My 2024 Presidential Election Preview
What the polls tell us about what to expect next week
Election Eve Update (11:00 a.m.): When I made my original post last week, it felt like Trump had the momentum after surging to a very narrow lead in some of the most crucial swing states and national polls. Since then, Kamala Harris has gotten some good polling results that have shifted this race back closer to 50-50:
These last-minute shifts could just be noise. But, there is also reason to think that Harris has real momentum among Latino voters and late-deciders. A well-respected and historically accurate pollster, J. Ann Selzer, released the results of a poll this week that showed Kamala Harris surging into a lead in Iowa!
Again, all of this could just be noise, but you will recall that in my original prediction, I leaned heavily on the “Big Mo” in picking who would win some of the closest states. I also pointed out that a shift back toward Harris of only 0.4% could be enough to sway the election. If in fact these latest polling results are indicative of something real, it could absolutely swing the results of this election. Here’s my updated projection based on the latest polling numbers:
The only real change is that I have flipped my Nevada and Wisconsin projections, with Wisconsin going blue and Nevada going red. While that is not enough to change the outcome, it does make this race much closer because it means that all Harris would have to do to win is to flip Pennsylvania, a state where Trump’s lead has fallen from 0.5% to 0.3% in the last week! This race is going to come down to the wire.
Original (10/30) Post:
This morning, I thought I’d take a break from the series explaining my voting decision and put on my political science hat to predict what we will see next week.
Swing States
Jacob Long, Assistant Professor of Mass Communications at the University of South Carolina has produced an excellent resource that tracks polling errors in national elections at the state-level going back all the way to 1998. On that page he compiled average polling error and bias over the two Presidential elections and one mid-term election between 2016 and 2020.
For purposes of my analysis, I’m going to label as swing states those states which are currently polling within that state’s historical margin of error using Jacob Long’s averages. Here is that list:
This list is probably a little over-inclusive. Nate Silver recently published an article explaining why this election will probably come down to just seven of these (the 2nd district in Maine, New Hampshire, and Ohio would require truly historic polling errors to go against the conventional wisdom). This list also includes all five states that flipped from Trump in 2016 to Biden in 2020: Michigan, Pennsylvania, Wisconsin, Georgia, and Arizona. That doesn’t mean that it’s impossible for there to be a surprise in a state not listed here. It just means that if there is, then we are likely dealing with a completely unprecedented polling error (beyond even what we saw in 2016 and 2020) and the election will likely be an astounding landslide for one candidate or the other.
Assuming no completely unprecedented polling errors on election day, then, we are looking at a clean slate map that looks something like this:
Now, let’s walk through these states one at a time. First, let’s look at the states where we have a pretty high degree of confidence about the result:
Ohio
Current Silver Bulletin Polling Average: Trump +5.9%
Current 538 Polling Average: Trump +8.0%
2016 Presidential Election Result: Trump +7.9%
2020 Presidential Election Result: Trump +8.1%
2022 Senate Election Result: Vance (R) +6.6%
Cook Political Report Rating: Solid R
Ohio was long considered a bellwether state in U.S. Presidential elections, the quintessential swing state, voting for the winning candidate in every election from 1964-2016. That changed when it cast its votes decidedly for Donald Trump in 2020. It arguably shouldn’t really be on this list of swing states as it would be fairly shocking to see it go blue, but to be consistent with my method, I’m including it here.
Projected Winner: Donald Trump
New Hampshire
Current Silver Bulletin Polling Average: Harris +5.7%
Current 538 Polling Average: Harris +5.1%
2016 Presidential Election Result: Clinton +0.3%
2020 Presidential Election Result: Biden +7.3%
2022 Senate Election Result: Hassan (D) +9.2%
Cook Political Report Rating: Likely D
Although the margin in the polling averages is slightly larger than historical polling error, I thought I’d include this state in my analysis in light of some recent headlines. It was a shocking display of the bad news Kamala Harris has been getting in the polls lately when a pair of polls came out last week putting this state in play. New Hampshire, which was once a surprising Republican strong-hold in New England, hasn’t voted for a Republican Presidential candidate since 2000. Despite the surprising polls we recently saw, the fundamentals here just don’t line up to expect a Trump win.
Projected Winner: Kamala Harris
Maine 2nd Congressional District
Current Silver Bulletin Polling Average: Trump +5.0%
Current 538 Polling Average: Trump +7.1%
2016 Presidential Election Result: Trump (margin not available)
2020 Presidential Election Result: Trump +7.9%
2022 Senate Election Result: N/A
Cook Political Report Rating: Likely R
Maine and Nebraska are the only states that apportion their electoral college based on congressional district results. Both states award two electoral votes to the winner of the state-wide vote, and apportion their remaining electoral votes with one vote going to the winner of the popular vote within each Congressional district. The 2nd Congressional District, while theoretically within the margin of error, appears likely to once again fall into Trump’s column as it did in 2016 and 2020.
Projected Winner: Donald Trump
Electoral Map Update
The three states (or districts) we’ve looked at so far, while technically meeting my definition of a swing state, are very likely to go in the direction I’ve projected here. If Trump loses Ohio or Harris loses New Hampshire, while not necessarily indicative of a major landslide, it would definitely signal major problems for their campaigns. The remaining seven states are much more difficult to predict. So, I thought now would be a good time to update the map:
As you can see, Kamala Harris is sitting at 226 electoral votes, 44 shy of the 270 needed to win. Donald Trump is sitting at 219 electoral votes, 51 shy of the 270 needed to win.
The next pair of states we will look at are the ones where polling averages have a net margin greater than 1.2%. Why 1.2%? Nate Silver did an analysis of the weight-average statistical bias of polls in the final 21 days leading up to Presidential elections going back to 2000 and found that, on average, there was a polling bias of 1.2% in favor of Democrats. Obviously, that number was famously higher in 2016 and 2020, but there was actually a bias in favor of Republicans in 2012 and in the 2022 Senate elections. All-in-all, I’m just looking for some rational cutting point and this seems as good as any.
Arizona
Current Silver Bulletin Polling Average: Trump +2.3%
Current 538 Polling Average: Trump +2.0%
2016 Presidential Election Result: Trump +3.5%
2020 Presidential Election Result: Biden +0.3%
2022 Senate Election Result: Kelly (D) +4.9%
Cook Political Report Rating: Toss Up
Arizona was one of the five states Biden managed to flip from Trump’s column in 2020. He did so by the skin of his teeth with a margin of barely more than 10,000 votes. Interestingly, this is one of the few states where the polling error may have actually worked in Biden’s favor as Donald Trump had a 0.6% advantage in the RCP polling average at this time in 2020. With that said, he has a stronger advantage in the polls now and it would take a polling bias in favor of Republicans greater than any we have seen in Arizona since before the data we have begins (1998) in order for Harris to pull of the upset.
Projected Winner: Donald Trump
Georgia
Current Silver Bulletin Polling Average: Trump +1.9%
Current 538 Polling Average: Trump +1.6%
2016 Presidential Election Result: Trump +5.1%
2020 Presidential Election Result: Biden +0.2%
2022 Senate Election Result: Warnock (D) +2.8%
Cook Political Report Rating: Toss Up
Georgia was another one of those states that flipped from red to blue between 2016 and 2020. It was also by the narrowest of margins: just under 12,000 votes. Going into that election, the RCP polling average actually showed Biden ahead by 0.8%, more than half a point higher than is actual margin of victory. This time around, Trump has a more significant lead going into election day. You only have to go back to 2012, though, to find a year where the Republican candidate underperformed his polling numbers by more than Trump’s current polling margin. This one is definitely in play, but I do lean toward Trump reclaiming Georgia this time around.
Projected Winner: Donald Trump
Electoral Map Update
With Donald Trump reclaiming two of those swing states that broke for Biden in 2020, let’s take a look at the updated map:
As you can see, Kamala Harris is still sitting at 226 electoral votes, 44 shy of the 270 needed to win. Donald Trump is up to 246 electoral votes, just 31 shy of the 270 needed to win.
The remaining states, quite frankly, are way too close to call based on polling data. They are all within the range of normal polling error. Still, let’s look at each state to see what we are dealing with.
North Carolina
Current Silver Bulletin Polling Average: Trump +1.1%
Current 538 Polling Average: Trump +1.1%
2016 Presidential Election Result: Trump +3.6%
2020 Presidential Election Result: Trump +1.3%
2022 Senate Election Result: Budd (R) +3.2%
Cook Political Report Rating: Toss Up
North Carolina is a really difficult race to predict. There was a lot of concern that Hurricane Helene could dampen turnout in North Carolina, especially in the most impacted areas of the state (which tend to be rural, predominantly Republican areas). The good news is that early indications are that turnout is looking really good in North Carolina. Hurricane Helene is still likely to have an impact here, though, as it has become one of the big issues on the minds of North Carolina voters, and could be impacting poll accuracy as it is very possible that those impacted are less likely to respond to surveyors. In other news, the Harris campaign recently cut ad buys in North Carolina which, in light of higher GOP turnout in early voting numbers, is being taken by some as a signal that they do not feel confident here.
Projected Winner: Donald Trump
Michigan
Current Silver Bulletin Polling Average: Harris +0.9%
Current 538 Polling Average: Harris +1.1%
2016 Presidential Election Result: Trump +0.3%
2020 Presidential Election Result: Biden +2.8%
2022 Senate Election Result: N/A
Cook Political Report Rating: Toss Up
Michigan is a state that Trump very narrowly won in 2016, but returned firmly to the blue column in 2020. The polls show Harris sitting on a lead right at 1%. Unfortunately for Harris, polling overestimated Democratic performance in Michigan by significantly more than that margin in both 2016 and 2020. That doesn’t mean the same pattern will necessarily repeat, but it is impossible to rule it out.
Projected Winner: Kamala Harris
Pennsylvania
Current Silver Bulletin Polling Average: Trump +0.4%
Current 538 Polling Average: Trump +0.2%
2016 Presidential Election Result: Trump +0.7%
2020 Presidential Election Result: Biden+1.2%
2022 Senate Election Result: Fetterman (D) +4.9%
Cook Political Report Rating: Toss Up
Scott Tranter, the director of data science at Decision Desk HQ, wrote in a memo that “whoever wins PA has an [approximately] 85% chance of being the next president.” While it is possible for either candidate to win this election without winning Pennsylvania, if you wanted to know who was going to win and you could only get the results from one state, Pennsylvania is your best bet. The polls show it has essentially been neck-and-neck for months.
The polling bias in Pennsylvania hasn’t been as strong in the last two Presidential elections as in some of the other midwestern states, either. And, this year, because its importance in this election is so widely known, it has seen an incredible quantity of polling. At the end of the day, this state is just impossible to predict. Donald Trump seems to have all of the momentum now, though, and, when it is this close, I tend to side with what President George H.W. Bush called the “Big Mo.”
Projected Winner: Donald Trump
Wisconsin
Current Silver Bulletin Polling Average: Harris +0.3%
Current 538 Polling Average: Harris +0.9%
2016 Presidential Election Result: Trump +0.7%
2020 Presidential Election Result: Biden +0.6%
2022 Senate Election Result: Johnson (R) +1.0%
Cook Political Report Rating: Toss Up
Wisconsin is another state that is really impossible to predict. They went for Trump by less than 1% in 2016, re-elected an incumbent Democratic Senator comfortably in 2018, went for Biden by less than 1% in 2020, then elected a Republican Senator by a margin of 1% in 2022. This state just keeps oscillating and will likely come down to get-out-the-vote efforts across the state on election day. This one could easily go either way, which is a problem for Harris who had a pretty significant lead in polls for most of August and September, only to see it evaporate in October.
Projected Winner: Donald Trump
Nevada
Current Silver Bulletin Polling Average: Harris +0.1%
Current 538 Polling Average: Harris +0.2%
2016 Presidential Election Result: Clinton +2.4%
2020 Presidential Election Result: Biden +2.4%
2022 Senate Election Result: Masto (D) +0.9%
Cook Political Report Rating: Toss Up
Nevada and its six electoral votes don’t typically grab a lot of headlines, but is probably the single tightest race in the polls going into next week’s election. Much like Wisconsin, Harris had a more significant lead in September and August, in line with what this state has looked like in previously elections, going for Clinton in 2016 and Biden in 2020. So, if I were consistent with what I said about Pennsylvania and Wisconsin, I would go with Trump who clearly has all the momentum after erasing Harris’s lead. I can’t get over the fact, though, that Nevada has not gone red since 2004. That likely indicates a strong ground game for Democrats.
Projected Winner: Kamala Harris
Final Prediction
If everything goes as I’ve predicted, Donald Trump wins the Presidency by a closer margin that what we saw in 2016 or 2020. It’s essentially the 2016 map all over again, but with Michigan staying blue.
I don’t want to oversell this prediction. As I mentioned before, all ten swing states/districts identified above are technically within the margin of error. Kamala Harris’s most likely pathway to victory is to hold Wisconsin (where she actually has a narrow lead in the current polling averages) and pull out a win in Pennsylvania. That only requires a shift in her favor of 0.4% in the polling average. That’s how narrow the margin is right now. And, with a week to go, it should be noted that is roughly equivalent to the amount of movement we’ve seen across the board in polling averages over the last week. Though this past week’s movement has been in Trump’s favor, who’s to say this next week won’t push back in Harris’s direction?
Let’s say there’s another polling error equivalent to what we say in 2020 with Harris underperforming her polls by about 4.1%. Donald Trump would win 312-226, a larger margin than he had in 2016. What if it went the other way? Kamala Harris would win 319-219, nearing Barack Obama’s margin of victory over Mitt Romney in 2012. Anything in that range would not be all that surprising based on current polling and past polling error.
When Will We Know?
The bad news is that there has been little done to effectively address the problems we’ve seen in recent years where key battleground states can’t be called for days following the election. The ones we are most likely to be waiting days for: Nevada, Pennsylvania, and Georgia. It is unlikely that either candidate wraps up 270 votes without any of those three states unless there is a significant partisan bias in the polling that results that leads to virtually all the battleground states going one direction or the other. Looking at my prediction above, if you take those three states off the board, Donald Trump would be sitting at 256 electoral votes waiting to see if he could get one of Pennsylvania or Georgia to fall his way.
Here’s the good news. It is possible the “red mirage” or “blue mirage” effect we have seen in recent elections (based on when the state reports early/absentee vote results—which have recently favored Democrats—relative to Election Day vote results—which have recently favored Republicans) will be less significant this year than in 2016 and 2020. That’s because more Republicans appear to be voting early than in previous years, greatly diminishing the partisan skew to those votes, likely making them look more like Election Day votes. That could vary by state, however, so pay attention to reporting on election day as to which votes are coming in and where the outstanding vote is before you make any rush judgments based on the early returns.
What About the House and Senate?
It looks very possible Republicans could get unified control of Congress and the White House for the first time since the first two years of Donald Trump’s first term.
In the House, it is essentially dead even. The Cook Political Report gives a slight advantage to Republicans:
If all the races went according to Cook’s projections and the toss ups split evenly, Republicans would come away with a very narrow 2-4 seat majority. This lines up pretty well with the FiveThirtyEight projection which gives Republicans a 53% chance of winning control of the House.
Meanwhile FiveThirtyEight gives Republicans an 89% chance of controlling the Senate. Right now, Democrats have a narrow 51-49 seat majority. 23 of those Democratic seats are up for election this cycle, though, to only 11 Republican seats. That means Democrats have a lot more ground to lose. The Cook Political Report does not currently rate any of the Republican-held seats as toss ups, indicating Republicans are likely to retain all of their current seats. Meanwhile, it rates 1 current Democratic held seat as a Solid Republican seat in the upcoming election, another leans Republican, and 4 are toss ups. That indicates Republicans stand to gain 2-6 seats in the Senate, likely giving them their largest Senate majority since Trump’s first term (though still not enough to overcome Democratic filibusters).