2016 Exit Polling - the third party perspective

Political pundits are pouring over the exit polls. Understandably, the focus is on who was supporting Clinton and who was supporting Trump.

But what about the approximately five percent of voters who voted for Johnson (~ 3%), Stein (~ 1%), and others (~ 1%)? What can we say about them?

The biggest question that is always asked is what would have happened if there was only two candidates on the ballot? Did the minor party candidates cost Clinton the election?

I think the CNN.com exit poll page has the most data of the major news sites. If you scroll all the way down, I think the screen capture of one of the items in the exit poll addresses that question.

This would suggest that if the 5% who went minor party, 63% would not vote on the POTUS if there were only two choices. Of the 37% who would make a choice, a slightly larger number would vote for Trump (21%) compared to Clinton (16%). Thus, the argument that Clinton was harmed by the presence of third party voters may not hold up. However, there is one big caveat: how was this question answered in specific swing states. Unfortunately, I don't have access to exit polling data of that specificity. 

What else could be said about the minor party votes in 2016?

Demographically, younger voters (18 to 39) were two-times more likely to vote for a minor party.  This demographic also supported Clinton while those 40 and older supported Trump.

We can speculate why this age divide occurred: perhaps some younger voters are less willing to accept the binary nature of presidential elections but as voters get older this changes.

Independents are also three times more likely to vote for a candidate from a minor party compared to those with a party identification. This is not surprising since they have opted out in terms of party registration. Trump won the independents 46% to 42%.

The 2016 voter totals for alternative party candidates was the highest since 1996 when Perot got 8% and the combined remaining other party candidates got almost 2%. Does this bode well for the future of alternatives to the Democratic and Republican parties? I suspect the high numbers reflects the historic unpopularity of the two main party candidates and not a trend toward greater support of third parties. 

In the CNN exit poll charts, there were many variations on questions of the qualities of the candidates (temperament, qualified, honest, favorable/unfavorable). Of these, the one shown below was striking. Only 2% of voters surveyed viewed both candidates favorably. If the voter had a favorable view of the candidate, they inevitably voted for that candidate. Of the 18% that held unfavorable views of both candidates, more went to Trump (47% vs. 30%) and 23% of these voters opted to vote for a minor party candidate.

Finally, late deciding voters (within the final month) tended to go to Trump but a certain percentage decided to support minor party candidates. 

In 2012, Johnson ran as Libertarian and Stein ran as Green and they combined to get 1.4% of the vote. In 2016, both improved on their 2012 numbers. However, one suspects that some of that increase is of a "protest" nature rather than outright support for them. Though there is an abstract desire for more options, even with historic levels of unpopularity for the two major candidates, 95% of voters still voted for one or the other.

Electoral College

Pretty much every four years essays are written for and against the electoral college. Go ahead and google and you'll get plenty of articles. This year (2016) and in 2000 when the electoral college result differed from the popular vote, the conversation is even more intense.

Whether the electoral college is "fair" or not depends on whether one accepts the premise of the system that the founders established.

My understanding is that the founders tried to balance population based representation and state based representation in our system of governance. Hence, the House is population based and the Senate is state based. Then the question became how to determine the Presidency? They opted to construct an electoral college that reflected both population and state based aspects by linking the number of electors to the number of House and Senate members per state. The net effect of this imperfect compromise is that the electoral college mostly mirrors the population but diminishes the impact of larger states and magnifies the impact of smaller states. Taking 2012 election data: California cast ~ 13 million votes while Wyoming cast ~ 250,000. Thus, from a population perspective California weighs 52 times more than Wyoming. But in the electoral college California gets 55 while Wyoming gets 3; thus, California's weight is 18 times larger. One can argue whether this is "fair" but that is the system we have and it was intended that way by the founders when, at that time, Virginia was the big state all the other states were worried would have too much influence. Additionally, should the electoral college fail to provide a winner, the Constitution calls for the decision to be made in the House which is the population based body yet the House members vote as states. Using the House to elect the president has very rarely happened (1800 and 1824) but the design again reflects an attempt to balance state and population based representation.

The Constitution did not set up a pure direct democracy for the national government but rather a representative democracy: as determined by population in the House, as determined by states in the Senate, as determined by a mixture of both in the electoral college mechanism for the Presidency.

Mixed feelings post-election

Am still processing the elections and am slowly trying to put my finger on where I am at. In the past, in some cases I voted for a candidate that would go on to be elected (1984, 1988, 2000, 2004) and so there was satisfaction in the result. In others, I voted for a candidate who lost and so there was disappointment (1992, 1996, 2008, 2012) but satisfaction that the political process was working.

This year was unique in that I knew I would feel no satisfaction in either candidate's victory and I am concerned our political system is failing. Clinton supporters would have felt elation at the first woman president. Trump supporters would feel the voice of the forgotten working class was heard. Reluctant Clinton supporters who were #NeverTrump would feel relief that Trump lost. Reluctant Trump supporters who were #NeverClinton would feel relief that Clinton lost.

But #NeverTrump #NeverClinton voters would feel disappointment no matter who would win.

I believe that George Will is also #NeverTrump #NeverClinton. He put it this way in his column this morning:

At dawn Tuesday in West Quoddy Head, Maine, the easternmost point of the United States, it was certain that by midnight in Cape Wrangell, Alaska, the westernmost fringe, there would be a loser who deserved to lose and a winner who did not deserve to win. [......] The simultaneous sickness of both parties surely reveals a crisis of the U.S. regime. The GOP was easily captured, and then quickly normalized, by history’s most unpleasant and unprepared candidate, whose campaign was a Niagara of mendacities. And the world’s oldest party contrived to nominate someone who lost to him. To an electorate clamoring for disruptive change, Democrats offered a candidate as familiar as faded wallpaper. The party produced no plausible alternative to her joyless, stained embodiment of arrogant entitlement. And she promised to intensify the progressive mentality. [......] The moving finger writes, and having writ moves on, perhaps soon to inscribe this: In 2016, Republicans won a ruinous triumph that convinced them that they can forever prosper by capturing an ever-larger portion of an ever-smaller portion of the electorate. This kamikaze arithmetic of white nationalism should prompt the president-elect to test his followers’ devotion to him by asking their permission to see the national tapestry as it is and should be.

Political earthquake

Did not see that coming.

As a #NeverTrump #NeverClinton voter, when I left the ballot box Tuesday morning, I knew the next President of the United States would not be someone I voted for.

Nate Silver at 538 did hedge a bit on the morning of election day with his analysis piece. Excerpt:
First, Clinton’s overall lead over Trump — while her gains over the past day or two have helped — is still within the range where a fairly ordinary polling error could eliminate it. 
Second, the number of undecided and third-party voters is much higher than in recent elections, which contributes to uncertainty. 
Third, Clinton’s coalition — which relies increasingly on college-educated whites and Hispanics — is somewhat inefficiently configured for the Electoral College, because these voters are less likely to live in swing states. If the popular vote turns out to be a few percentage points closer than polls project it, Clinton will be an Electoral College underdog. 

As the night wore on, it became apparent that it was a "perfect storm" that Nate Silver described.
First, Clinton underperformed in the Obama coalition demographics.
Second, Trump's "silent voters" turned out in much larger numbers.
Third, the electoral college map yielded what is looking to be a very close popular vote total in favor of Clinton but a small but solid electoral college win for Trump.

No matter who would win last night, the nation is divided. If Clinton had won, there would still be about half the voters who did not vote for her. Trump has won and there are half the voters who did not vote for him.

No matter who would have won last night that individual will have the challenge of finding it in themselves to have the big heartedness to reach out to those who opposed their candidacy, the open mindedness to bring into the executive branch appointees who can work with people, and the humility to seek the welfare of the whole nation and not just personal ambition.

November 2016 Election Forecast

Why should all the pundits have a corner on the market on forecasts? This pajama wearing blogger can forecast too! Below are two graphs from Real Clear Politics.

Graph 1 is the 3 month data on Clinton Trump in a "two-way" race.

Graph 2 is the 3 month graph of the "four-way" race.

There are two striking features from graph 1.

1. Clinton's numbers bounce between 45% and 49%. {hitting the mark with 47.6%}

2. Trump's number bounce between 40% and 45%. {stunning at 47.4%}

This would suggest Clinton has the advantage as the high end of Trump's numbers are at the low end of Clinton's numbers - the two lines never crossed. This scenario could happen resulting in a very long night. UPDATE: Nate Silver of 538 explains the probabilities due to the uncertainties in polling and the action of undecided voters. But nonetheless, the most likely outcome is that Clinton will edge out Trump by 2 to 3%.

There are two striking features from graph 2 regarding the two minor party candidates.

1. Johnson's numbers maxed out around 9% and have declined to 4% as election day approaches. 

2. Stein's numbers peaked near 4% but is now closer to 2% as election day approaches.

This suggests that when Clinton had the larger lead, more voters entertained the idea of a minor party candidate. But as the race has tightened, probably about half of those voters have moved back to one or the other main party candidates or will cast no vote on POTUS.

Thus, what happens on November 8?

Green party candidate Nader got almost 3% in 2000 in what was known to be a close race. This indicates that Green voters are fiercely loyal. Thus, I suspect, the Stein supporters will stick with her and be joined by Sander supporters who refuse to support Clinton. My prediction Stein gets 1.7% of the vote.

As for the Johnson/Weld ticket, I think they continue to lose support. Their repeated "foot in mouth" episodes have alarmed the loosely adherent supporters. How does 3.6% sound? Because there are some states that are very strongly Clinton or Trump, the Libertarians may get some protest voters. However, the dream of a historic non-major party performance is gone with the overall tighter race between Clinton and Trump. Of course, it will still be historic for the Libertarian party in specific. 

Thus, final figures:

48.6%   Clinton {currently 47.6% likely to increase some due to California}

45.9%   Trump {currently 47.4%}

3.6%     Johnson {3.2%}

1.7%     Stein {1.0%}

0.2%     Others {0.5%}

But of course, it is the electoral college that actually determines the presidency.

Map 1 is the Real Clear Politics map with the "toss-up" states in gray.

Map 2 is the my adjustments to the RCP map.

Some of the "toss-up" in map 1 are probably not that close. Colorado, Maine, New Mexico I think are in the Democratic column as is Pennsylvania, Virginia and Michigan. Philadelphia is so strongly Democrat that Republicans can't get enough supporters in the rest of the state to off-set that advantage {Clinton support was too little too late and called to Trump breaking the "blue wall"}. Same in Detroit thus keeping Michigan safe for Clinton {Trump up 10K votes but still counting as of Thursday}. And the suburbs of northern Virginia just outside of DC is probably too strongly for Clinton for Trump to compensate with showings in the rest of the state {Clinton support came in and edged Trump}. Though Georgia is probably trending away from the GOP, don't see Trump losing that state.

As you can see in map 2, Trump would have to win all the gray locations to get to 270.

Thus, the balance hangs with these 8 elections:

Poll closing time (EST)

7:30 

North Carolina {this race was close but eventually called for Trump later in the evening}

Ohio {this race was not close and that was a sign there was something happening}

8:00 

Florida {this race was close but eventually called for Trump later in the evening}

Maine CD2 {I don't know when it was called for Trump}

New Hampshire {ultimately on Wednesday or Thursday called for Clinton but it didn't matter}

9:00

Arizona {ultimately on Wednesday or Thursday called for Trump but he had exceed 270 already}

10:00

Iowa {this race was close initially but eventually called for Trump later in the evening}

Nevada {this race was close but eventually called for Clinton later in the evening}

Probably sometime around 11pm, North Carolina, New Hampshire or Nevada will be called for Clinton thus ending the night of drama. {The drama started with Minnesota and Wisconsin were too close to call a clear sign something was happening in the "blue wall." Eventually, Clinton edged Trump out in Minnesota but the shocker was Wisconsin going Trump}. 

Final map:

Will Clinton have any "coat-tails" to bring the Senate into Democratic hands?

It is going to be close, RCP has 8 races as toss-ups. I think the Democrats take Nevada, Wisconsin, and Pennsylvania. While the GOP take Indiana, North Carolina and Florida. The cliff-hangers will be Missouri and New Hampshire. I am calling it 50-50 split with NH going GOP and Missouri Democrat. With the VP Kaine, the Senate will organize with the Democrats in control.

And so ends the season for LA Galaxy, Elfsborg, and Falkenberg fans .....

The Galaxy got the 1-0 win at home in leg one. However, in leg two, Colorado got a wonder strike to even up 1-1 on aggregate. Didn't see the game but from the reports, it seemed to be a struggle for both sides with Colorado generally getting the better of the chances. However, it wound up in PKs and Colorado won it. LA Galaxy off-season is going to be interesting as that is now two years in-a-row the team has fizzled out.

Meanwhile, in Allsvenskan, the final matches were held today. Malmo had locked up the championship previously and the last day was left for some final jiggering of the finishing rank for a few teams. Elfborg finished 5th and Falkenberg finished 16th and will be relegated. Looking forward to following the two clubs in the next year. Haven't followed the Superettan before. Will see if FFF can bounce back after a dreadful campaign.

Thus, now my soccer fan energies will be directed toward US Men's National Team, US Women's National Team, and Liverpool FC.

The USMNT have a huge showdown with Mexico on 11/11 as part of World Cup Qualification for the CONCACAF.

On the USWNT side, there aren't any consequential matches in the near term as the 2016 Olympics have concluded as have the 2015 Women's World Cup. However, with the stunning upset in the Olympics, there are some things to consider on the USWNT side. The top priority is finding a successor to Hope Solo. Additionally, as for any team, finding the balance of veterans and working in new talent.

Meanwhile, LFC is top of the table after a 6-1 win over Watford with Klopp saying "stay cool" and indeed fans would be right to not go crazy this early in the season with title talk.

Local Items for November 2016 - LA County & Culver City Measures

There are a few local matters that are on the ballot (aside from the 17 statewide measures). So let’s put on our policy hats and compare methods of taxation. Which of the following taxes sound “fair” (recognizing that fair is often in the eyes of the taxed)?

A parcel tax of $0.015 per square foot of property (Measure A).
A sales tax of $0.005 per dollar (Measure M).
A parcel tax of $99/year on single family residents, $69/year on each unit of a multi-family dwelling, and $1096/year per acre of non-residential land (Measure CW).

Voters in LA County will decide on the first two items and Culver City residents will vote on the third.

First, one must decide whether these taxes increases are for a good cause. Measure A is for parks in LA County. Measure M is for transportation infrastructure in LA County. Measure CW is for clean water in Culver City. In my book, these are all necessary and proper functions of government.

Second, one must look at details like implementation. If I understand correctly, the rate of property tax is fixed in California; thus, if a tax is to be based on property, they have to be some version of a parcel tax as seen in Measure A and CW. https://ballotpedia.org/Parcel_tax

The LA Daily News objects to Measure A on tax fairness grounds and that there is no sunset provision. http://www.dailynews.com/opinion/20161021/vote-no-on-measure-a-la-county-should-try-again-on-parks-funding-endorsement One of the complaints about parcel taxes is that they are unfair. In this case, a 3000 sq. ft. home in a rural part of LA County that might fetch $200,000 in the open market would be charged the same parcel tax amount as a 3000 sq. ft. home in an upscale part of LA County that might garner $2 million. Thus, the fairness argument has some validity. However, in California, the only way to raise money for specific needs through property owners is through the parcel tax. Thus, unfortunately, the tax fairness complaint is sort of moot. I suppose a sales tax would be more fair (as in Measure M) but that is usually reserved for raising huge amounts of tax revenue - nearly $900 million in the case of Measure M. In any case, Measure A raises nearly $100 million annually. I’ll be voting yes on measure A.

The LA Daily News also objects to Measure M for not including a sunset provision. http://www.dailynews.com/opinion/20161025/vote-no-on-measure-m-la-countys-permanent-transit-tax-endorsement The no sunset provision isn’t too bothersome to me in that transportation infrastructure is going to be an ongoing issue so a dedicated funding source may well be needed. However, it is bothersome that the initial Measure R did have a sunset (2039) and they are backing away from that and this new measure is going to raise taxes on top of the existing Measure R. Because our transportation infrastructure is such a mess, I’ll be voting yes on measure M but with the awareness the political powers that be have not be fully upfront about the costs and/or are seriously over-running their previously planned budgets.

Measure CW is highly local and no major newspaper is covering the story. The city government’s rationale for the measure is described here. No official objection was filed for inclusion in the ballot booklet. http://www.culvercity.org/city-hall/information/election-information/ballot-measure-information/clean-culver-city The problem of storm water run-off and other clean water issues have been around a long time. With this parcel tax, about $2 million will be raised annually. We just got to bite the bullet and do something about it. I’ll be voting yes on measure CW.

Crazy and unlikely but possible electoral college scenarios

Using the RCP electoral college map.

Here is today's "no toss-up" map showing a very narrow Clinton victory.

Change New Hampshire and you get the 269 tied scenario! That means the election goes to the House of Representatives with two candidates qualified for the House of Representatives to choose from.

What happens if McMullin pulls off the upset in Utah and Nevada and Virginia flip? You get no one with 270 and three candidates qualified for the House of Representatives to choose from.

538 has a similar map to the RCP "no toss-up" map with the exception that 538 has Nevada in the Clinton column.