Internal polling conducted by Democratic U.S. Senate candidate Andrew Romanoff's campaign shows the former state House speaker closing the gap with former Colorado Gov. John Hickenlooper to within 12 percentage points with a dozen days to go before ballots are due in the June 30 primary.
Romanoff trails Hickenlooper 51% to 39%, according to a telephone survey conducted June 16 and 17 among likely primary voters, pollster Andrew Myers said in a polling memo obtained by Colorado Politics.
That's down from a 49-percentage point spread between the two Democrats in an internal October poll conducted by another firm for the Romanoff campaign. That survey showed Hickenlooper, who entered the race in late August, ahead 68% to 19% among likely primary voters, the campaign said.
"It’s pretty profound what we’re seeing," Myers told Colorado Politics. "As I look at this race, for all intents and purposes, I think Hickenlooper is like the incumbent here. He’s dropped to 51% — that’s the danger zone."
Hickenlooper and Romanoff are facing off to determine which Democrat challenges Republican U.S. Sen. Cory Gardner in the fall.
In his polling memo, Myers said the survey results make clear why national Democrats are pouring money into the primary to support Hickenlooper, who has been beset with weeks of bad headlines over ethics violations and gaffes.
"Hickenlooper’s 'bad month' (as the media has labeled it), coupled with Romanoff’s introduction of himself to this electorate, has created a seismic shift in the fundamentals of this contest," Myers said.
The Democrats' Senate Majority PAC spent just over $1 million this week on TV advertising to counter Republican ads attacking Hickenlooper after the state's Independent Ethics Commission determined the former governor violated a constitutional ban on gifts when he accepted travel on a private plane and in a limousine. Hickenlooper was also found in contempt by the commission for defying a subpoena.
Hickenlooper has also taken heat in recent weeks for verbal stumbles — "I've stepped on my own words," he said during the pair's final debate on Tuesday — including comparing the plight of busy politicians to galley slaves in a 2014 video clip resurfaced this week by opponents.
Scant polling conducted last fall showed Hickenlooper leading Democratic rivals by overwhelming margins, but no publicly released polls have been available since the crowded primary field narrowed to Hickenlooper and Romanoff in early May.
The Hickenlooper campaign hasn't released its internal polling, but a source close to the campaign told Colorado Politics that its numbers don't reflect a race as close as Myers' poll found.
"The bottom line here is that Romanoff has seized the momentum, and he is well-positioned to pull off a remarkable upset in this contest," said Myers, who has been polling Colorado voters for 20 years.
"As this campaign has moved forward and voters have learned more about Hickenlooper, as well as the choice they have with Romanoff, Hickenlooper has lost very serious ground and fallen into the danger zone for a politician as well-known as he is."
Myers said the contest narrows further to a 10-point race among the poll's respondents who say they're most likely to vote.
After voters hear a paragraph describing campaign ads currently airing, he said, Romanoff takes the lead by 6 percentage points, with 48% to Hickenlooper's 42%.
"Again, a terrible sign indeed for Hickenlooper as he enters the final stretch of this race," Myers said.
Myers Research & Strategic Services polled 500 likely voters in Colorado's Democratic primary, with live operators contacting voters using land lines and cell phones. Democrats accounted for 60% of the sample and unaffiliated voters made up 30%. The margin of error is 4.4 percentage points.
According to a breakdown of the polling sample's composition provided to Colorado Politics, 49% of respondents described themselves as liberal, 38% as moderate, and 8% as conservative, with 5% who wouldn't say. Those under 50 made up 37% of the sample, and 61% were 50 and older. By race, 65% were white, 19% were Latinx, 6% were African-American, 1% were Asian or Pacific Islander, 1% were Native American, 2% were something else, and 6% declined to say.
This developing story has been updated.