowa poll, in the late spring of ’95. Dole was in first with 57%, followed distantly by Sen. Phil Gramm of Texas at 11%. None of the eight other candidates or potential candidates tested by the Register were in double digits.

Fast-forward to the caucuses in February 1996, and that gigantic Dole lead had all but evaporated. He ended up winning, but only barely, with a mere 26% of the vote. Television commentator Pat Buchanan finished second with 23%, and former Tennessee Gov. Lamar Alexander was right behind him with 18%. For Dole, it was a jarringly close call, one that would portend a shocking defeat to Buchanan in New Hampshire eight days later — although ultimately Dole would go on to win the nomination.

So, yes, Dole did squander nearly all of that 46-point lead that he’d started out with. And for all we know, Trump could end up surrendering most or all of the 23-point Iowa lead he’s now sitting on.

But a look one level deeper than the horse-race numbers uncovers some key differences between the two situations.

The first is that Dole’s initial support was clearly soft. In that 1995 Register poll, only 28% of Dole’s supporters said their minds were made up to vote for him; 67% said they were open to backing someone else.

The numbers for Trump in the 2023 poll are exactly reversed: 67% of his current supporters say their minds are made up, and they are not open to the idea of supporting other candidates.

Plus, 72% of respondents overall in that 1995 poll said their minds weren’t yet made up, compared with 52% in 2023. That means the atmosphere was more volatile and the potential for Dole to lose significant ground was apparent from the start. To compare the polls is to recognize that Trump may have already locked down more support than Dole ended up receiving in the actual caucuses.

The media-political universe of 1995 was also vastly different than today, with cable news still a relative novelty — there was no MSNBC or Fox News yet — and the internet in its infancy. In other words, it was harder for candidates to gain exposure to wide audiences.

This is another reason Dole, with two previous Iowa campaigns under his belt, opened with such an enormous polling lead. The 1995 Register poll found that 98% of Iowa Republicans knew enough about him to have an opinion — and that 86% viewed him favorably. Only two other candidates had at least 70% name recognition among Iowa Republicans: Buchanan (83%) and Gramm (70%). Half of the 10 names included in the Register poll were unknown to more than 50% of Republicans. 

Our new poll, by contrast, shows multiple candidates with near-universal familiarity to Iowa Republicans: 98% have an opinion of Trump, 97% for DeSantis, and 95% for former Vice President Mike Pence. More than half of the likely caucusgoers in our poll expressed favorable or unfavorable views of all eight candidates who were onstage for this week’s debate.

In the media-political universe of 2023, candidates tend to be more familiar to voters much earlier than they used to be. So while simple name recognition clearly accounted for a big chunk of Dole’s ’95 polling lead, Trump is dominating against a field that Iowa Republicans know a lot more about.

Ultimately, the case of Dole’s near-disastrous plunge may not actually offer that much hope to Trump’s foes. The situations are just too different.

Then again, there still may be a silver lining here for them. It has to do with the much higher name recognition that candidates now tend to enjoy.

DeSantis, as noted, is already almost universally known to Iowa Republicans. And, at least so far, they like the Florida governor, with 66% expressing a favorable view of him in the new poll, a point higher than Trump’s number.

And DeSantis isn’t alone. Sen. Tim Scott of South Carolina is viewed favorably by 59% of Iowa Republicans and unfavorably by just 17% — just about half of Trump’s 33% unfavorable score. It’s not translating into massive support right now, but there is broad goodwill among Iowa Republicans for at least two of Trump’s challengers, and it’s possible more will join the club after Wednesday’s debate.

And where there’s broad goodwill, there’s always hope for a campaign that they may yet convert it into support — the kind of support that could give Trump a run for his money. 

All of which is a long way of saying: It’s not what’s similar between 1995 and 2023 that should give non-Trump Republicans hope. It’s what’s different.