Today’s reasons for calm

1) There’s good reason to believe that many polls are substantially under-estimating Hispanic voter.  See these two tweets from Adrian Gray.  I would suggest that with the huge variance noted in the second tweet, rather than the average, the right estimate for Hispanics is closer to the high end of +HRC.  Also, common sense and lived experience should tell us that Trump will substantially under-perform Romney with Hispanic voters.  I would argue that any poll that suggests otherwise is probably wrong.

2a) Nate Silver is smart and good with stats, but he’s not God.  It’s entirely possible his model is over-estimating Trump’s chances.  For example, Drew Linzer is smart and good with stats, and his model currently gives HRC 90% chance of winning and shows her winning Nevada and Florida (Silver currently has these just barely tipping Trump).  Honestly, my instinct is to take what I find on Kos with almost as much skepticism as what I see on Fox and I question Linzer’s judgment in choosing to ally with them, but his Votamatic pre-dates the Kos association and, more importantly, there’s an argument to be made that he’s more accurate than Silver.

2b) I was not actually familiar with Pollyvote till Linzer tweeted it, but I checked it out last night and their intriguing methodology (averaging across a variety of types of data/information on the election) has proven remarkably accurate in the past 3 elections.  Yeah, only 3 elections, but less than 1% off in national two-party vote each time.  They give HRC a solid lead.

3) Trump has a tough road without Nevada and all indications are that Democrats have built a near-insurmountable lead in early voting.  This also relates to point 1.  Poll averages show Nevada a toss-up, but that may be due to systematically under-estimating Hispanic voting.

So, HRC a sure thing?  Nope.  But I’d suggest well better than the 65% or so she’s doing at 538.



About Steve Greene
Professor of Political Science at NC State

5 Responses to Today’s reasons for calm

  1. Mika says:

    Yes, I agree. Sometimes it feels like everyone who didn’t believe in Trump’s chances at the primary is bending over backwards not to underestimate him now.

    • Steve Greene says:

      Yes. Though, as with my non-stick coating analogy the other day, he really is doing surprisingly well. Even to those of us who understand the power of partisanship (or thought we did).

  2. Brian says:

    Nitpick, but I think you mean “substantially underperform Romney”, not “substantially perform Romney”. Unless you think Trump is going to do Romney impersonations to woo Hispanic voters.

  3. Brian says:

    Also, Silver says that 538’s model does factor in early voters, because the model includes polls that do call-backs to see if previously polled individuals have already voted. But he didn’t give details about how the model would know which polls capture this info, or how they would be weighted. I doubt he’s assigning any special weight to such polls. I’ll be curious to see whether they make any enhancements to their model to account for early voting metrics in future elections forecasts.

Leave a Reply

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

You are commenting using your 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

%d bloggers like this: