The terrific electoral vote mapping site 270towin.com has an excellent new page for seeing the updated national and state polls using simple pie charts. On one clean, uncluttered, easy to scan page, you’ll see how the accumulated recent averages of national and state polls stand as of that day. Each pie chart shows the current support for Clinton, Trump, and the “others”. You can start with the latest national numbers, then scroll across and down the page to see who has the largest slice of pie in each state. The folks at 270towin take the average of all available recent polls, so you need not drill down into the bottomless questions of poll credibility, etc. Below we’ll use the new page to gain some powerful state-by-state insights to this extraordinarily complex contest.
How to set up the web page
On the new 270towin polls summary page look for the SORT button (on a forest green background) just above the “National” pie chart on the lower left of the top row of the pie charts display. Above the SORT button is a check box to show the 3rd party and undecided breakout in each pie chart. Check the box and wait for the page to auto-refresh.
Now, select another useful option. To the right of the green SORT button you’ll see a set of radio button choices Choosing “Competitiveness” will bring the Battleground and other closely-contested states up toward the top of the array display, saving scan time. Select that button and wait for the page to auto-refresh again.
Now we’re ready to put this new 270towin tool to work.
Tip: As you scan down the page, watch the gold/green/gray slices of the Battleground state pies
The new all-polls-on-one-page illustrates how the combined third party/undecided “other” voters are making the estimation of “likely voters”, and the final November win/lose percentages so difficult this year. (Note that 270towin’s “other” is the same as my “3PU” group name.) In a recent post I’ve dwelt at some length on estimating the likely voting behavior of the 3PUs, so we need not belabor the problem again here. My “net-net” conclusion was that, when all the votes are in on November 9th, “slippage” from the 3PUs will be due to half the Stein, Johnson, and (some) undecided voters deciding that Clinton is safer for the country than Trump. These last-minute decisions will be the 2016 version of the expected traditional “natural” movement away from the third parties as voters decide they want their vote to count.
How can we use the pie-charts page to gauge the potential impact of the expected wave of last-minute voter bloc decisions, especially in the key swing states that will actually decide the national election?
To answer this question for a given state, you’ll first need to add the percentages and subtract from 100 to deduce the Green Party’s voter preference share in each pie chart. In Nevada for example, add Clinton’s 41%, Trump’s 43%, Johnson’s 8%, and the undecided (as of today) 7%. Subtracting from 100% yields Stein’s current share, 1%, and thus, a total of 16% 3PUs.
In past elections where there was a significant third party and/or a large proportion of undecideds, roughly half of the total 3PUs have gone to each of the two main candidates. So, if that happens again this year, who cares about the gold/green/gray “others” pie slices in 270towin’s terrific one-page array?
You do, if you’re a Progressive or Democrat.
Remember please, my conclusion a few days back was that very few of these 3PU switcharoo votes will go to Trump:
Trump has all the supporters at this point he’s likely to get. He has so poisoned himself in terms of character, competence, and temperament that very few 3PU voters are going to change their mind in the polling booth and vote for him. The Libertarians are uniformly opposed to him as are the Greens, so he can, at best, hope for a few undecideds to take him up on his “Whaddya got to lose?” offer.
Unlike in previous elections, where we’d expect the undecided group to split roughly evenly between the candidates, I suggest that the electorate this year is so exceptionally angst-ridden we could see the undecideds splitting 40% for Clinton, 30% for Trump, and 30% failing to vote. Combine this breakout with my assumption that half of the 3Ps will switch to Clinton while the other half stick with their 3P candidate, and we have a wholly different take on the likely final results in November than the simplistic reading of the current polls by the TV commentators would indicate. In a phrase, I’ll call this the “November surprise” of 2016.
Is a “November surprise” a real possibility?
Yes, as can be seen when we do a state-by-state “flip-ability” estimate for the key Battleground and other closely-contested states.
Assuming our last-minute voter behavioral analysis model is basically correct, let’s scan the array of state pie charts on the new 270towin pie charts page. When we do, we see how, in state after state, a switch of just a small portion of the 3PUs (shown in the gold, green and gray slivers at the top of each chart) to Clinton in a few states would decide those states for the Dems. If this same “November surprise” effect occurred in just a few of the red-leaning states, it would turn the overall national contest from the so-called “dead heat” the TV media is selling to a Democratic rout of Trump and his Republican fellow travelers.
Again, let’s use Nevada for a test case. Above we saw that the projected tally based on polls is 41% for Clinton, 43% for Trump, with a total of 9% for the 3Ps and 7% undecideds. My analysis therefore is half of the 9% goes to Clinton, while the other half toughs it out. Meanwhile of the 7% undecideds, I’ll call 3% to Clinton, 2% to Trump, and 2% not voting, That would give Clinton 41 + 4.5 + 3 or 48.5% total to Trump’s 43 + 2 or 45% total, with 4.5% staying with the third parties and 2% not voting. From a 2 point deficit using today’s polls, Clinton pulls out a comfortable win on November 8th.
Now let’s examine the other truly key Battleground states using the new 270towin array.
- If just one-third of North Carolina’s 15% 3PU voters decide at the end to be safe with Clinton, she flips the state’s 15 Electoral Votes (EVs)
- Ditto with Ohio, where a quarter of the state’s 18% 3PUs would give her a squeaker, but a win is a win…
- And in Iowa (!), now being Trumpeted as a likely win for the Orange windbag — a 25% Clinton final decision among the large 20% of “other” voters would move Iowa into the Clinton column
- A 25% last-minute decision for Clinton among Florida’s 12% 3PUs would let her take Florida by a margin larger than President Obama’s in 2012.
And it may not end there. Look at solid red Texas (where Clinton is actually now leading in the registered voter polls): a last minute decision for Clinton by just 25% of Texas’ huge 27% “other” category would give her an even-money chance of nabbing the state’s 38 EVs. In Georgia, the same scenario is possible, if a third of the state’s 15% “other” voters opt at the last minute for Clinton — not inconceivable considering the Atlanta urbanites’ distaste for Trump, coupled with an expected surge in African-American turnout. In Arizona, the story is even worse for the Republicans, since just a twenty percent last-minute decision for Clinton among the state’s enormous 24% “other” bloc would flip this red state to blue.
What will drive 20 to 40 percent of 3PUs in these key states to change to Clinton? The deep Clinton ground game in most of them will be a huge facilitator, but mostly it will be the realization by voters of the likely dangerous consequences of allowing Trump to win. For most of these last-minute switchers, it will start and end with the Supreme Court, and the prospect of as many as three Trump appointments, turning our country into a fascist paradise for the super wealthy and their corporations.
Yeah, I understand: my “November surprise” scenario sounds too good to be true. And it may well be. Admittedly, my handy-dandy formula for estimating the average final split percentages in the state results will be “off” for any given state. And my estimate of 50% of the 3P votes going at the end to Clinton may be too high. Ah, but it could also be too low: if the fear of Trump in the Oval Office continues to grow, we’d see perhaps 70% of the combined Green/Libertarian voters switching. Considering the large proportion of 3P voters in today’s polls, a 70% average move of these voters to the Democrats would seal Ohio for Clinton, in addition to her wins in Florida, New Hampshire and North Carolina. Game over, Orange guy; back to the penthouse and your other reality show.
A warning for the faint of heart
We cannot predict with any certainty how large a percentage of the 3PU/”other” folks will, in the end, switch to Clinton in any given state. Nonetheless, IMO the argument that a) there will be a significant degree of “others” slippage in these final weeks and b) that the great majority of the switcheroos will go for Clinton are sound hypotheses.
But understand the nature of the last-minute decision process: whatever slippage comes Clinton’s way will not be evident until the votes are counted. It means the contest is likely to be a nail biter all the way to the end, with the results not fully known until the early hours of November 9th.
We also need to remember that the “likely voter” models of the various pollsters would be wrecked (statistically) if as few of 5% of Republicans stay home in disgust or protest. If an additional say, 5% of Trump’s core voters decide he’s hopeless beaten and don’t bother voting, that would further deflate the likely voter estimates for the Republicans. And if, say 5% more of his supporters finally realize what a ruthless, truthless con man he is, we can envision a shocking tidal wave of switches to Clinton. (I’m thinking primarily of the non-college white women here, especially the ones who have a daughter.) The combination of these three sub-bloc behaviors on Election Day would not only defeat Trump: they’d have a profound impact on the Republican candidates for the House and Senate. But we’ll deal with that prospect in a future post.
Ironically, should the former Sanders supporters ultimately decide not to vote, the impact on Clinton would not be significant, since these hard-care voters are now supporting the Libertarians, the Greens, or are milling about in the undecided category. The larger point is that likely voter numbers are suspect this year, and could well have a bearing on my estimates above.
Understand as well that an “October surprise could throw off my “November surprise” estimates. But it’s not clear what such a surprise might be, nor can we say it would help Trump more than Clinton (as the TV talking heads and pundits mostly seem to assume.) I’d bet we’ll see more leaked emails, and that they won’t significantly affect the results. The possibility of a major external terrorist attack cannot be ruled out, of course. But such an event could turn out to be to Clinton’s benefit, as scared voters opt for the more stable, experienced, and competent candidate. My personal favorite October surprise.is that after two debates, or even just one, Trump whines the debates are rigged, and walk away from another face-off. That would definitely cost him permanently in the polls.
No way around it; there’s weeks of uncertainty ahead! But here’s a silver lining to keep you calm. Note that I’ve only listed eight of the states that could conceivably be in play, using my behavioral analysis of the “others” category in these states. Just one red-to-blue flip among these states due to pro-Clinton 3PU slippage would effectively block Trump from any path to 270 EVs.