Anchorage-based pollster and political consultant Mark Hellenthal released a wide-ranging poll on Wednesday, which shows, among other things, that if the election were held today, SB 21, or the oil tax legislation that was passed by the Legislature, would likely be repealed by a slim margin. It also showed that while Dan Sullivan has the lead in the U.S. Senate primary, it might not be as wide a lead between Sullivan and Mead Treadwell — the two leading candidates — as Sullivan’s internal polling has suggested.
The poll was conducted June 17-28 and includes 392 registered Alaskan registered voters who were contacted both on landlines and via cell phones. The margin of error is 4.9 percent, Hellenthal said, though that number goes up when primary voters are carved out in the poll.
“All other Alaska research firms contract their fielding work to ‘outside’ firms,” Hellenthal wrote in the poll’s introduction. “Therefore, their results are only as accurate as the ‘Outside’ firm’s quality controls.”
The poll shows that Sullivan is up by only 4.3 percent over Treadwell among those who are likely to vote in the Republican primary. Sullivan had 34.8 percent, Treadwell 30.5 percent and Miller 21.4 percent, with 13 percent still undecided. Remember, however, that this is a primary-vote carve out, which reduces the number of respondents substantially and increases the margin of error likewise.
Those are the horserace numbers. Dig deeper into the crosstabs and there’s more disturbing news for Sullivan. Treadwell has a much lower negative number than does Sullivan. A whopping 30 percent pf all voters have a negative view of Sullivan, while Treadwell only has a 10 percent negative rating. Miller, not surprisingly, has a 47 percent negative rating. These are all voters, including Democrats.
Hellenthal said that the attacks on Sullivan by Begich and the Democrats have hurt Sullivan. “They’ve bruised him,” he said. And in the meantime, everyone has left Treadwell alone.
“Someone’s going to have to go negative on Treadwell,” Hellenthal said. “If no one goes negative, Sullivan loses to Treadwell,” he predicted.
The other surprise in Hellenthal’s poll was that despite the millions spent on those who are fighting repealing the oil tax bill passed in 2013, repeal is winning by 2 percentage points by all voters, and oddly enough, by 5 percentage points by those who are planning to vote in the Republican primary.
Hellenthal’s theory is that all of the ads put out by the “No on 1” campaign have actually neutralized the message, particularly because in all the ads he’s heard, the “paid for” disclosure is at the end, and most of the ads so far have been paid for by the oil companies.
“What the public is left with is that BP, ConocoPhillips, and Exxon are saying that it’s a good thing for you if we give them more money,” he said.
Contact Amanda Coyne at amandamcoyne@yahoo.com
Mark-
Or we could propose a constitutional amendment to weaken our silly initiative process. I think it was a bad joke that the delegates never envisioned being used so often.
The cruise ship tax and the 90 day session are both examples of why Alaskans shouldn’t have such ease to vote on things.
If you’ve known him so long, then one would guess you’d know how to spell his name correctly.
Marc Hellenthal and a broken clock.
What are two things that are right twice a day?
Alex, I’ll take “other useless crap” for 400.
I heard Dan Sullivan speak and he did not lambast Begich. people in the audience did but her kept it quite high road. He spoke of what he could do with his experience. No matter how much progressive propaganda Amanda and her people spread the truth is you should go listen to Dan Sullivan yourself and see and hear what he does and says…. and if Tony Knowles is voting NO for all the right reasons as in to keep Alaska financially solvent shouldn’t we all vote NO to keep Alaska Strong! the answer is YES
Thanks for the report, Amanda. This is interesting stuff. I would like to say that it is a shameful world in which we live where a candidate has to “go negative” on another candidate in order to win. If a candidate has enough positive attributes, he/she should be able to sell themselves. I am a consummate cheerleader and believe that you can support your team without booing the other; this goes for sports or politics!
Also, people act like they personally own the oil that is underground in Alaska. Quit being so selfish and self-serving! Alaskans are rewarded for living in an oil-rich state and not one of us is responsible for the existence of these valuable resources. If we do not meet the producers part way and allow them a few tiny tax allowances, they will never invest in us and then we will have ZERO benefit. If you read the agreement, they are still paying huge sums to our state. Everyone wins: the producers make a profit; Alaskans reap the benefits; our country is self-sufficient. Look at the greater good and let’s do what we can to ensure drilling. Please, vote No on 1.
If one follows “Truth Teller’s” logic, one would dismiss all polls. The point is the confidence interval for a 600 sample is only 4% tighter than for a 267 sample. At a 50% observation, one can be 95% sure that the true population figure is with + 4% or 46% to 54% for a sample of 600 and for a 267 sample + 6% or 44% to 56%. You double the cost for very little gain!
Marc Hellenthal
Hellenthal & Associates
(907) 276-1001 (voice)
(907) 277-2315 (fax)
(907) 570-1252 (cell)
hellenthal@gci.net
Mr. Hellenthal’s admission that Sullivan could be ahead of Treadwell by 14 points or behind Treadwell by 6 points is useless, absurd and worthless information. Defending this as useful or relevant information iis making you look silly. I think you are staistically correct; however, my wild guess on where the race is would probably be more specific in accuracy that your work product. Why did you choose to share something so problematic and open to criiticism?
To suggest that Hellenthal is a dummy is dumb. I have know Mark for more than 20 years and find him to be very politically astute. His methodology was probably influenced by economics and costs. He probably had a client who wanted to do too much in one poll with a limited budget. The problem, which Amanda Coyne points out, is that with a small sample size, as he carved out for primary election questions, just creates a statistical ± that makes it difficult to narrowly predict where voters are because of the huge deviation. Accordingly, Sullivan could be anywhere from 4.9 points ahead to 10 points ahead. Again, Hellenthal’s numbers are not wrong, we just simply don’t know exactly what they are because of the huge error factor. Nonetheless, I like the report and information presented. The information is valuable.
Sorry, Lynn my comments were meant for Lee Raymond.
Mr. Helenthal,
I don’t think your response was intended for me?
Lynn Willis we polled all voters because we also had an interest in General election candidates and issues.
Marc Hellenthal
Dan Sullivan’s 4.3% lead is within the margin of error (+ 5.7% see below). It is not the case as Mark Davis states that “Treadwell trails Sullivan somewhere between 12 and 4 points.” Treadwell could be ahead of Sullivan by 6.1% (best case for Treadwell) or behind Sullivan by 14.5% (worse case for Treadwell).
Survey results are commonly released with a given sample size and margin of error for that sample. The margin of error reported is for a worse case percentage result (50.0%).
Typical survey samples are samples of 600, 384, and 267. The margins of error for the previous samples are + 4.0%, + 5.0%, and + 6.0%, respectively.
Why pay thousands of dollars more for an increase in accuracy of + 1.0%? Larger sample sizes are chosen for the need for accurate internal xtabs, not for the overall accuracy of results.
Each value of a variable has a margin of error. For example in the survey reported above, the total sample size is 392 (+ 4.94%). 325 (+ 5.43%) are eligible to vote in the Republican Primary, and 269 (+ 5.97%) are likely Republican Primary voters.
The table below shows the actual margins of error for the actual percentages found in out survey. In statistical language, one says one can be 95% sure that the actual results for Dan Sullivan’s observed percent of 34.8% is between 29.1% and 40.5%, Mead Treadwell’s percent of 30.5% is between 25.0% and 36.0%, and Joe Miller’s 21.4% is between 16.5% and 26.3%.
Margin
Of Survey
Candidates Error Percent Range
Dan Sullivan + 5.7% 34.8% 29.1% to 40.5%
Mead Treadwell + 5.5% 30.5% 25.0% to 36.0%
Joe Miller + 4.9% 21.4% 16.5% to 26.3%
Undecided + 4.1% 13.3% 9.2% to 17.4%
Worse case + 6.0% 50.0% 44.0% to 56.0%
Marc Hellenthal
Hellenthal & Associates
Ummm…. why didn’t hellenthall just poll primary voters? Seems like that would solve the problem.
Instead, he polled some people that don’t vote in primaries. Good work dummy.
Are some Alaskans assuming a “yes” vote on Ballot Measure 1 is a vote to support SB21? That could be a problem for opponents of repeal. Alaska law requires that a “yes” vote will cause the change requested by referendum or initiative to happen. Therefore a “yes” vote causes SB21 to be repealed.
This requirement for ballot wording is a constant source of confusion and the legislature might want to look into offering a question to the voters. For example “Should SB21 be repealed?” or “Should SB 21 remain in effect?”.
“To suggest that somehow Treadwell was only trailing by 4 or 5 points is a stretch where the number, given the sample/error size, could be twice that or more.”
The swing pulls both ways Marky Mark…”to suggest”…
Polls are statements drawn off of quantifiable data, how does your gut feeling compare to science, not to mention when you post it under an alias? You should have said, “I don’t like that this poll has Treadwell trailing within the margin of error, I want Dan to win, and everyone else needs to think he can.”
We like polls when they favor our candidate of choice and don’t when they don’t. In the senate race, I don’t have a favorite quite yet; nonetheless, there are a few issues with this poll that I would like to highlight. First, I have always thought that Hellenthal’s work, within the framework of his methodologies, was accurate information. I suspect this poll follows that history. Also, I thought it was very perceptive of you to recognize and point out to your readers that the statistical deviation on primary races increased significantly from a significantly smaller sample size. In other words, my estimation is that Treadwell trails Sullivan somewhere between 12 and 4 points. Given the carve out, as you pointed out, makes it impossible to know if Treadwell is closing the gap or of Sullivan is continuing to expand his lead. Unfortunately for us political junkies that waste too much time reading and analyzing polls, the small sample makes it virtually impossible to evaluate the race with meaningful specificity. Having said this, I still applaud Hellenthal’s work as I believe he remains in the statistical error rate, albeit that it is huge. What I take the greatest breech is the length of time it took to complete the surveys. After all, polls are snapshots , not moving pictures. Still, I think Hellenthal has proven he knows politics here in the 49th state and would caution folks of being dismissive of his work. What I find most interesting is his his results related to the oil tax question. But, the only poll that matters is the poll at the ballot box on election day. Thanks Amanda for sharing and putting the results in perspective. To suggest that somehow Treadwell was only trailing by 4 or 5 points is a stretch where the number, given the sample/error size, could be twice that or more. Most reporters wouldn’t have a clue of this type of nuance which is one of the reasons that separates you from others. Keep up the good work.