Election Observation Potpourri #2

A list of election/numbers observations to peruse as you bite your nails and obsess over the polls.

1: Is something happening in New Hampshire? Maybe so. We had three polls released yesterday in New Hampshire with Trump +5, Trump +2 and Tied. Another came out this morning showing Trump +1 in a head to head match up and tied in a four way race. New Hampshire is the only key state where we’ve seen real movement over the last week. By ‘real movement’, I mean a key state where there’s been substantial movement which could change the outcome. Going from +12 to +7 probably doesn’t matter. New Hampshire is a Blue Wall state. As I said yesterday, that’s very concerning. It is also important to remember that while it has been a fairly reliable presidential state for the Democrats, it is also an extremely white state. That means a lot of the Democrats most consistent voters – African-Americans, Hispanics, Asian-Americans, etc. – are just not there. If Trump is going to breach the Blue Wall anywhere that’s a pretty good place to start.

2. Is it a bad sign that the Clinton campaign is doing its final campaign event in the probably must win state of Pennsylvania? Probably not. If the campaign were moving toward a blowout, they might well be elsewhere. But a better explanation is that Pennsylvania is one of the few swing states with no early voting. Most of the barnstorming you’re seeing now is angled at turning out and banking early votes. It’s a logical place for the campaign to finish up. I don’t read that much into it.

3. The PollTracker Average for Nevada currently shows Donald Trump with a 2.8 percentage point lead. But the number crunchers I trust are pretty much unanimous that early voting has been so strong for Democrats that they are getting close to or perhaps have even put it away. It’s probably accurate to say that most early vote number crunchers think that Trump would now need a massive-verging on unrealistic election day turnout to turn the tide. I’ve seen no other early vote state where there’s so much unanimity on this point. Of course, early voting is too imprecise a measure to ever say these things with certainty. But based on what I’ve heard I am fairly confident that Democrats win the presidential and senate races in Nevada. The disparity between the public polls and the results on the ground may tell us something about underestimated Hispanic turnout, though Nevada has proven fairly challenging for pollsters in recent elections. It also may point to another shadow issue – the Democratic field advantage. Those may both apply in other states beside Nevada.

4: I see few signs the people watching the numbers inside the Clinton campaign are worried about the state of the race.

5: The early voting situations in North Carolina and Florida remain very tight, touch-n-go, uncertain. I see good signs and not good signs. I lack the knowledge and math skills to disentangle them or judge with any certainty which ones matter more than others. What is clear is that after last week and very early this week when African-American turnout numbers were lagging the 2012 marks, those numbers have clearly started to rebound. In North Carolina especially this is yet more evidence that the restrictions on early voting aimed at limiting African-American turnout worked. What’s not clear is whether it worked simply in moving that voting to closer to election day or in reducing African-American turnout overall. What’s clear is that the early voting electorate looks more like the Obama coalition each day in these two states, whether it’s moving there fast enough, I’m not sure. From reading people I trust, one big thing that makes it hard to judge in Florida at least is that changes in the law has increased the amount of early voting overall – converting some same day voters to earlys. That makes apples to apples comparisons very hard. My gut tells me Florida and NC are trending toward Clinton (if only trending in the sense of more positive evidence coming in). But that may be more my gut than hard numbers I can point to.

6: I have not pulled together all the numbers yet. Some of it is stuff I hear from number crunchers even though I haven’t seen the hard data. But we seem to be seeing a lot of evidence from lots of states that public polling may have (and may still be) underestimated turnout and pro-Democratic margins of Hispanic voters.