A forecast of the 2024 US presidential election -- are the assumptions and results reasonable?

Politics & Elections
1

The post below is my original work. It is based on nothing more than past presidential election results from Wikipedia and the mathematical application of the eight assumptions laid out below. Compared to the work of professional pollsters and election modelers, this is a very simple and, frankly, perhaps naive model. Nevertheless, I have enjoyed the process of putting this all together.

My apologies for the length of this OP. It is my hope that it can be the basis of fruitful discussion about how the 2024 presidential election can reasonably be expected to turn out. Since this entire scheme sprung from my own head, there are obviously blind spots and things I’ve failed to consider.

…

ASSUMPTIONS

Summary: In the upcoming 2024 presidential election, overall voter turnout nationwide is likely to fall below 2020’s record-setting level. Support for the Democratic candidate (Biden) will likely stay stable, while support for the Republican candidate (Trump) will likely drop slightly. Third parties will pick up some of the ex-Republican votes as the kind of ‘none of the above’ votes that were relatively prevalent in the 2016 election.

  1. US voter turnout, after seven consecutive elections between 36.5% and 38.0% (1956-1980), started becoming more chaotic in 1984. Since then, turnout percentage has trended upwards, but with more volatility. 2020’s percentage of ~48% was the highest ever, jumping nearly six points over 2016.

  2. Accordingly, it is reasonable to expect voter percentage to drop from 2020’s all-time high. Arbitrarily calculating the sum of 2016 & 2020 vote counts, then dividing by 2, roughly splits the difference between 2016’s and 2020’s turnout percentages (about 45% predicted for 2024).

  3. While Nikki Haley will not win the 2024 Republican nomination, her level of support and ability to attract funding reliably indicate that Donald Trump’s voting base is at least marginally smaller than it was in 2020. Accordingly, we will build in an arbitrary 5% drop in the Republican candidate’s 2024 vote count (see below for formula).

  4. Third-party candidates accounted for 5.7% of the vote total in 2016. In the 2000 election, that percentage was about 3.7%. In 2004, 2008, and 2012, it was below 2.0%. In 2016, many voters were disatisfied with both Trump and Hilary Clinton. In 2020, there was a groundswell in Biden’s favor to mobilize and defeat Donald Trump which dropped the third-party percentage back below 2.0% once again. Looking at the lay of the land in 2024, there is apparently a level of disatisfaction with the prospect of a Trump-Biden rematch. Accordingly, it is reasonable to expect an increase in the number of third-party votes. We will add in an arbitrary 5% increase to the third-party 2024 vote count (see below for formula).

  5. This assumption is admittedly more tentative: There seem to be multiple factors simultaneously pushing the 2024 vote percentage for Biden downward and propping them upward. Instead of attempting to account for these factors individually, I just threw them into a hat together (admittedly not a rigorous approach) and called them, collectively, a push. So for the purposes of this OP, the predicted 2024 Democratic vote count will only be affected by the lowered predicted turnout in assumption #2 above (see below for formula).

  6. Since the 2024 presidential election will certainly come down to a small number of swing states, I will apply the five assumptions above to six swing states: Pennsylvania, Wisconsin, Michigan, Georgia, Arizona and the ripe-for-turning North Carolina. Three other states sometimes considered “swing states” I will reckon as safe for one of the major parties and thus not consider them in this analysis: Nevada for the Democrats, Ohio and Florida for the Republicans.

  7. Based on the assumptions above, the predicted 2024 major-party votes and third-party votes are derived by these formulas:

  • Democrats: ([2016 votes]+[2020 votes])/2
  • Republicans: (([2016 votes]+[2020 votes])/2)*0.95
  • All Others: (([2016 votes]+[2020 votes])/2)*1.05
  1. The predicted total sum of 2024 presidential election votes cast is based on a sum of the results of the three formulas shown in assumption #7.

…

RESULTS

So what do these assumptions yield? I’ve built a little handy-dandy Excel sheet which can crunch the numbers quickly. It’s an oversimplified approach, but convenient for roughly testing numerical assumptions (e.g. what if Trump loses 10% of his 2020 support while Biden stands pat?).

Here’s what the 2024 election looks like following the assumptions above – scroll down in the preformatted-text box to see forecasts for all six swing states:

**USA popular vote**

	    2024 count	 2024 pctg
-------------------------------
  Dem	73,568,508	  50.95%
  Rep	65,174,181	  45.14%
  3rd	 5,645,351	   3.91%
Total  144,388,040   100.00%


**Pennsylvania**

	   2024 count	2024 pctg
-------------------------------
D	    3,192,335	 49.87%
R	    3,015,493	 47.10%
3	      193,923	  3.03%
Total	6,401,751	100.00%

Democrats won PA by 80,555 in 2020. Predicted to win by 176,842 in 2024.


**Michigan**

	   2024 count	2024 pctg
-------------------------------
D	    2,536,440	 50.18%
R	    2,341,463	 46.32%
3	      176,564	  3.49%
Total	5,054,466	100.00%

Democrats won MI by 154,188 in 2020. Predicted to win by 194,977 in 2024.


**Wisconsin**

	   2024 count	2024 pctg
-------------------------------
D	    1,506,701	49.11%
R	    1,432,347	46.69%
3	      128,794	 4.20%
Total	3,067,842  100.00%

Democrats won WI by 20,682 in 2020. Predicted to win by 74,354 in 2024.


**Georgia**

	   2024 count	2024 pctg
-------------------------------
D	    2,175,798	 48.91%
R	    2,161,705	 48.59%
3	      111,372	  2.50%
Total	4,448,876	100.00%

Democrats won GA by 11,779 in 2020. Predicted to win by 14,093 in 2024.


**Arizona**

	   2024 count	2024 pctg
-------------------------------
D	    1,416,655	48.36%
R	    1,384,191	47.25%
3	      128,408	 4.38%
Total   2,929,254  100.00%

Democrats won AZ by 10,457 in 2020. Predicted to win by 32,464 in 2024.


**North Carolina**

    	2024 count	2024 pctg
-------------------------------
D	    2,436,804	48.62%
R	    2,432,668	48.54%
3	      142,461	 2.84%
Total   5,011,933  100.00%

Republicans won NC by 74,483 in 2020. Democrats predicted to win by 4,136 in 2024.

…

The predicted 2024 electoral college map would thus look like this (link to make your own map). It’s basically the 2016 map all over again, with North Carolina flipped to Biden and the electoral votes per state updated after the 2020 Census:

…

…

I can poke a hole in this scheme here and there already. Looking at Michigan specifically: How much do concerns over Gaza affect Arab-American votes in the state? Still, this forecast model does allow Biden to lose over 180,000 votes in Michigan and still win the state’s electoral votes.

Thank you all for any comments, criticisms, and suggestions.

2

I sure hope you’re right, but I think current polling is more reliable that data from 4 and 8 years ago… And that current data doesn’t look good for Biden. All the swing states are leaning to Trump. https://www.statista.com/statistics/1428865/general-election-swing-state-polling-biden-trump-us/

Plenty of time for things to change, but at this point I don’t see any reason to rejoice.

3

Good post, but I think the enduring lesson of 2016 and 2020 is that you always have to budget in a certain Trump factor for safety’s sake and assume Trump will overperform whatever he’s expected/predicted on Election Night.

4

There’s nothing unreasonable there. That is definitely a perfectly plausible prediction. Though also so is a much narrowed Biden win or a Trump victory of varying degrees.

5

That speaks directly to assumption #3 (that the Trump voter base appears to have shrunk slightly since 2020). To me, even if Haley is not neck and neck with Trump, her demonstrated level of support in the primaries makes me think Trump is flaking off supporters. Not a ton, a few percent. But that would be enough.

6

I don’t think this follows. Most of Haley’s support seems to be coming from voters who never voted for Trump to begin with, but rather a coalition of Never Trumpers and independents and Democrats crossing over to vote in the Republican primaries. Over 70% of her voters in the New Hampshire Republican primary were not even registered Republicans.

Also, plenty of candidates have faced spirited challengers in the primary only for those voters to come home in the general election. Trump himself only took about 45% of all the votes cast in Republican primaries in 2016. And yet in the general election he added over 2 million voters compared to the previous Republican ticket.

7

@Horatius was good enough to post this in the active “anti-Trump resistance” thread (which are in turn cross-posts from other P&E threads). Trump’s support is showing cracks.

EDIT: To summarize the link – (1) Biden’s campaign out-raised Trump’s in January 2024 by almost 2 to 1, and (2) the conservative CPAC convention has notably poorer attendance this year than in recent years.

8

That’s OK. We’re counting on small percentages here, not a sea change. If “most” of Haley’s support come from the groups you mention … some percentage comes from 2020 Trump voters. And that ‘some’ doesn’t need to be large to be impactful.

9

A five percent drop in his general election vote total attributable solely to Nikki Haley supporters would be enormous. And as I mentioned, even the tiny, tiny fraction of Haley primary voters who voted for Trump last time will come home in the general election when the choice is between Trump and Biden.

10

Maybe this should be its own thread. I am not trying to hijack, I really think it fits here.

I think the assumptions made in the OP are reasonable enough, realizing that there is enough uncertainty to make the results so blurry that, while reasonable, the opposite is also reasonable. Unfortunately, I don’t believe it will matter who gets the most votes.

Major Hijack of thread, please do not reply. {What Exit?}

My forecast for the 2024 Presidential Election is that it will be surrounded by so much chaos that an attempt will be made to throw it to the House to decide. From Trump’s viewpoint, the only way to escape the legal mess in which he has managed to embroil himself is to regain the Presidency. That is a must-have. The only sure-fire way to win the Presidency is to have the House decide, where he owns all the Republican Representatives, and Republican-majority states outnumber the Democratic-majority states.

It’s the same thing he tried in 2020, except that he started too late with the planning stage. It wasn’t until after the election that he came up with the half-baked plan to cause enough chaos to force a House vote. This time, it will be planned from the start. Expect so much interference, computer issues, lawsuits, and rampart pandemonium in the vote counting centers that even solid Biden states won’t be able to certify their results. If there’s one thing Trump can do it’s create a mess.

I don’t like the prospect, but I don’t see any way to stop it.

12

Everything I just hid is hijacking this carefully written OP. There are probably other threads for such conjecture and opinions or you can start a new one.

But try to avoid going off-topic so far for cleanly written OPs especially.

I will reopen the thread in 16 minutes.

13

This topic was automatically opened after 15 minutes.

14

Nevada is currently Trump country:

2024 Nevada: Trump vs. Biden Polls

I’m not going to say the OP forecast will prove wrong. But as of today, the Democratic ticket is behind. And, unlike in most cycles, the great majority of voters have made up their minds, or at least know what way they lean, early.

Could events occur to change which way leaners lean? Sure. But the discussed forecast makes assumptions favorable to the Democrats.

15

Agreed. Other groups must make up part of that five-percent drop. Disaffected conservative voters that just stay home, or leave president blank while voting down-ballot, or who vote third-party … they can all be part of that five percent.

A five percent reduction in Republican support is not a sine non qua number, in any case. Change that percentage to 4.8%, and North Carolina flips back to the Republicans. At 4.4%, Georgia flips back. That would still leave Biden with a 287 - 251 Electoral College win – uncomfortable, to be sure, but a victory none the less.

Below are the other break-even percentages for the remaining four swing states listed in the OP. A reminder: These percentages are the amount of Republican support that can be lost in each swing state and the Republicans still be able to eke out a victory in a hypothetical 2024 general election with roughly a 3% drop in overall turnout.

  • Arizona - 2.8%
  • Wisconsin - 0.01% (Republicans have no wiggle room here)
  • Pennsylvania - 0.0% (Republicans have to gain significant support to win PA)
  • Michigan - 0.0% (Same as PA)

Even if Biden lost NC, GA, and AZ … he’d still win in the Electoral College 276 - 262.

16

Not yet – Biden is drawing closer as the polls progress forward in time. The biggest outlier poll that makes up 1/3 of the RCP Average score is a Trump +11 from four months ago (~11/1/2023). Since then, Trump has not surpassed 48% in the polling. There have also been several margin-of-error polls (statistical dead heats) since that +11 that RealClear (to their credit) lists, but does not include in their average.

17

Heh – give Nevada to the Republicans, as well, and that still leaves a 272-268 nail-biter in the Democrats favor.

Republican support doesn’t have to crater – it really just needs to ebb a bit, and that if we assume (as in the OP) that Democratic support stays more or less static.

18

Perhaps you should just split it off as a new thread.

19

In recent months, three non-partisan polling organizations have been polling Biden-Trump in Nevada. The rolling average is based on the latest poll from each organization.

With so few pollsters, of course there is room for error.

Re the one poll where Trump was +11, outliers exist, but arguing them away is a questionable analytic practice likely to increase the error.

20

I sadly agree. I also agree with the OP’s prediction that third party votes will make up a bigger percentage this year, and I think it’s fair to say that third party candidates always pose a greater risk to Democratic candidates than to Republican candidates, both because Republican voters tend to be more loyal, and because the rules for contingent elections by the House, should that occur, favor the Republicans to a preposterous degree. We can think of a contingent election as the Electoral College on steroids.

As for the Electoral College itself, likewise baked into the Constitution, the prospects of eliminating that are slim to none, because the very process of altering the Constitution, itself, also strongly favors Republicans. There’s a reason there haven’t been any amendments in recent decades, unless you want to count the Twenty-Seventh in 1992, and the subject matter of that Amendment was pretty much outside of the space of the ongoing confict between left and right.

Before then, up until conservative states began revoking their previous ratifications of the Equal Rights Amendment, recent Amendments had generally trended in a liberal direction, in cases where that was relevant to their subject matter.

We all felt so relieved and happy in December 2020.

Yesterday I was crunching some of the numbers from that election. It turns out that Biden’s victory came down to about 315K votes out of ~25M total votes from six battleground states. This works out to about 1.2%. Compared to the total national vote of about 155M, it’s .02%

That’s how much Biden really won by last time, 0.02% of the national vote, because the system was (and is) so heavily tilted against him.