Poll shows Ranked Choice deciding thin Gideon victory, Biden electoral win in CD2

SOMESVILLE, Oct. 30, 2020 – In Wall Street parlance, an “odd lot” is the aggregation of small batches of trades which are not as easily exchangeable like a “round lot” which are standardized units of 100 shares.

In Maine’s Ranked Choice Voting, we have some of the same characteristics of odd lots – third-party voters who don’t fit standardized units of the two-party system. (Forgive the obvious double entendre.)

Odd lots are important because they are glimpses into the sentiments of small investors. One share of GOOG (Google) is about $1,500. One hundreds shares (round lot) is $150,000. Most Americans cannot buy 100 shares of GOOG.

Small investors hold stock for the long-term and some companies consider them worth fighting for. Apple, whose shares are almost 40 percent owned by individual investors, did a 7-to-1 stock split last spring, cutting its per share stock price from $645 to $92 overnight. CEO Tim Cook said he wanted to make shares “more accessible to a larger number of investors.” 

Similarly in Maine, politicians discount third-party voters at their peril, especially this year, as odd lots most likely will decide the race for the United States Senate and the electoral vote for president in the state’s Second Congressional District.

Here’s why:

In the 14 polls of the Maine Senate race since February, only once has Sara Gideon polled more than the 50 percent needed to win outright without a second round of ranked choice voting. And that poll by Quinnipiac College in Early September has been discounted as an outlier which had Gideon up by 12 percentage points. Very few experts believed it was credible.

A poll of 1,200 Mainers released Thursday by SurveyUSA had Gideon up by only 1 percent 46-45. SurveyUSA is considered the best of the crop of pollsters in Maine this year. (In 2019 It received a grade of A from Nate Silver’s fivethirtyeight.com, which said SurveyUSA prediction of elections had a 90 percent accuracy rate.) http://www.surveyusa.com/client/PollReport.aspx?g=b3c59e14-b400-40ad-919c-1f42ca3ae8b6

This is where the odd lots come into play. When asked who their second choice would be in this race, Gideon received 10 percent of the vote as opposed to Collins’s 6 percent. An overwhelming percentage of Green Party Candidate Lisa Savage’s supporters said they would give their vote to Gideon.

“Under Maine’s rules, the candidate with the fewest votes, (Max) Linn, is eliminated and those who chose Linn as their 1st choice are examined to see who was their 2nd choice,” SurveyUSA stated. “After Linn’s supporters are reallocated, the contest stands, Gideon 48%, Collins 47%, Savage 5%. Still: no one candidate is at 50%.”

Then the poll concluded:

“So, the state looks to see who the supporters of the candidate with the next fewest votes, (Lisa) Savage, picked as their 2nd choice … when Savage’s votes are reapportioned among the surviving 2 candidates, we have a winner:

* Gideon, 51%.
* Collins, 49%.

The same scenario plays out in the race for the electoral vote in the Second Congressional District. The poll shows Biden holding a 48-45 edge in CD2 but not a majority.

“But it’s a tough, uphill slog for the President the way SurveyUSA sizes up the situation. When voters from ME-02 are asked their 1st choice for President, Biden, at 48%, is as close as he can be to a 1st-ballot win but falls short of the 50% needed. Trump is at 45%, and 3 other minor-party candidates combine for 5%. The last-place finisher on the 1st ballot, Rocky De La Fuente, of the Alliance Party, the 2nd-to-last-place finisher, Howie Hawkins of the Green Party, and the 3rd-place finisher, Jo Jorgensen of the Libertarian Party, are all batch-eliminated. When their 2nd-choice (and in certain cases 3rd-choice) picks are re-allocated in real time, in the instant runoff, we have a winner in ME-02:

Biden, 51%.
Trump, 49%.

“Like all opinion research studies, this SurveyUSA poll offers estimates, which come with a plus or minus,” SurveyUSA said. “Given that 509 likely voters were interviewed in ME-02, it is possible for Trump to eclipse Biden in ME-02 and still have that be consistent with these findings. And in reverse, it is also possible that Biden will win the ME-02 on 1st ballot, with no need for ranking.”

In all Maine has had 14 polls on the Senate race with 12,235 respondents since February. That is a massive sample. Susan Collins did not win a single poll. Besides the SurveyUSA poll the others were:

Colby College*10/21 – 10/25879 LV3.34743Gideon +4
Pan Atlantic*10/2 – 10/6600 LV4.54740Gideon +7
Bangor Daily News*9/25 – 10/4466 LV4443Gideon +1
Colby College*9/17 – 9/23847 LV3.44541Gideon +4
Boston Globe/Suffolk*9/17 – 9/20500 LV4.44641Gideon +5
NY Times/Siena*9/11 – 9/16663 LV5.14944Gideon +5
Quinnipiac9/10 – 9/141183 LV2.95442Gideon +12
Bangor Daily News7/28 – 8/9500 RV4338Gideon +5
Quinnipiac7/30 – 8/3807 RV3.54743Gideon +4
Colby College7/18 – 6/24888 LV3.94439Gideon +5
PPP (D)7/2 – 7/31022 RV3.14642Gideon +4
PPP (D)3/2 – 3/3872 RV3.34743Gideon +4
Colby College/Socialsphere2/10 – 2/131008 RV3.24342Gideon +1

Half of MDI voters already returned ballots …

SOMESVILLE, Oct 31, 2020 – Claire Woolfolk and her team this morning were busy tabulating the absentee ballots which have been returned to the municipal building in Northeast Harbor where she is the town clerk. Maine allows the early tabulation of ballots for release on election night. A similar process was underway in the three other towns on MDI. As of this writing nearly half of the voters have cast their ballots. That is certainly expected to increase by Tuesday. This chart shows the number of registered voters in each town as of this week, by party affiliation.


BALLOTS OUT59198951929105009

GSJ conducted this informal survey of the town clerks’ offices on Wednesday and Thursday. The total number of registered voters, 9,748 for MDI, isn’t likely to increase much as most have already registered.

Bar Harbor Democrats have lapped the Republicans who make up only 18 percent of the electorate in that town. The GOP doesn’t fare much better in the other towns. MDI is about as blue as any big Northeastern metro and has become an important linchpin for Democrats running for office in the Second Congressional District of Maine.

Mount Desert’s out-sized influence on national politics …

SOMESVILLE. May 11, 2020 – Maine’s Second Congressional District was the only one which gave Donald Trump an electoral vote from the Northeast. And our county, Hancock, was the only one in CD2 Trump did not carry. Jared Golden is the district congressman because of the heavy support from Mount Desert. He won by only 3,500 votes in 2018.

And your vote will help decide whether the United States Senate will stay a Republican-controlled chamber. Hardly a day goes by when the national press fails to include Maine in their breathless coverage.

But the fact is that Maine is now regarded as a critical battleground state, and District 2 (in green below) – sans Portland and Augusta – is really the entire battleground.

This image has an empty alt attribute; its file name is maine_us_congressional_district_2_since_2013.jpg

Hillary Clinton carried District 1 easily, and because she won the majority of the popular vote in Maine she also received the two at-large electoral votes. But she won the state by only 22,000 votes and she lost District 2 to Trump by 10 percent.

District 2 has other distinctions. The Trump victory in 2016 marked the first time Maine has split its electoral votes (3-1) – one of only two states with that mechanism. The other is Nebraska, with five votes. The Second District also has eight counties which were part of 206 “pivot counties” in the country which went from supporting Barack Obama in 2012 to Trump in 2016.

Pivot counties
 Trump margin of victory in 2016Obama margin of victory in 2012Obama margin of victory in 2008
Androscoggin County, Maine9.38%12.78%15.22%
Aroostook County, Maine17.19%7.62%9.58%
Franklin County, Maine5.47%18.41%20.29%
Kennebec County, Maine3.58%13.46%14.78%
Oxford County, Maine12.94%14.73%16.04%
Penobscot County, Maine10.91%2.93%5.12%
Somerset County, Maine22.67%1.68%5.70%
Washington County, Maine18.44%1.60%1.01%

Mount Desert Island is clearly out of step with the rest of District 2 with the exception of liked-minded Blue Hill.

If each vote carried a weight, there is none in the United States which indexes higher than a vote here on the Quietside. That is the reason I registered to vote here instead of Connecticut, where I have another house.

A reader challenged my assertion that Mount Desert made the difference in Jared Golden’s 2018 victory in Maine’s Congressional District 2:

“I dispute your conclusion that Golden won because of Mt. Desert support.  For the first time in Maine history, we had ranked choice voting.  Will Hoar (son-in-law of Dick Wolfe) and Tiffany Bond (both independents) amassed about 8% of the vote and thus, when those votes were added to Golden’s total,  the key 1% majority on the “second ballot”.

I thanked the reader for holding my feet to the fire when it comes to accuracy.

My reply:

First of all, not all of Bond and Hoar’s votes went to Golden. About 10,000 of Bond’s votes went to Golden and 5,000 to Bruce Poliquin. Hoar’s percentage was a bit higher for Golden. Counting their votes in the second ballot put Golden over the top by 2,906 votes, from a first ballot deficit of about 2,000. But a closer look at the vote count for Hancock County, particularly the four Mount Desert towns where Golden won by huge margins, shows a difference of about 3,400 votes on MDI. Bar Harbor went 2,060 for Golden versus 638 for Poliquin. Without that overwhelming support from these four towns, the votes from the second ballot from the two Independents wouldn’t have mattered.

I do regret one oversight. I should have included our neighbors to the west as like-minded. Blue Hill, Deer Isle, Penobscot, Stonington, Castine, Brooklin, Brooksville, Sedgwick and Surry also should be credited for the Golden victory.

So, while I acknowledge that the sui generis characteristics of Maine’s Ranked Choice voting made for a compelling story, the truth is that the influence of Mount Desert in the Second District was as stated in earlier report.

Here is how Hancock County voted in 2018:

                                                           Golden         Poliquin

Bar Harbor2060628
Blue Hill1050477
Cranberry Isles2819
Cranberry Isles5412
Deer Isle665308
Great Pond1222
Mount Desert863317
Osborn — T22 MD Twp20
Southwest Harbor562303
Swans Island85101
Verona Island126124
Winter Harbor127142

Maine Senate race nastiest of all, study says

SOMESVILLE, Oct. 15, 2020 – The race between Susan Collins and Democratic challenger Sara Gideon ranks first as the most negative campaign with over two-thirds of the advertising on television being pure attack, according to the Wesleyan University media project.

Gideon has been more likely to go negative in her advertising, with more than half of her ads solely attacking compared to over a third of Collins’ ads.

Iowa’s contest between Joni Ernst and Theresa Greenfield ranks a close second at 61.5 percent negative overall despite the fact that Greenfield has not aired any pure attack ads and Ernst’s percentage of pure attacks is only 42 percent. The bulk of the negativity comes from outside groups and the party committees in that race. The Arizona, Alaska and North Carolina contests round out the top five in negativity with overall half of their ads being negative.

Table: Most Negative U.S. Senate Races (Pure Attack Only)

StateAiringsNeg %Neg %
(Dem Cand only)
Neg %
(Rep Cand only)
SOURCE: Based on ongoing Wesleyan Media Project coding of Kantar/CMAG data, which is subject to change.

This chart looks at the tone of Senate advertising in 2020 in comparison to the two previous presidential cycles. As measured through attack and contrast ads, 2020 is more negative during the comparable last few weeks in 2012 and 2016. On the other hand, pure attack spots are slightly down as a share of all ads in September and in the full cycle-to-date in 2020 compared to 2012 and 2016.

Numbers include ads aired on broadcast television and national cable between January 1 of the off year and September 27 of the election year in each cycle (left panel) and between September 5-27 of the election year in each year (right panel). Numbers include candidate, party, and group-sponsored ads.

This chart looks at the trend in negative messages (combining attack and contrast ads) in Senate races since the summer. As is often the case, negative spots make up a larger share of ads as the campaign progresses, such that in the last week they accounted for over 70 percent of all Senate ads. For a few weeks in late August, pro-Democratic ads were more negative than pro-GOP ads. This has reversed since the beginning of September, with about 4 of every 5 pro-Republican ads containing an attack on a Democratic Senate candidate.