New polls have the race staying at the status quo, which is good news for Biden who is leading.
A couple of interesting tidbits. Polls in Ohio are strange with Rasmussen and Zogby, both of which tend to favor Trump, having Biden up by 4 and 2 respectively. Meanwhile, morning consult has Trump up by about 5. So, Ohio is strange. High rated pollsters have not done any there in a little while so it's hard to get a sense of the race. Ohio is a must win for Trump. For Biden, it doesn't matter all too much at this point. Biden would only really need Ohio if he lost Pennsylvania, and it's hard to foresee a scenario where Biden wins Ohio but loses Pennsylvania.
Biden's leads in New Hampshire and Nevada are looking a little thin. NYT has him up by only 3 in NH. This was a state Hillary won by less than a percent in 2016. In Nevada, Times has Biden up by 4. That's more than Hillary at this time, who wound up winning the state by about 2%. If Biden lost these two states it would give him less wiggle room in the rust belt states.
The good news for Biden though has less to do with how much he's up, and more to do with how many voters are decided. According to the latest Fox News poll, Biden is up by 5 with 97% of the vote accounted for. Monmouth University also has 97% of the vote accounted for, with Biden up by 7. Around this time last year, Monmouth showed that about 92% of voters were decided. On election day, the times had 91% of voters decided and Fox had 95. This trend also holds up on a state level. For instance in Pennsylvania, on election day, the polls showed about 92-93% of people had decided. Polls being conducted now in Pennsylvania show about 96% of voters have decided.
So, that leads to two conclusions. First, there's a bit of a less chance of a last minute shift. In 2016 there were enough undecided voters that if they broke for Trump, they could have tipped the scales, even if the polls were accurate. Last minute news bad news from Clinton didn't help. However, this time around there are far less undecided voters. If Fox's poll is right, it wouldn't matter if every undecided voter decided on Trump, it wouldn't be enough. Trump has to actually change the minds of Biden voters.
And honestly... that doesn't seem likely. After four years of Trump, it's hard to see a huge amount of people changing their mind on him, absent a huge scandal rocking the Biden campaign. And... that doesn't seem likely either. Protests/riots haven't seemed to shift public opinion and I'm not sure a vaccine would either. Meanwhile, a lot of people with dirt on Trump are deciding to air it now (there's no better time to sell a book). The best hope for Trump is that the polls are really really inaccurate.
Edit: A more detailed analysis of Monmouth's University's latest poll seems to bear out my theories.
In regards to Trump, 37% say they are certain they'll vote for Trump, and 1% say they're very likely to. 50% say they are not at all likely to vote Trump. In regards to Biden, 43% say they are certain, 4% very likely, and 40% say they're not at all likely. About 7% for each candidate say they're somewhat likely. The interesting thing is that the "not at all likely" number has not shifted much for either candidate.
Trump needs to nationally get to within about 3 to have a chance in the electoral college. He has 38% of solid votes to Biden's 47%. About 1% say they probably won't vote (surely more will not). Then there's about 14% who are wishy washy but leaning towards one candidate. If we assume that Trump takes 60% of that wishy washy group, and the rest vote as expected that would lead to 46.4 for Trump and 52.6 for Biden. If the national numbers turn out like that it is exceedingly unlikely that Trump will win in the electoral college. Hillary won the popular vote by 2.1% and Trump won by about 1% in Pennsylvania, Florida, Wisconsin, and Michigan. Hard to imagine the national numbers shifting that much and none of the change being felt in any swing states.
Last edited by JWeinCom - on 14 September 2020