For months political prognosticators have predicted Donald Trump’s political demise — but is it possible that polls are actually underestimating the Republican frontrunner’s support?

The theory was first propounded by John Phillips, a California-based conservative columnist and radio host.

“This means that nontraditional news consumers and nontraditional voters made a point to tune in and see what Trump had to say, yet many of these Trump supporters won’t be considered in the polls,” Phillips recently wrote in a column, pointing to the fact a jaw dropping 24-million people turned into the first Republican debate on Fox News, most likely because of Trump’s presence.

“Take the PPP poll, for example,” he went on, speaking of a recent Public Policy Polling survey which showed Trump with a sevem percentage point lead over his nearest Republican rival in Iowa. “It reflects only the opinion of ‘likely’ Republican primary voters. They define ‘likely’ voters as those who are self-described ‘regular’ primary voters. Therefore, any new person brought into the process by Trump wouldn’t be counted, underestimating his actual support.”

Phillips suggested that the Trump phenomena might be similar to the elections of Arnold Schwarzenegger and Jesse Ventura, both non-conventional celebrity gubernatorial candidates in California and Minnesota, respectively.

“In the 1998 election for Minnesota governor, an Oct. 20 poll conducted by the Star Tribune had Democrat Hubert Humphrey III in the lead with 35 percent, Republican Norm Coleman just behind him with 34 percent and wrestler Jesse Ventura in third,” Phillips pointed out. “On Election Day, Ventura won with 37 percent of the vote.”

There are obviously differences between Trump, Ventura and Schwarzenegger. Ventura, for instance, got elected as an independent in 1998 and though Schwarzenegger was elected as a Republican in 2003, he ran in a recall election that boasted 135 candidates.

Still, could Phillips’ theory have legs — could Trump really have a more commanding lead than surveys currently show? Some pollsters and political analysts tell The Daily Caller it’s not unthinkable.

“I don’t know if polls are underestimating or overestimating Trump’s support. Either is possible,” Princeton professor Sam Wang, who runs the Princeton Election Consortium, told TheDC in an email. “It depends on whether his appeal to people who don’t usually vote will last until the primaries start in earnest.”

Wang then pointed to a CNN poll of the Republican primary field released Tuesday that suggested Trump’s appeal might be far more long lasting than many political pundits once believed.

“Previously, he was strong only as a first choice,” Wang explained. “Now his support is starting to show some depth. If that depth holds up, there is a chance his surge will last longer than analysts were expecting in July.”

In the new poll, Trump not only leads all Republican contenders with 24 percent of the vote, he is also the top “second choice” among respondents with 14 percent support.

Thomas Holbrook, a political science professor at the University of Wisconsin-Milwaukee who studies political polls, says while Trump’s poll numbers are tricky to analyze at this stage, there may be something to Phillips’ theory.

“To be honest, I’m not sure what to make of Donald Trump’s poll numbers right now, at least in terms of what they might portend for the actual caucuses and primaries,” he told TheDC in an email. “If you look at the polls on pollster.com, Trump does do somewhat better when the samples are based on all adults instead of likely voters.”

“However, the thing about appealing to people who don’t usually vote is that they don’t usually vote,” he cautioned. “So I’d be hesitant to say that he will outperform the polls when voting actually takes place. This far out, though, I just think it’s really hard to say much about the elections based on the polls.”

Republican pollster Glen Bolger said that if any of the polls are underestimating Trump’s support, it’s more likely to be the state polls, not the national polls.

“Well, there’s two ways to look at it,” Bolger, who is currently working for a super PAC that supports former Texas Gov. Rick Perry, told TheDC in an email. “First, your premise about Trump attracting some people who don’t usually vote may well be true. However, many of the national polls are very loose in their definition of primary voters – they take anyone who is an identified Republican or an Independent who leans Republican. Thus, the national polls already account for this.”

“The state-specific polls (Iowa, New Hampshire) do drill down to likely primary voters (caucus goers in IA), so those might miss some Trump voters,” he added.

Bolger, though, went on to caution that “we don’t know whether these Trumpites who aren’t usually primary voters will do at turning out in primaries/caucuses.”

“It’s one thing for a Ventura voter to show up in a November election, it’s another for a Trump voter to show up for a February or March election,” he added. “I’m not saying they won’t – I’m just saying we don’t know.”

Political guru Larry Sabato, the head of University of Virginia’s Center for Politics, is skeptical the polls are undercounting Trump’s support, though he does believe Trump’s best chance to win is to “mobilize nonvoters or those not previously connected to party nominations.”

“But so far, a large majority of the polls have been using very small samples of Republicans (and Republican-leaning independents) to draw conclusions about Trump and the GOP field,” he said in an email. “The high probability is that the polls are already including loads of people who are very unlikely to show up in low-turnout caucuses and primaries — especially caucuses. So caution should be the rule for now.”

Jay DeSart, a polling expert at Utah Valley University, was even more wary of the theory than Sabato.

“It’s difficult to say, but I’m inclined at this point to think that Trump has probably hit a ceiling of support,” he said in an email. “This is functionally different than the Schwarzenegger and Ventura phenomena because we are still in the nomination phase and you’ve got a very different dynamic there than you do in the general election phase.”

Pointing to the recent CNN poll that shows 58 percent of Republicans believe the party has a better chance of winning the White House with a nominee other than Trump, DeSart added, “while there may be some room for Trump to expand his support, it will still be difficult for him to consolidate much beyond what he has now.”