by Rod D. Martin
September 5, 2018
Rasmussen today has the Generic Ballot at D+4 (46-42). This is within the safety zone — Republicans historically win the House even at D+5.
But it’s worth noting that it’s also part of a worrisome trend. Rasmussen had the race dead even two weeks ago. And other (less reliable) polls have it much worse.
The catch? Well, there are a lot of catches. For example, yesterday’s ABC News poll has it D+14 (52-38). Yeah, that sounds terrible…until you realize that ABC’s poll has about half the sample size it needs for a good poll. And also samples “registered voters” — which tend to lean D — rather than “likely voters”, which tend to reflect what’s actually going on.
Oh, and if history is any indication (because the WaPo story conveniently hides the details), this poll probably intentionally grossly oversamples Democrats. How do I know? Because they did it all through the fall of 2016, even producing a poll not long before the election that showed Trump losing and Hillary winning by a whopping 12 points.
“Oh but Rod: other polls show a similar Democrat lead.” Yes they do. And all of them have similar issues.
I’m not thrilled about these numbers. But I know how this game gets played. Fake (or fake-ish) polls are used to demoralize our side and turn out their side (to borrow their phrase, “suppress our turnout”, and our fundraising, and our activism, and whatever else they can suppress). They just ludicrously claim that having a secure election suppresses turnout, while in fact, they’re using supposedly neutral news sources to do the real thing.
It didn’t work in 2016; or at least, it didn’t work enough. Don’t let it work this time either.