Let me start by saying that I know this will be moved to round table and that’s fine but if I just start the post there no one will ever see it.
Those of you who really know me know that deep down, at my core, I’m just a political science nerd and geek out on polling numbers and voting trends and political history and other things of that nature. This is NOT a thread to discuss who we want to win or who should win or what policies we like etc. This is a thread for the handful of posters here who enjoy political science based discourse and are genuinely interested in voting patterns and predictions and can remove their emotions from the conversation. I’m also doing this so that I can brag about it if I’m right and also so all of you can make fun of me if I’m incredibly wrong. So consider it a gift from me to the board.
So my methodology is pretty simple: I’m looking at the aggregate polling numbers for the closest states and comparing them to the aggregate polling numbers the week of the election in 2016 vs the real results in 2016 and looking for trends/differences and making a prediction based on those trends. One of the great myths of the 2016 election is that all of the polls got it completely wrong - that’s not quite the case. The results were, mostly, within the normal range of historical polling accuracy - they just all happened to break one specific way on election day.
The most immediately noticeable trend is that there are far, far fewer that are polling in the “undecided” or “other” category than there were at this time in 2016. This is significant because the large amounts of undecideds are the ones that won the election for the President in 2016. The vast majority of those polling in that category woke up on November 8 and decided to vote for the Republican candidate. In 2020, however, many of those same people seem to have already decided and those have tended to break for the Democratic candidate in 2020.
So in my opinion, those remaining “undecided” voters will still break heavily for the President on election day, there are just far fewer of them to break for him this time around. Those who voted for the President in 2016 that have changed their minds are already baked into the polling numbers for the Democratic nominee. I would expect most of these states to have a smaller margin of victory for the Democrats than the polling currently indicates. I’m also factoring in a trend in the last several election cycles that conservatives are under represented in polling.
If we look, for example, at Arizona. In 2016 the President led the Democratic nominee by 3 points with 9% polling as undecided. The President then carried that state by 3 points, meaning that the undecideds mostly broke evenly for both candidates. In 2020, the polls are virtually dead locked with 6% undecided - so those 3% that are ”decideds” in this cycle are already represented in the Democrat’s poll numbers. If the undecideds again break about evenly, then this will be a very closely contested state and is the true definition of a “toss up.”
There are 2 things that Arizona calls attention to that I think are important to note for this cycle: One is the changing electorate in some traditionally Republican states and the second is the impact of tight senate races.
We have seen some red states get bluer, and some blue states get redder, via migration and policy shifts. Cities like Las Vegas, Phoenix, Dallas-Fort Worth, Atlanta, and Raleigh have seen continued population increases in traditionally Democratic voters. We’re also seeing Democrats continually abandoning the blue collar “blue dog” Democratic voters they’ve relied on in the past in Ohio, MIchigan, and Pennsylvania.
Arizona (like GA, NC, TX, and IA) has a hotly contested senate race that also seems to be influencing the polls. The Democrats have a much larger lead in the senate race in Arizona than they do in the Presidential race, and it seems very likely that in a matter of just a couple of years Arizona will go from having 2 Republican senators to 2 Democratic ones. Because of the lead that the Democrats have and have consistently had in that senate race, I would say that while this state is very much a toss up for President, I find it likely that Arizona is blue on election night.
In Colorado we saw the Democratic nominee lead by 3 points with a whopping 15% of undecideds and then she won that state by 5 points. This is one of the few states where the Democrat results were better than the polls. In 2020, there are only 8% undecided in Colorado, with the Democrats leading by 12%. So the continued migration of Democratic voters into the Denver area has put Colorado squarely into the blue state category after being considered a swing state for several cycles.
Florida has been, and still is, maybe the most difficult state to predict. In 2016 the President won Florida by about 113,000 votes - which in Florida is pretty much a landslide. Going into that election night the President and the Democratic nominee were tied with 8% undecided and the President carried the state by 1%. In 2020, the Democrats lead here by 3 points with 5% undecided. So even if we account for a favorable break for the President on election day, it becomes difficult to see how the President can make up that gap simply given the smaller volume of undecided voters. Its important to note that only about half of the voters that made up the “undecided” portion in 2016 actually voted. Moreover, the Florida polls have been steadily moving further in the Democrats’ favor, with some polls having him up by five points. A couple of months ago I was very skeptical that the Democrats could carry Florida, but based on this week’s numbers, it seems likely. Honestly though, I would probably not expect any official winner in Florida for several weeks, especially if the outcome of the electoral college depends on it.
Texas, like Arizona, has experienced a blue wave of migration as Austin, Houston, and DFW all continue to be some of the most rapidly growing areas in the country with the majority of Texas implants being blue voters. The President led this state by 12 points in 2016 with 12% undecided and won by 9 - a break for the Democrats surprisingly, likely because the Republicans spent so little time and money campaigning in the state. The President now leads by 3 points with 5% undecided. So again, we see that many of the previous cycle’s “undecideds” have broken to the left. LIke Florida, however, there seems to be too small of a volume of undecideds for the Democrats to make up the gap. The senate race in Texas is also not as close as it is in other red states. The President will carry Texas again - but it is worth keeping an eye on the turnout. More people have already voted in Texas than voted in 2016 and high voter turnout (especially high early voter turnout) tends to be a good thing for the Democrats. But I have a really hard time imagining Texas going for the Democrats on election night. Just don’t see it. But it may be a very tight race here in 2024.
Iowa is another state that has a very hot senate race. But the undecideds here broke huge for the Republicans in 2016, with the President winning by 9 points after leading by 3 heading into election day. The polls are tied right now with 6% undecided vs 13% from 2016, but given the lead that the Republicans have in the senate race and given the huge break I expect for the President here on Tuesday, IA will almost certainly stay red.
Georgia not only has one tight senate race but 2 tight senate races. Democrats have been dreaming of flipping Georgia for years and have waited and waited for that Atlanta population to grow enough to turn Georgia blue and they just might be there now. You can find a post in my post history from not too long ago where I snicker at the idea of the Democrats flipping Georgia, but man the polls have pushed hard in their favor since then. The Democrats lead here by 3% with 5% undecided - the same polling numbers they’re seeing in Florida. In 2016 the President led by 5 points with 7% undecided and won by 5 points. So even in that cycle the undecideds broke fairly evenly and I’d expect them to here again. They’ve also seen very high voting turnout in Georgia, and I think Georgia will in fact turn blue with at least one of the senate races going to the Democrats and possibly both. If that does happen, it would truly be a landmark year of change in Georgia as a Democrat hasn’t won a statewide race in that state since 2004 and that was Zell Miller - who, IIRC, actually campaigned for President Bush that year and spoke at the Republican convention.
North Carolina broke heavy for the President in 2016 and was carried by the Republicans in 2012 despite the Democrats pushing with tremendous fervor there in both cycles. This is another state where the Democrats are performing very well in the senate race, but I just can’t see North Carolina going for the Democrats here. The PresIdent and the Democrats were tied in NC on election day with 8% undecided and the President won by 4 points - a massive break for him among undecided voters. The Democrats lead by 1 point here with 5% undecided so again, I think we have to figure that the undecideds will break bigly for the President and give him the win here.
I would be very surprised if Ohio turned blue on election night. The President won this “swing state” in 2016 by a huge 8 points after leading by 3 going into the night. Now the candidates are in a virtual tie with 6% undecideds - one of the highest percentages of undecided voters in the country. There were 11% undecideds in 2016 and I would expect again in 2020 that the President will get the votes needed on election day to carry the state.
The real story of the 2016 election, and of this election, will be told in PA, MI, and WI. Prior to 2016 NONE of those states had gone red since 1988, so flipping all three was quite unexpected. If the Democrats can flip them all back, then they win. If they can flip just two of them back, they likely still win. To win re-election the President really needs to carry all three of these states. He “can” mathematically win without them but if he loses all three then he has to win every single other swing state. PA is really, truly, a must win state for the President.
Pennsylvania is the closest of the three. The Democrats lead by 6 points with only 4% undecided. This means that pretty much every single undecided voter is going to have to show up and vote for the President on election day. Of course, that’s kind of what happened here in 2016. The Dems led by 4 points going in with only 8% (a small amount by that year’s standards) and lost the state by less than 1%. The same happened in Wisconsin where the Dems led by an incredibly 7% only to lose by, again, less than 1% after the 13% undecided voters swung for the President. Michigan - same story. Dems up 5 with 11% undecided and they lose by less than 1%.
But in MIchigan and Wisconsin, the Democrats currently lead by 7 and 9 points, respectively, with only 5% undecided in each state.
And that is the rub for the President. In 2016 he flipped 3 states that hadn’t been flipped in 6 election cycles and won all 3 of them by less than 1%. To win re-election he must now win all 3 of those states again - that he won by less than 1% - despite trailing by more now than he did in 2016 and despite there being far less undecided voters now than there were in 2016.
The numbers, simply, aren’t favorable for the President. They weren’t favorable for him in 2016 but they are less so now not because of the lead that the Democrats have, but because so many voters have already decided and, in fact, 2/3rd of the voters from 2016 have already voted.
And the fact that there aren’t many undecided voters out there, I think, makes a lot of sense given the strong sense of division in our country. People have dug in their heels and decided who they are supporting. There have been very few moments that have really moved the polls around a lot. The majority of voters had their mind made up long, long ago as to who they were voting for. The cake was in the oven, so to speak, before the conventions, before the first debate, before the President got Covid, before scandalous emails leaked, and on and on.
I have the Democrats winning the Presidency with 335 electoral votes to 203. The more serious question, to my mind, is will the margin of victory be big enough in enough states that the election cannot be reasonably contested in the courts? There’s no doubt both sides will aggressively pursue whatever legal means they have to affect the outcome of the election, and the supreme court has already ruled twice against Republican efforts in PA and NC regarding how long after election day mail in ballots can be counted. Notably, one of the decisions was 5-3, with Justice Kavanaugh siding with the majority, so Justice Barrett being on the court would not have mattered. The other was a 4-4 decision, and so we don’t know what the outcome would have been had Justice Barrett been on the court. Traditionally, the supreme court has been very cautious when it comes to overriding state court decisions on voting and historically supreme court justices who are feared to act with political motivations do not do so (at least not to the degree their opponents worry they will) once they are on the bench - particularly when it comes to issues like voting. I think Chief Justice Roberts really, really, really wants to avoid a situation where there’s even the perception that the court decided or impacted the outcome of the election.
I normally end My Thoughts (Long) with a Go Tigers! But this post is meant to be impartial and not endorse a specific candidate or party. So, I‘ll just end it with...
Go Vote!
Those of you who really know me know that deep down, at my core, I’m just a political science nerd and geek out on polling numbers and voting trends and political history and other things of that nature. This is NOT a thread to discuss who we want to win or who should win or what policies we like etc. This is a thread for the handful of posters here who enjoy political science based discourse and are genuinely interested in voting patterns and predictions and can remove their emotions from the conversation. I’m also doing this so that I can brag about it if I’m right and also so all of you can make fun of me if I’m incredibly wrong. So consider it a gift from me to the board.
So my methodology is pretty simple: I’m looking at the aggregate polling numbers for the closest states and comparing them to the aggregate polling numbers the week of the election in 2016 vs the real results in 2016 and looking for trends/differences and making a prediction based on those trends. One of the great myths of the 2016 election is that all of the polls got it completely wrong - that’s not quite the case. The results were, mostly, within the normal range of historical polling accuracy - they just all happened to break one specific way on election day.
The most immediately noticeable trend is that there are far, far fewer that are polling in the “undecided” or “other” category than there were at this time in 2016. This is significant because the large amounts of undecideds are the ones that won the election for the President in 2016. The vast majority of those polling in that category woke up on November 8 and decided to vote for the Republican candidate. In 2020, however, many of those same people seem to have already decided and those have tended to break for the Democratic candidate in 2020.
So in my opinion, those remaining “undecided” voters will still break heavily for the President on election day, there are just far fewer of them to break for him this time around. Those who voted for the President in 2016 that have changed their minds are already baked into the polling numbers for the Democratic nominee. I would expect most of these states to have a smaller margin of victory for the Democrats than the polling currently indicates. I’m also factoring in a trend in the last several election cycles that conservatives are under represented in polling.
If we look, for example, at Arizona. In 2016 the President led the Democratic nominee by 3 points with 9% polling as undecided. The President then carried that state by 3 points, meaning that the undecideds mostly broke evenly for both candidates. In 2020, the polls are virtually dead locked with 6% undecided - so those 3% that are ”decideds” in this cycle are already represented in the Democrat’s poll numbers. If the undecideds again break about evenly, then this will be a very closely contested state and is the true definition of a “toss up.”
There are 2 things that Arizona calls attention to that I think are important to note for this cycle: One is the changing electorate in some traditionally Republican states and the second is the impact of tight senate races.
We have seen some red states get bluer, and some blue states get redder, via migration and policy shifts. Cities like Las Vegas, Phoenix, Dallas-Fort Worth, Atlanta, and Raleigh have seen continued population increases in traditionally Democratic voters. We’re also seeing Democrats continually abandoning the blue collar “blue dog” Democratic voters they’ve relied on in the past in Ohio, MIchigan, and Pennsylvania.
Arizona (like GA, NC, TX, and IA) has a hotly contested senate race that also seems to be influencing the polls. The Democrats have a much larger lead in the senate race in Arizona than they do in the Presidential race, and it seems very likely that in a matter of just a couple of years Arizona will go from having 2 Republican senators to 2 Democratic ones. Because of the lead that the Democrats have and have consistently had in that senate race, I would say that while this state is very much a toss up for President, I find it likely that Arizona is blue on election night.
In Colorado we saw the Democratic nominee lead by 3 points with a whopping 15% of undecideds and then she won that state by 5 points. This is one of the few states where the Democrat results were better than the polls. In 2020, there are only 8% undecided in Colorado, with the Democrats leading by 12%. So the continued migration of Democratic voters into the Denver area has put Colorado squarely into the blue state category after being considered a swing state for several cycles.
Florida has been, and still is, maybe the most difficult state to predict. In 2016 the President won Florida by about 113,000 votes - which in Florida is pretty much a landslide. Going into that election night the President and the Democratic nominee were tied with 8% undecided and the President carried the state by 1%. In 2020, the Democrats lead here by 3 points with 5% undecided. So even if we account for a favorable break for the President on election day, it becomes difficult to see how the President can make up that gap simply given the smaller volume of undecided voters. Its important to note that only about half of the voters that made up the “undecided” portion in 2016 actually voted. Moreover, the Florida polls have been steadily moving further in the Democrats’ favor, with some polls having him up by five points. A couple of months ago I was very skeptical that the Democrats could carry Florida, but based on this week’s numbers, it seems likely. Honestly though, I would probably not expect any official winner in Florida for several weeks, especially if the outcome of the electoral college depends on it.
Texas, like Arizona, has experienced a blue wave of migration as Austin, Houston, and DFW all continue to be some of the most rapidly growing areas in the country with the majority of Texas implants being blue voters. The President led this state by 12 points in 2016 with 12% undecided and won by 9 - a break for the Democrats surprisingly, likely because the Republicans spent so little time and money campaigning in the state. The President now leads by 3 points with 5% undecided. So again, we see that many of the previous cycle’s “undecideds” have broken to the left. LIke Florida, however, there seems to be too small of a volume of undecideds for the Democrats to make up the gap. The senate race in Texas is also not as close as it is in other red states. The President will carry Texas again - but it is worth keeping an eye on the turnout. More people have already voted in Texas than voted in 2016 and high voter turnout (especially high early voter turnout) tends to be a good thing for the Democrats. But I have a really hard time imagining Texas going for the Democrats on election night. Just don’t see it. But it may be a very tight race here in 2024.
Iowa is another state that has a very hot senate race. But the undecideds here broke huge for the Republicans in 2016, with the President winning by 9 points after leading by 3 heading into election day. The polls are tied right now with 6% undecided vs 13% from 2016, but given the lead that the Republicans have in the senate race and given the huge break I expect for the President here on Tuesday, IA will almost certainly stay red.
Georgia not only has one tight senate race but 2 tight senate races. Democrats have been dreaming of flipping Georgia for years and have waited and waited for that Atlanta population to grow enough to turn Georgia blue and they just might be there now. You can find a post in my post history from not too long ago where I snicker at the idea of the Democrats flipping Georgia, but man the polls have pushed hard in their favor since then. The Democrats lead here by 3% with 5% undecided - the same polling numbers they’re seeing in Florida. In 2016 the President led by 5 points with 7% undecided and won by 5 points. So even in that cycle the undecideds broke fairly evenly and I’d expect them to here again. They’ve also seen very high voting turnout in Georgia, and I think Georgia will in fact turn blue with at least one of the senate races going to the Democrats and possibly both. If that does happen, it would truly be a landmark year of change in Georgia as a Democrat hasn’t won a statewide race in that state since 2004 and that was Zell Miller - who, IIRC, actually campaigned for President Bush that year and spoke at the Republican convention.
North Carolina broke heavy for the President in 2016 and was carried by the Republicans in 2012 despite the Democrats pushing with tremendous fervor there in both cycles. This is another state where the Democrats are performing very well in the senate race, but I just can’t see North Carolina going for the Democrats here. The PresIdent and the Democrats were tied in NC on election day with 8% undecided and the President won by 4 points - a massive break for him among undecided voters. The Democrats lead by 1 point here with 5% undecided so again, I think we have to figure that the undecideds will break bigly for the President and give him the win here.
I would be very surprised if Ohio turned blue on election night. The President won this “swing state” in 2016 by a huge 8 points after leading by 3 going into the night. Now the candidates are in a virtual tie with 6% undecideds - one of the highest percentages of undecided voters in the country. There were 11% undecideds in 2016 and I would expect again in 2020 that the President will get the votes needed on election day to carry the state.
The real story of the 2016 election, and of this election, will be told in PA, MI, and WI. Prior to 2016 NONE of those states had gone red since 1988, so flipping all three was quite unexpected. If the Democrats can flip them all back, then they win. If they can flip just two of them back, they likely still win. To win re-election the President really needs to carry all three of these states. He “can” mathematically win without them but if he loses all three then he has to win every single other swing state. PA is really, truly, a must win state for the President.
Pennsylvania is the closest of the three. The Democrats lead by 6 points with only 4% undecided. This means that pretty much every single undecided voter is going to have to show up and vote for the President on election day. Of course, that’s kind of what happened here in 2016. The Dems led by 4 points going in with only 8% (a small amount by that year’s standards) and lost the state by less than 1%. The same happened in Wisconsin where the Dems led by an incredibly 7% only to lose by, again, less than 1% after the 13% undecided voters swung for the President. Michigan - same story. Dems up 5 with 11% undecided and they lose by less than 1%.
But in MIchigan and Wisconsin, the Democrats currently lead by 7 and 9 points, respectively, with only 5% undecided in each state.
And that is the rub for the President. In 2016 he flipped 3 states that hadn’t been flipped in 6 election cycles and won all 3 of them by less than 1%. To win re-election he must now win all 3 of those states again - that he won by less than 1% - despite trailing by more now than he did in 2016 and despite there being far less undecided voters now than there were in 2016.
The numbers, simply, aren’t favorable for the President. They weren’t favorable for him in 2016 but they are less so now not because of the lead that the Democrats have, but because so many voters have already decided and, in fact, 2/3rd of the voters from 2016 have already voted.
And the fact that there aren’t many undecided voters out there, I think, makes a lot of sense given the strong sense of division in our country. People have dug in their heels and decided who they are supporting. There have been very few moments that have really moved the polls around a lot. The majority of voters had their mind made up long, long ago as to who they were voting for. The cake was in the oven, so to speak, before the conventions, before the first debate, before the President got Covid, before scandalous emails leaked, and on and on.
I have the Democrats winning the Presidency with 335 electoral votes to 203. The more serious question, to my mind, is will the margin of victory be big enough in enough states that the election cannot be reasonably contested in the courts? There’s no doubt both sides will aggressively pursue whatever legal means they have to affect the outcome of the election, and the supreme court has already ruled twice against Republican efforts in PA and NC regarding how long after election day mail in ballots can be counted. Notably, one of the decisions was 5-3, with Justice Kavanaugh siding with the majority, so Justice Barrett being on the court would not have mattered. The other was a 4-4 decision, and so we don’t know what the outcome would have been had Justice Barrett been on the court. Traditionally, the supreme court has been very cautious when it comes to overriding state court decisions on voting and historically supreme court justices who are feared to act with political motivations do not do so (at least not to the degree their opponents worry they will) once they are on the bench - particularly when it comes to issues like voting. I think Chief Justice Roberts really, really, really wants to avoid a situation where there’s even the perception that the court decided or impacted the outcome of the election.
I normally end My Thoughts (Long) with a Go Tigers! But this post is meant to be impartial and not endorse a specific candidate or party. So, I‘ll just end it with...
Go Vote!
