Flunking out of Electoral College
An unbiased and informative view of polling, the Electoral College and the coming presidential election. Part 2 of 2.
The election of Donald Trump in 2016 in a lot of ways was a disaster. Trump pulled us out of an important treaty with Iran, appointed Supreme Court judges who have damaged women’s reproductive health (among other things), and fanned the flames of social unrest. But I believe the Democratic Party’s response to Trump’s election was also a disaster.
The Fog of War
When Hillary Clinton lost the 2016 Presidential Election to Donald Trump, a lot of Democrats lost their minds. Clinton had had a favorability rating of 54.5% (much higher than more recent politicians) and she resoundingly won the popular vote. Yet she lost every swing state, leading to Trump winning the Electoral vote and becoming president.
Below is Nate Silver’s prediction just before the election, followed by the actual electoral vote count.


It’s the Enthusiasm, Stupid!
Silver predicted that five swing states—Wisconsin, Michigan, Pennsylvania, North Carolina, and Florida—would go to Clinton but they went to Trump instead. Why? Well, because he’s a huckster. But why did Clinton lose? The short answer is she ran a lousy campaign, and when she lost she and her supporters pointed fingers everywhere—racists, sexists, xenophobes, homophobes, gun nuts, Russia, and the FBI—everywhere except at themselves.
But for the long answer let’s look at the Electoral College.
Expanding the Vote
The American revolution really was revolutionary. Among many radical achievements the revolutionaries expanded voting from the narrow property holder classes, in some states, to all free male adults.
But there’s a problem with doing that.
For an election back in colonial days the voters could convene at the courthouse on the designated day, and the sheriff could just tell the people who want candidate A to stand over to one side and the people who want candidate B to stand over to the other. Not much room for fraud! The elected sheriff is counting bodies that everyone can see. If he tries to cheat by picking the side that has fewer voters, not only will everyone know he’s cheating, but the side that’s he is cheating is numerically larger, and so will have a definite edge if a brawl ensues. Plus the sheriff will eventually be running for reelection himself, so cheating the majority in plain sight is a bad idea.
When it comes to public voting like this, however, it probably won’t be very close. When it becomes clear that one side is winning many people who might have voted differently will switch to be on the winning side. A slim majority quickly grows.
But people don’t just choose a side to be with the majority, which brings us to the biggest weakness with public voting. If you owe someone money, if you need a job, if you want to impress someone you probably will want to be seen to vote however they vote. In actual public voting the rich and powerful can have massive influence over everyone else. No one wants to lose their farm because they’re seen voting for someone the banker hates. That’s why public voting only really works when everyone is a property owner and has independence to stand up to pressure.
So to expand voting for a more democratic, inclusive world—as the revolutionaries wanted to do—they needed ballots. Secret ballots. So powerless people can vote their conscience without powerful people seeing. Most secret ballots are some form of written ballots. Written ballots do require voters to read a bit, but it doesn’t take much literacy to distinguish between a series of names. So if the vote is on people, there’s a list of names, and the voter can circle a name and put the ballot into the slot of a box where it will join lots of other ballots. No one will know how anybody else voted.
Of course written ballots mean someone has to make sure the ballots are printed, each voter gets one and only one, the ballots are correctly put into the box, the boxes are secured, and of course the ballots must be counted.
Which raises new problems. While paper ballots reduce the potential corrupting influence of public pressure, they also increase the possibility of fraud: ballot stuffing, ballot destroying, and miscounting (deliberate or accidental). To circumvent these challenges there needs to be a time and place for the counting with observers. This is what Republicans disrupted in the 2000 Florida recount. The lack of this is what many Kerry supporters were suspicious of in the 2004 presidential election in Ohio, and what Trump supporters were suspicious of in the 2020 election.
Electoral College
These potential problems are easy to understand to anyone who thinks about them, and the founders of the U.S. thought about them.
Enter the Electoral College.
Every presidential election rolls around and I talk about the Electoral College and it seems to set people off. But in my experience few people think through how to hold an election on a continental scale in the time before telegraphs or telephones. It’s the 1780s. Picture carrying boxes of paper ballots on the back of horses or mules over roads for days and weeks with who knows how much ballot stuffing and ballot dumping, or just ballots lost in crossing rivers or storms, or stolen. Who transports them? How are those people chosen? Wherever they converge how is anyone able to judge which of the people carting the votes is legit and which isn’t?
It’s a logistical nightmare.
Instead—to minimize possible fraud—the fledgling U.S. had each state pick electors—however they deemed feasible—and the electors would gather and publicly vote for the president.1 That would make the system more democratic than just having the U.S. Congress choose the president, but minimize the ballot shenanigans, and best of all when irregularities might be suspected, voters would blame it on their state governments or their state electors rather than the federal government. Since state governments are closer to the voters, it’s easier for voters to turn officials suspected of wrongdoing out of office.
Let’s get the usual complaints out of the way:
Yes, since the invention of the telegraph there are plenty of other ways to conduct elections. The Electoral College is like powdered wigs and buckle shoes—a weird vestige of the past that we don’t need to follow.
However, the federal and constitutional laws regarding the Electoral College have not negatively impacted either political party in any significant way.
Despite what you may have heard the Electoral College was not created to preserve the power of slave owners. It had nothing to do with chattel slavery, enslavers, or the enslaved.2
The Electoral College was not a system picked instead of a popular vote since there was no way at the time to have a popular vote.
The Electoral College was a compromise between the big states and the small states, without which the small states would have refused to join the Union. We would be a continent of mid-sized powers often at war. Giving the small states more of a voice than strictly determined by population keeps them from getting crushed.
The weighting of smaller states votes has not favored either the Democrats or Republicans in any presidential election. Both parties win both large states and small states.
Since it’s creation the Electoral College has been modified in many many ways so that while it looks like a Rube Goldberg machine it has not led to outcomes that violate the popular will at the federal level.
Finally, and most importantly:
The reason the electoral college outcome often differs from the popular vote outcome is because MOST STATES GIVE ALL THEIR ELECTORS TO THE WINNING CANDIDATE. This is entirely due to state laws and has nothing to do with federal law or the constitution.
To repeat: The reason popular and electoral votes differ so much is that in 48 of 50 states whoever takes 51% of the vote takes 100% of the states’ electors. That is entirely due to state not federal law. That means winning several swing states by 51% beats winning fewer swing state by 90%. Hillary Clinton won the states she won by huge margins and lost the states she lost by slim margins, so she won the popular vote by a lot, but that doesn’t matter. Trump won more electoral votes, so he won.
The 2024 Electoral Vote Picture
But enough about the past. What’s happening next week in the next exciting episode of the Electoral College?
Well, bearing in mind all the challenges of polling I discussed here, let’s take the swing states one by one.
Here is the 2024 map from site 270 to win with states where one party or another is far enough ahead that we can be reasonably sure our imperfect polling methods are good enough:
Maine & Nebraska
You’ll notice that Maine and Nebraska divide up their electoral votes. They are the only non-winner-take-all states in the nation. One effect of this is that candidates don’t campaign there as much since it won’t make as much difference as winning a winner-take-all state. So you see why in a lot of ways voters in Maine and Nebraska would be better off being winner-take-all states. That’s why there are so many winner-take-all states.
Four Basics
Before we get into the swing states here are four patterns you should know about.
In 2016 and 2020 Trump supporters were underrepresented by polling models. This meant the outcome of the election differed from projections by several percentage points.
In 2022 after the Dobbs Decision, which gave states the ability to outlaw abortion, pro-choice women were underrepresented by polling models. This meant the outcome of the election differed from projections by several percentage points.
Trump seems to have subcontracted his Get Out the Vote (GOTV) effort to a for-profit corporation. This will probably hurt his chances.
Trump bad-mouthed a lot of mainstream (often GOP) poll workers in the last go around, and no one knows how that will impact how they do their jobs.
One (Probably) Big Thing
One little-discussed pattern in presidential elections is that the voting machines have to be distributed by someone in each state. Usually the Secretary of State. In 2016 Republicans oversaw the elections in Michigan, Wisconsin, and Pennsylvania and Trump won all three of those states by a percentage point or two. In 2020 Democrats oversaw the elections in Michigan, Wisconsin, and Pennsylvania and Biden won all three of those states by a percentage point or two.
Coincidence?
Probably the difference is just making sure that your own people have been able to register, and your strongest districts have working voting machines and enough of them so your faithful constituents won’t have to wait in line too long. Anyway call me a crazy conspiracy theorist but I’m betting it makes a difference of a point or two!
So let’s run down the swing states:
Wisconsin
Wisconsin’s voting machines are handled by the Democratic Secretary of State, Wisconsin went for Biden last go around, and Harris seems to have a point or two edge, so…
WI definitely goes D!
Nevada
Nevada’s voting machines are handled by the Democratic Secretary of State, Nevada went for Biden last go around, and Harris seems to have a point or two advantage, so…
NV definitely goes D!
Georgia
Georgia’s a fun bit of skulduggery. Governor Brian Kemp first ran for the office in 2018. He happened to be the Secretary of State at the time and thus in charge of the voting rolls for his own election. Suffice it to say, Kemp made good use of his office:
Political scientists Michael Bernhard and Daniel O'Neill described Kemp's actions as the worst case of voter suppression in that election year… [Kemp] purged 1.4 million inactive voters from voter rolls during his tenure, including 668,000 in 2017.
By such means Kemp ended up winning a runoff by 54,723 votes.
It was a national scandal. Kemp’s replacement Secretary of State Brad Raffensperger defended him, and the national media soon went on to other things, but when it came time for the 2020 election, not wanting to get caught with their hands in the cookie jar a second time Kemp and Raffensperger ran the cleanest election in Georgia history and the state voted for Joe Biden. This caused Trump to throw a tantrum and tell Georgians it was fraud which seems to have demoralized Republican voters enough that two Democrats won runoffs to the Senate, which gave the Democratic Party a Senate majority. Then Donald Trump, now trying to overturn the election, told Kemp and Raffensperger in recorded conversations to find him some votes somewhere. But K & R couldn’t rig another election so soon after 2018, and Trump has had it in for them ever since.
After this the pro-Trump GOP in the state has been enraged and stripped Raffensperger of his control of voting machines or something like that. I don’t understand who is doing what down there. But the state is run by the Republicans and I can’t see them letting another get away from them. So…
GA definitely goes R!
Which gets us to:
Arizona
I don’t know anything about Arizona. Biden won and a Democrat controls the voting machines. But Trump is a couple points ahead, so let’s assume everything cancels out.
AZ probably goes R!
Michigan
This year Democrats control the voting machines in Michigan and the state went for Biden. But there’s a big Muslim vote in the state that carried Biden, and they aren’t too happy about Biden and Harris giving unqualified support to the Israeli genocide. You might claim that Trump is worse, but remember that enthusiasm and turnout are critical to elections. If the Muslim voters stay home…
MI maybe goes R!
North Carolina
North Carolina has a Democrat in charge of the ballots, but did not vote for Biden. North Carolina was also just hit by hurricane in the western part of the state. That might make them appreciative of Biden or angry at Biden. The area is traditionally Republican so is it harder for them to vote? Also didn’t the Republican legislature passed some more restrictive registration laws? But I may be wrong in that. So who knows?
NC maybe goes R!
Pennsylvania
A Republican controls the voting machines in Pennsylvania, but that Republican is an anti-Trump Republican. So if that voting machine discrepancy is the natural outcome of Secretaries of State just looking after their own, this should give Trump an advantage anyway. However, note above when the Republicans in Georgia decided to hold a clean election Biden won. Both Democratic and Republican parties at various times have engaged in all sorts of voting fraud and corruption, so what if the voting machine discrepancy is not just an unconscious bias and instead the post-Karl Rove GOP engaging in deliberate voter suppression? In that case PA might go to Harris.
One other issue in PA is that there are something like 17,000 mail-in ballots that were thrown out but whose mailers may, according to PA courts, still vote on election day. However, the GOP is planning to take that to the U.S. Supreme Court. If it gets there the Supreme Court will side with Trump and not let those voters vote.
Wouldn’t it be something if North Carolina and Michigan go for Harris and it all comes down to those re-do on the mail-in votes in Pennsylvania, and that goes to the Supreme Court as in 2000. And as in 2000 the SCOTUS will put the Republican in office.
PA maybe goes R!
The Sweep Factor
According to these musings Harris would have to win all three of the “maybe” states, or win two and flip Arizona. So Trump is likely to win.
But elections aren’t like sports betting. We know the factors in a game. We don’t know the factors in voting. Unlike games all we can say with certainty is:
What pollsters think—whatever it is—is probably wrong.
Americans, as I said here, have already decided who they want, but pollsters have no foolproof way to figure out what they’ve decided. We don’t know who will turn out to actually vote, and more than that we don’t know how to weigh our sampling. In 2016 and 2020 there were Trump voters who apparently wouldn’t admit their preference to pollsters (or who disproportionately wouldn’t participate in polling at all). Pollsters think they’ve corrected their models this go around. A similar phenomenon happened with pro-choice women in 2020. Pollsters think they’ve corrected their models this go around. Since in the last several elections significant voting blocs have been undercounted, it’s reasonable to suspect that blocs of voters are being undercounted now. Or that previously undercounts have been overcorrected for. Certainly predicting Michigan, Pennsylvania, and North Carolina are impossible. Pollsters don’t know how to weight the various samples. Probably Michigan and Pennsylvania will depend on the reality of GOTV, pro-choice women, and whether Trump voters are still not being mathematically sampled properly. In that light I would give all three states to Harris.
Republicans seem almost universally to have adopted the position that if they lose that means there was election fraud. That must be a comforting position to hold. If Harris loses I suspect a lot of Dems will say the same. They’ll lament sexism, racism, fascism, and the like. It’s true that there are many voters who are sexist, racist, and even fascist. But more than that there are voters who don’t care if other voters or even if candidates themselves are sexist, racist, or fascist. They’ll vote for someone because they want tax breaks, or because he’s funny, or because it’s funny to see him irritate other people, or because they expect some material benefit. If Dems aren’t willing to court some of those voters, every election becomes a cliffhanger, and Dems will never win by enough to have a mandate to govern.
The question isn’t whether Americans are willing to elect a black woman. The question is are voters in Michigan, for example, willing to elect a black woman who promises to give weapons to Israel. The question is why would any a significant majority of non-wealthy voters believe that any major politician in America today is really trying to serve their interests? The Democrats promise to fight for us and stand up for this and that, but they do not promise to do what we want and what’s in our best interest. The Republicans don’t either.
That’s why the election is so close.
Thanks for reading Blame Cannon! New posts every Wednesday, plus weekend bonus posts as life allows.
The gist came from the Achaean League.
As far as the 3/5 compromise, it reduced rather than in increased the power of slave states. Slave states did not get full count of their enslaved persons even though free states got full count for the children and women and others who could not vote. But in any case that does not impact elections today since it was cancelled by the Civil War Amendments.