Still Too Close To Call

I have learned some important lessons as a veteran pollster. First, don’t ever try to call a race that is too close to call. I tried that in 2004 by calling a John Kerry victory even though my polls showed President George W. Bush with a narrow lead. Second, no matter who wins, the pundits will say the polls were all wrong. That was the case in 2016 when the polls were actually quite good. And third, if a party wins 30-40 seats all by one or two points, it was still a very close election and yet also a landslide. In other words, let’s all be nuanced. And let’s all recognize that many voters will make up their minds on Election Day — even deciding whether or not to vote.

In short, polls are much better describing the now and offering trend lines for tomorrow. With all of that said, two days before Americans vote, Democrats appear to be leading in enough districts to have a shot at a majority of seats in the next Congress. But, as I just described this past  Friday, too many of those leads are still too close to make an argument of a certain victory.  And as many as 9%-10% of “likely voters” are still not sure. Add to that: the amount of negative advertising on all sides is not really encouraging those holdouts to vote one way or the other. A likely voter who thinks everyone stinks does not offer us a clear path to prediction.

One thing is certain. President Donald Trump has the capacity of making everything about himself and the cable networks are enjoying doing just that. So this referendum on Tuesday is as much a statement about him and his leadership as it is about your local candidates.

So let’s look at some of the latest data. According to a brand new Washington Post/ABC News Poll, 71% of registered voters feel the economy is in “good” or “excellent” shape and 44% give President Donald Trump a positive job approval — which is where his average is in a number of recent polls. That is not a great number, but be assured that his side has already and will continue to vote this cycle. Democrats have to win 15 of the congressional districts that are considered to be toss-ups, which appears to be doable. However, just a quick survey across those districts shows too many that are simply too close — Florida 15 is tied, New York 22 and New York 19 are 1 point races, just as in Alaska-At-Large and Pennsylvania 1. Nevada 4 is a two-point race and so on. So the real factor again will be the numbers of younger voters who turnout. Thus far, less than 5% of the total early voters have been under 30. Will they come out on Tuesday?

What good is a pollster who cannot say the Sunday night before an election that he doesn’t know what is going to happen? A whole lot better than one who makes a wild prediction when he shouldn’t. Or another who doesn’t have enough data or one who seemingly has too much data — but is doing so on races too close to call.

A Democratic majority can certainly happen but let’s not rely on the same folks who were so sure that Hillary Clinton was going to be our 45th President. I used to be a risk taker and am proud of that. I am equally proud of not being a risk taker when results of polls are as close as these are. I hope you can appreciate this.

Maybe more to say on Election Day.