2016 Democrats, Poll Watch, South Carolina, Uncategorized

SC Poll Watch: Episode 3 (Bernie isn’t Close Yet)

February 14, 2016

There is no indication Bernie Sanders is in position to deal Hillary Clinton a knockout blow in South Carolina on February 27. Nothing is impossible in politics, but the current trend line shows a Hillary victory.

Even if he wins Nevada, which is extremely possible, it would take a larger bump than he received after Iowa or New Hampshire to make this truly close. There are a few levels of result:

Bernie Victory (even 1 vote margin is an earthquake)

Hillary Narrow Win (under 5 points)

Hillary Win (around 10 points)

Strong Hillary Win (about 15 points)

New Hampshire-Style Landslide (20 to 25 points)

Total Blowout (over 25 points)

Of the above outcomes, the top 3 favor Bernie. New Hampshire was his ideal environment, South Carolina is Hillary’s. Both Iowa and Nevada have some characteristics that favor each.

Nevada is less white than Iowa, but has younger voters. Both campaigns have advantages in a caucus. Bernie’s voters are more fired up, important when voting requires more participation than a primary. Hillary is still more organized on the ground, also crucial.

If Bernie does noticeably better in South Carolina than Hillary did in New Hampshire, it’s a moral victory for him. Especially after a win or another effective tie in Nevada. Technically, a 15 point loss is less than 22, but with more time elapsed, it would get scored as a draw.

The bottom two outcomes favor Hillary. A 20-25 point defeat means Bernie was no more successful in winning over African-Americans after making his first push than Hillary was with Millennials. Splitting the difference doesn’t work. Clinton has a huge super delegate advantage. Bernie won’t get them to change his mind without making inroads.

A 25% plus defeat indicates the Bern is simmering and he’s already peaked. Would not bode well for March 1, or his chances of winning a the nomination without some help from the FBI.

He has two weeks to climb the ladder. According to the Real Clear Politics average, he begins at the bottom, with a 31.2% deficit. Is this true?

WARNING: We don’t have enough data for full confidence. I’m doing some inferring and estimating here. It feels about right, but….

The two current polls are from CBS/YouGov and ARG. We’ll start with ARG, as this is their first South Carolina survey. They aren’t real detailed in their data breakdowns, but all favor Hillary.

She leads with men and women, voters 50 and over and 18 to 49, among whites and African-Americans, Democrats and Independents/Republicans who plan to vote in the Democratic primary. Overall, it’s a 38% lead (65/27) with 7% undecided.

Absent the detail, I’m assuming Bernie leads with white men between 18-29, and quite possibly all voters in that age group. Nowhere near enough to compete. Can we trust ARG?

In Iowa, they were pro-Bernie, showing him up 3 points in multiple surveys over the final two weeks. Only CNN/ORC was more consistently bullish on his position. Quinnipiac had similar results. CBS/YouGov had Bernie barely ahead. Everyone else had Hillary winning.

If you adjust for how the O’Malley caucus vote was distributed and the difference between actual voters and the state delegate equivalent numbers the Iowa DNC releases, my best estimate is ARG wound up slightly higher on him than the result.

In New Hampshire, they leaned pro-Hillary. Their final tracking poll was Hillary -9, the final result was -22. ARG consistently showed Bernie with a smaller lead than the median pollster. After South Carolina, we’ll know which one was the outlier for ARG.

In the meanwhile, the most positive possible spin for Bernie would let you give him 6 of the 7 undecided points, and deduct 10 points from the gap as a New Hampshire mistake tax (figure they were 3 high in Iowa, 13 low in NH).

So 38 minus 5 minus 10 = 23 point deficit. We can speculate on his chances of closing the gap, but first a look at the other poll.

The “good” poll for Bernie has him down 19 (59/40). CBS/YouGov is normally optimistic about Bernie. His best Palmetto State poll results are theirs. His CBS/YouGov numbers were ahead of the Real Clear Politics average in both Iowa and New Hampshire.

There’s a structural reason for this. Like CNN/ORC, they are not limiting themselves to a pre-existing list of registered voters. For a candidate like Bernie, who gains from new registered voters and voters who used to ignore primaries and caucuses, this methodology helps.

The same effect shows up on the GOP side with Donald Trump. Mind you, this lean doesn’t mean they’re wrong. The New Hampshire results were very close to what the most optimistic polls for Sanders and Trump indicated. There’s a more than decent chance this poll is mostly correct.

The problem is he’s still at a big deficit. What about momentum? There isn’t any evidence of any. Things could change if Bernie wins in Nevada, but they did not change after he outperformed expectations in Iowa, nor after a New Hampshire landslide.

The last YouGov poll in New Hampshire was actually before Iowa voted. Hillary wound up with 38%, exactly where the voters put her. Bernie moved up from 57% to 60% by capturing O’Malley voters without a home.

Check out YouGov in South Carolina over the past several months and see where the major changes are:

February: Sanders -19 (40% support)

January: Sanders -22 (38%)

December: Sanders -36 (31%)

November: Sanders -47 (25%)

October: Sanders -43 (25%)

September: Sanders -23 (23%)

Tying in Iowa and winning in New Hampshire made little difference. Bernie picked up two points and closed the gap by 3. Between mid-December and mid-January, he made significantly more progress.

Voters started giving the Bern consideration in August/September. This was back when Joe Biden was a fixture in polls. The reason Bernie only trailed Hillary by 23 points with 23% support was the presence of Joe.

As you may remember, October was Hillary’s best month of the campaign. Biden opted out, Hillary did very well in the first debate, and won the Benghazi hearing. As a result, South Carolinians gave her the entire amount of support Biden held.

At this point, Hillary was ahead everywhere, even in New Hampshire by a small margin. November was more of the same. People began dusting off their articles about inevitable Hillary.

You can see Bernie did make some progress in South Carolina between mid-November and mid-December. He was over 30% for the first time, though he still trailed by well over 30 points. The big move was between then and January.

Near as I can tell, the proximal reason was Donald Trump going after the Clintons, specifically calling Hillary out for attacking Bill’s accusers. Between Christmas and mid-January, Bernie improved everywhere. In Iowa, this pulled him roughly even, in New Hampshire, it gave him a wide margin. In the Palmetto State, it brought him within 22.

Absent a change in the contours of the race, Bernie isn’t going to catch up. If you want to build a scenario where he either wins, or moves far enough up the ladder for a moral victory, some combination of the following is necessary:

Nevada creates momentum that Iowa and New Hampshire did not

It’s possible. If a block of voters are waiting to see Bernie do well in a non-white state before committing, he could get more of a bounce.

YouGov indicates 67% of Palmetto State Democrats are locked in and another 26% are pretty sure, so this is more a 3 to 5 point possibility, not 10.

Hillary is further weakened by outside forces (Trump, investigation, etc.)

Donald Trump may well feel the need to chase after Hillary again if his margin in South Carolina shrinks. There’s no guarantee of equal effectiveness this time, but he is the largest variable so far.

You can probably assume those most influenced by Trump or the email scandal have already moved to Bernie or are undecided, so again, don’t expect a huge shift.

Prolonged exposure to the Bern in South Carolina changes hearts and minds

Bernie ran a lot of media and made plenty of appearances in Iowa and New Hampshire. The push began well before January.

Perhaps the January poll move wasn’t just Trump and was an accumulation of past media and appearances. Maybe he reached a tipping point.

Very skeptical. His South Carolina numbers moved at the same rate at the same time, just from a lower level. Hillary is trying very hard in South Carolina too, and held her ground over the final two weeks in Iowa and New Hampshire.

Furthermore, Bernie was in Colorado yesterday. He’s at least as focused on March 1 as South Carolina. Give him a couple points for an ad blitz if you must, but it’s hard to justify more.

If absolutely everything goes right for him, maybe Bernie makes it close. FiveThirtyEight has her today at 98% to win. There’s good reason for that.

At this point, anything over the 15 point expectations break-even line is a big accomplishment.

As soon as something changes, I’ll return with a report. For now, it’s all about Nevada. That one is up for grabs and far more likely to influence the rest of the campaign, both as a data point and talking point.




Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s