Nate Silver at 538: "Hillary Clinton Is Stuck In A Poll-Deflating Feedback Loop"

[Attorney in Texas]

Nate Silver, who is very robust about Clinton's prospects, has some serious analysis for her campaign to consider:

Feedback loops can produce self-reinforcing (but usually temporary) booms and busts of support. For instance, a candidate who has some initial spark of success, such as by doing well in a debate, can receive more favorable media coverage. That, in turn, can beget more success as voters jump on the bandwagon and his poll numbers go up further.

Candidates can just as easily get caught — or entrap themselves — in self-reinforcing cycles of negative media attention and declining poll numbers. Hillary Clinton looks like she’s stuck in one of these ruts right now.

The Washington Post’s David Weigel recently observed that voters were hearing about only three types of Clinton stories, all of which have negative implications for her. First are stories about the scandal surrounding the private email server she used as secretary of state. Next are stories about her declining poll numbers. And third are stories about how Vice President Joe Biden might enter the Democratic presidential race.

Weigel isn’t exaggerating: For roughly the past two months, voters have heard almost nothing about Clinton apart from these three types of stories. I went through the archives of the news aggregation website Memeorandum, which uses an algorithm to identify the top U.S. news stories of the day. I tracked whether there was a Clinton-related headline in one of the top three positions at 11 a.m. each morning and, if so, what the subject was.2 You can see the results below:

The whole analysis is insightful and well worth a full read beyond this excerpt.

[Jester Messiah]

1. So if she wants to turn things around...

She needs to get out there and do something positive. Not just say stuff, but DO stuff, noteworthy enough to grab the focus away from her negatives. I guess dancing with Ellen wasn't quite substantive enough to close that particular deal.

That leads one to ask though... what sort of task could she undertake at short notice that would yield a big enough return to grab focus in a positive way?

[hootinholler]

2. The spontaneity plan should do the trick

[JackInGreen]

3. She's doing it

Like the public event she canceled to go pass the hat. Spontaneous.

[Avalux]

6. I know! Go to Dallas for a fundraiser and public event, then cancel the public event.

That is a surefire way to turn things around.

[hifiguy]

14. Jebus Haploid Christ riding a unicycle and juggling fish.

The people running her campagn are as utterly useless as DWS or a collection of left-handed football bats.

Not that there's much of a difference there.

[Jester Messiah]

15. My guess is they got their positions much in the same way DWS did

Kissing the right asses and playing "the game." Actually having any skill at one's supposed position? Entirely optional, and probably rigorously disincentivized.

[arcane1]

32. What's that about?

[Avalux]

35. Here's a link:

http://trailblazersblog.dallasnews.com/2015/09/hillary-clinton-in-dallas-tuesday-for-fundraiser-scraps-public-event.html/

[arcane1]

36. Hmm, I wonder what the "conflict" is.

Am I the only one creeped out by the spokesperson cheerily saying “We’re going to raise a significant amount of money” from 150 people?

[Avalux]

37. Who knows; but it appears as if she choose the smaller event with lots of cash involved.

That's not a good message, no matter how you slice it. I'm not creeped out I'm disgusted.

[frylock]

12. One way NOT to do that..

talk about feedback loops.

Hillary Clinton in Dallas Tuesday for fundraiser, scraps public event

[HereSince1628]

20. But Nate says it's the media doing this to her

According to him none of it is her doing, none of it is her fault. And -THAT- is part of the problem, it's always the fault of a vast conspiracy of haters.

[Jester Messiah]

38. The media just want to sell ads

They do that by being as interesting or, failing that, controversial as possible. If Hilary gave them something to run with, they'd run with it. Until she can give them something more interesting than serverghazi or her death-spiraling poll numbers, that's what they're going to focus on.

[HooptieWagon]

4. The 15 scheduled Democratic debates will put her in a positive light.

Oh, wait....

[magical thyme]

5. maybe she could try sending direct shipments of her favorite brand household and personal stuff...

also some "lighter" interviews for the ladies. You know, Kim Kardashian, Kanye and stuff.

That oughta do it.

[Ichingcarpenter]

7. Well I guess this space should be added to the Hillary Board Game

[azmom]

29. Hahahaha.

[magical thyme]

8. and aside from his obvious dig at Bernie in the opening 'graph, there is this:

There are a number of ways the spiral of negative stories could end:
■New news stories could disrupt the cycle.6
■Biden could opt out of the race and possibly also endorse Clinton.

■The trickle of new revelations on the email story could stop — as it largely did from April through June.

Except that each month a new tranche of emails will be released, for several months to come. Iirc, at least January. And there is her upcoming testimony. Possible testimony by Bryan if they offer him full immunity (which I expect they will). A report from the FBI on what they've found.

■Clinton could lift her poll numbers, perhaps temporarily, with an aggressive advertising spend.

$4M spent in NH resulted in another 10% drop in the polls. The more people see her, Nate, the more they see through her. Get it?

■Clinton could hit some bedrock of support — her most loyal voters — beyond which her poll numbers wouldn’t decline much further.7

Like in '08?

■Clinton could fall far enough that the “Clinton comeback” story becomes more compelling to the media than the “Clinton in disarray” story, as happened late in the 2008 Democratic primary campaign.

Too little too late in '08, and with a seemingly better alternative. Same as '16.

But keep clinging to your "establishment candidate" over the "insurgent candidate" who could win IA and NH but lacks support anywhere else. Which, btw, he lacked in IA and NH until he started campaigning there, and then that support skyrocketed. And now, btw, he's starting to campaign "anywhere else" so can only go up in standings.

[Gothmog]

9. Again this feedback loop will not stop Hillary Clinton from being the nominee

The loop really does not matter until we get past the Oct 22 testimony before Howdy Gowdy committee and Joe Biden announces that he is not running. These two events will help break this loop and will be in plenty of time for the primary season.

[Attorney in Texas]

23. The only 2 things that can stop Clinton are Clinton and Sanders. Biden is a non-issue (either the

Clinton campaign totally fails -- not a bad news feedback loop, but a total failure like Bush and Walker and Rubio are currently failing -- and Biden comes in BECAUSE CLINTON KILLED OFF HER OWN CAMPAIGN or Biden has no chance and probably no interest in a bare-knuckle fight with Clinton for the nomination).

If Clinton does not kill off her own campaign, the only one standing between her and the nomination is Sanders backed by the progressive wing of the party.

I'm hopeful.

[Gothmog]

25. I agree that Biden is not a factor

Biden is rumored to be delaying his decision until late October or early November which means that he is not likely to get into the race until and unless there is an implosion in the Clinton campaign http://www.democraticunderground.com/1251593782 This time frame is too late to mount a serious challenge to a viable Clinton campaign but is plenty of time to be the nominee if the only opposition is Sanders. Contrary to the amusing speculation by some Sanders supporters, Biden is not going to get into the campaign in order to split the establishment vote so that Sanders will be the nominee.

As for Sanders, time will tell. As you know in Texas one of the key voting blocks in the primary is the African American vote. I was at a bundler event for the Clinton campaign hosted by some African American professionals. From what I have seen, Clinton has strong support in the African American community right now. In 2008, I was on my county's nomnation and credentials committee and I saw the depth of the support for Hillary Clinton in the Latino community. Clinton won the Texas primary but lost the caucus round in 2008.

BTW, I signed up for the Clinton legal/Victory counsel program. This program is far more organized than the Obama voter protection program including a great memo on what lawyers can do for a federal campaign in terms of providing services without running afoul of the Federal campaign finance regulations. I was happy to see that my actions for Kerry Edwards in 2004 and Obama in 2008 and 2012 were all proper under these guidelines.

[WillyT]

10. K & R !!!

[Agschmid]

11. Really good read, thanks.

I really respect Nate his analysis is often spot on.

[kenfrequed]

13. Am I wrong?

Or did Nate Silver just do that whole article without mentioning Bernie Sanders by name even once?

[Attorney in Texas]

17. You're wrong. "She’s fallen behind Sanders in most polls in New Hampshire and some polls in Iowa."

[kenfrequed]

34. Ah...

Was it just a drop in blurb? Was any ink spent on talking about how it could be more Sanders gains than Hillary losing? Because framing it as Hillary losing is sort of implying those supporters somehow were hers or that she was entitled to them.

[kenn3d]

16. Sure. It's just the (usually) temporary bad x, bad y, and bad z negative feedback loop stuff.

All serious polling analysts are familiar with this phenomenon. It's perfectly normal.

I too found it amusing that Nate's latest "serious analysis" sounds curiously like hedging his earlier bets in the light of serious evidence that he's been wrong, and still includes only one sentence referencing Senator Sanders:

" She’s fallen behind Sanders in most polls in New Hampshire and some polls in Iowa, and she increasingly also trails Republicans in hypothetical head-to-head matchups."

I wonder if this analysis considers how many candidates have lost the IA caucus and the NH primary, and then gone on to win the Presidency?

Otherwise he only finds the Sanders campaign worthy of mention in 2 footnotes to the article:

1. In his formulation, Weigel grouped Biden-related stories with Bernie Sanders-related stories. I tend to think of Sanders-related stories as falling in the “bad polls” category instead, at least in the way that they’re usually framed by the media. But it doesn’t really matter — the point is that there’s a lot of overlap between these types of stories and that they tend to reinforce one another. ^
7. It’s possible that Sanders will soon brush up against the limits of his support — about one-third of Democratic voters nationally are white liberals, and Sanders’s support is concentrated overwhelmingly among that group.

So it seems that Nate is now saying that his consistently "robust" Clinton inevitability forecasts maybe weren't completely wrong... She might still be inevitable, and Bernie's plateau may possibly still be coming soon...
It's just that her continuously bad personal headlines, bad polling data, and bad opponents (mostly the one who's not in the race) have her in a kind of negative feedback rut right now...

That's his analysis?
Okay then.

So I guess it's a good thing she still has plenty of money... Because after that predictably stupid Brocksmear debacle gets added into this already toxic mix, she may soon find her future donations falling into the same kind of negative feedback loop that her polling trends are in.

I suspect that the Sanders campaign may possibly have some (more) surpri$ing re$ults to report in the upcoming FEC cycle also btw. And I'd bet the polling trends will continue to show diminishing gaps wherever he isn't already in the lead.
There has been more than plenty of notice given... to not underestimate him.

If Hillary's camp fails to heed such notice, "they could easily get caught — or entrap themselves" in some kind of self-reinforcing cycles of negative somethingoranother, and it sure won't be Nate's fault.

[PowerToThePeople]

18. No one thought the Titanic would sink

Once it started to, people got off it as fast as they could.

[Fearless]

19. But either she wins or he was wrong in his prediction... such dilemmas.

[Attorney in Texas]

21. That's not how statistics works. If Clinton has a 70% chance of winning (which is what the betting

markets and polls suggest AT THE MOMENT), then if you could run the election 10 times, she would lose 3 times and win 7 times.

Assessing a candidate's chances of winning at 70% defines that candidate as a favorite but is not a guarantee.

Moreover, any statistician would expect the odds to fluctuate up and down when assessing a candidate's odds 5 months before the first vote in the first contest is cast.

I suspect that Clinton's odds will rise higher and sink lower between now and the Iowa caucus. My hope is that Sanders can parley a win in Iowa and a second win in New Hampshire into an upset victory, but we ought not lose sight of the fact that a Sanders win would be an upset and Clinton is (currently) the favorite to win the nomination.

[Gothmog]

27. Silver has discounted the effect of winning both Iowa and New Hampshire

Both states are 90+% white voting states and are no representative of the key states that Sanders needs to do well in if he wants to be the nominee. If Sanders is unable to broaden his appeal from the narrow demographic segment currently supporting Sanders, then Clinton will be the nominee

[Fearless]

33. He's doing quite well thanks

[99Forever]

22. Polls are propaganda tools, nothing more, nothing less.

I don't much buy any of them, as whomever posts them, has an agenda, ALWAYS.

That said, the trends they show, are more indicative of what's really going on. YMMV

[Attorney in Texas]

24. If you are a Sanders fan, you ought to like polls because many use polls to separate O'Malley from

Sanders.

If O'Malley was polling at 20%, this would be a hugely different race. Not because 20% polling would mean anything in and of itself, but because O'Malley would get much more attention and would raise more funds if he were polling at 20%. If O'Malley were polling at 20% now, it would help him build the infrastructure so that he might be more competitive in February and March when it matters.

[99Forever]

28. I'm me, not some made up strawman.

You don't get to tell what I "like" or "don't like." I explained my position, I'm sorry that didn't fit whatever narrative you are pushing.

"If O'Malley was polling at 20%... yada yada" And if pigs had wings they would be birds.

[Gothmog]

26. Romney and Rove believed that in 2012 and were surprised when Nate was right

Both Rove and Romney convinced themselves that Nate Silver was wrong and that the polling were skewed. Both opinions were wrong. Silver and others are doing a decent job on predicting and modeling these contests.

As to relying on trends, Silver's models factor in trends but also look at things like demographics. Sanders is doing well in two states with 90+% white populations. Sanders is not polling well in states with more diverse populations and most people believe that Sanders will not be the nominee unless he can broaden his appeal to more than the current base of supporters. The Super Tuesday states do not look favorable to Sanders.

[99Forever]

30. ...

[Major Hogwash]

31. I've come to the conclusion that Nate Silver is a vampire zombie.

He sucks and wants to eat our brains at the same time.
How in the hell is anyone supposed to control what the headlines say?

Gimme a freakin' break!