FL-Clearview Research: Scott + 2

[Terry the Fat Shark]

http://www.tampabay.com/florida-politics/buzz/2018/03/12/rick-scott-edges-out-bill-nelson-in-new-poll/

Scott - 43.3 (43)
Nelson - 41.3 (41)

Margin of error is 3.58 points.

[Terry the Fat Shark]

One thing to note about this:

As an example, Clearview used a turnout model based on 41 percent of voters being Republicans, 39 percent Democrats and 20 percent with no party affiliation or being registered with other parties. Vancore said the model would have ordinarily used a 4 percentage-point turnout advantage for Republicans, but Clearview scaled it back to 2 points because of recent election results, higher enthusiasm among Democrats and the "natural tendency" for mid-term elections to favor the party out of the White House.

"A few recent polls released to the media have shown samples that seem to anticipate more Democrats voting than Republicans," Vancore said. "While that could possibly be the case, we see little evidence for it at this time. "

[BudgieForce]

One thing to note about this:

As an example, Clearview used a turnout model based on 41 percent of voters being Republicans, 39 percent Democrats and 20 percent with no party affiliation or being registered with other parties. Vancore said the model would have ordinarily used a 4 percentage-point turnout advantage for Republicans, but Clearview scaled it back to 2 points because of recent election results, higher enthusiasm among Democrats and the "natural tendency" for mid-term elections to favor the party out of the White House.

"A few recent polls released to the media have shown samples that seem to anticipate more Democrats voting than Republicans," Vancore said. "While that could possibly be the case, we see little evidence for it at this time. "

I thought it was just a safe assumption.

[Gass3268]

They are basically admitting that they are using 2014 turnout model. This is my biggest fear for the fall. Some firms are going to assume a 2014 level turnout and restrict their sample to create one, while others will be more open and allow the respondents to determine the results. This is what happened in Virginia last November.

[UncleSam]

Well in 2016 more registered Rs voted than Ds. Florida has a very high number of registered Rs and a small number of Is (relative to the country), which is why it is a polarized but right leaning swing state.

In other words, this is a very reasonable assumption in a neutral year (whereas Ds will outnumber Rs in even many right leaning states simply because there are a lot more Ds than Rs nationally in terms of registration). The problem is that this is unlikely to be a neutral year, and it is therefore likely that Ds will be at least even if not outnumber Rs in the poll that matters.

If you think 2018 will be like 2016, then Scott is up 2, basically.

[Gass3268]

Well in 2016 more registered Rs voted than Ds. Florida has a very high number of registered Rs and a small number of Is (relative to the country), which is why it is a polarized but right leaning swing state.

In other words, this is a very reasonable assumption in a neutral year (whereas Ds will outnumber Rs in even many right leaning states simply because there are a lot more Ds than Rs nationally in terms of registration). The problem is that this is unlikely to be a neutral year, and it is therefore likely that Ds will be at least even if not outnumber Rs in the poll that matters.

If you think 2018 will be like 2016, then Scott is up 2, basically.

Every piece of data possible has shown that 2018 will not be a neutral year.