Can Anyone Put Their Thumb on an American Election?

I saw the Supreme Court decision the night it came down and initially had a pretty simple reaction: Good. That's settled. The Trump administration had tried to put new requirements on mail voting into effect through the U.S. Postal Service before the November 2026 election. Federal judges stopped them.

And on September 14, the Supreme Court refused to let the administration put the new system into effect for the 2026 election. So that particular fight appeared to be over for the November 2026 election. But then I started thinking about the larger question. What exactly prevents a president — or either political party — from putting its thumb on the machinery surrounding an American election?

I'm not talking about somebody sneaking into a counting room and changing 20,000 votes. I'm talking about something much less cinematic. Who decides who stays on the voter rolls? Who determines what identification you need? Who decides where ballots can be dropped off? Who draws congressional districts?

Who decides which ballots arrived correctly and which ones don't count? Who watches the polling places? Who certifies the results? And what happens when lawyers challenge those decisions before or after Election Day? I realized I didn't really know. So I went looking.

And what I found is simultaneously reassuring and a little unsettling.

This distinction matters. The Supreme Court did not issue some sweeping ruling declaring that the federal government can never regulate election mail. The Court denied the government's request to lift a lower-court injunction and allow the Postal Service rule to operate during the 2026 election. The unsigned order was brief.

The Court said the government was unlikely to succeed on its challenge to the preliminary injunction and that the other factors required for emergency relief didn't favor letting the rule take effect. Justice Brett Kavanaugh wrote separately. And his explanation is important. Kavanaugh said there was at least a “fair prospect” that the Postal Service's final rule fell within its statutory authority.

But he concluded that applying it to the 2026 elections would be arbitrary and capricious because state and local election officials did not have “sufficient time” to reasonably implement it. Justices Samuel Alito and Clarence Thomas dissented. And it didn't necessarily close the door on every part of the policy forever.

It stopped the rule from being applied to the 2026 election. That's when this became more interesting to me. Because while everybody was watching the fight over mail ballots, there were several other election fights happening at the same time.

I think this was my biggest misunderstanding. We talk about “the election” as though somebody in Washington flips a giant switch. There isn't one. American elections are extraordinarily decentralized. States establish many of the rules. Counties, cities and other local election authorities administer much of the actual voting.

Congress has constitutional authority over important aspects of federal elections. Federal agencies enforce federal election and civil-rights laws. Courts resolve disputes about all of them. Political parties and campaigns participate, challenge rules, recruit observers and bring lawsuits — but they don't administer the election itself.

So if you're worried about somebody interfering with an election, asking: “Can the president control the election?” isn't really the right question. The better question is: Where are the pressure points?

And there are quite a few.

This one surprised me because the fight is already enormous. The Justice Department has been seeking detailed voter-registration information from states around the country. As of September 2026, the administration had sued 29 states and the District of Columbia seeking voter-registration records. The Justice Department says it needs voter information to enforce federal election laws, including laws requiring states to maintain accurate registration lists.

And there's an important point here that can get lost in the politics: Voter rolls should be accurate. People who aren't legally eligible to vote shouldn't be voting. People who have died shouldn't remain active voters indefinitely. Duplicate and outdated registrations should be cleaned up. But there's another side to that responsibility.

Eligible Americans shouldn't be removed from the rolls because a government database got something wrong. And that isn't merely theoretical. A whistleblower disclosure made public on September 14, 2026, alleged that a Department of Homeland Security effort searching for noncitizen voters was relying on data that could wrongly implicate U.S. citizens.

DHS disputes the criticism and says the initiative is a legitimate effort to identify unlawful voting. That gives us the real question. It isn't: Should voter rolls be accurate? Of course they should. The question is: How do we determine accuracy, how reliable is the information we're using, and what happens when the government gets it wrong?

Because removing somebody who shouldn't be registered is election maintenance. Removing somebody who should be registered is potentially taking away an American citizen's vote.

Then there are the rules themselves. Voter identification. Proof of citizenship. Mail-ballot requirements. Drop boxes. Registration deadlines. Early voting. Signature verification. Provisional ballots.

None of those things automatically equals voter suppression. Every election requires rules. And supporters of tighter requirements make a straightforward argument: Voting is extraordinarily important, so verifying that voters are eligible and ballots legitimate is reasonable. Critics make an equally important point: A security rule that prevents legitimate voters from voting isn't harmless simply because it was created in the name of security.

This may be the least visible battlefield. According to Axios, as of September 2026, Republicans were involved directly or indirectly in roughly 200 election-related legal battles around the country. Those cases involve things like voter identification, voter lists, ballot drop-off locations and access for election observers and challengers. Republican officials describe the effort as election integrity: enforce the laws, prevent illegal voting and make sure observers can watch the process.

Critics argue that some of those cases could unnecessarily restrict lawful voting. But here's something else worth knowing: Election litigation itself isn't unusual. Republicans sue. Democrats sue. Campaigns sue. Voting-rights organizations sue. States sue the federal government.

Sometimes those lawsuits are exactly what protects voters. In fact, the Postal Service case that started this entire discussion is an example of courts reviewing an election rule before it could take effect. So counting lawsuits doesn't tell us whether something improper is happening. We have to look at what each lawsuit is actually asking a court to do.

This one operates before anybody casts a ballot. You can conduct an election perfectly and still dramatically influence its partisan consequences by deciding where the district lines go. That's redistricting. Normally congressional maps are redrawn following the census. But by 2026, the country was experiencing an unusual wave of mid-decade redistricting.

Republican-controlled states had pursued new maps intended to improve Republican prospects. Democratic-controlled states had pursued or considered their own responses. And courts have become involved here too.

In September 2026, Missouri provided an unusually complicated example. The Missouri Supreme Court ruled that voters could decide whether to approve a new congressional map and ordered the state to continue using its 2022 map for the November 2026 election. A federal judge then intervened in the dispute, and two days later the U.S. Supreme Court stayed that federal court order while the litigation continued. The back-and-forth illustrates how redistricting disputes can move through both state and federal courts — and how the applicable map can remain contested surprisingly close to Election Day.

Whatever party is drawing the lines, the underlying problem is the same: At some point, politicians can begin choosing their voters instead of voters choosing their politicians.

That's why gerrymandering deserves scrutiny regardless of which party benefits.

Here's another distinction worth making. Election observers are normal. Federal election monitors can be normal. Political parties recruiting people to observe polling places is normal. Voter intimidation is not.

Preventing a lawful voter from casting a ballot is not. And using governmental authority to intimidate voters would raise profoundly different legal questions from legitimate election observation. In 2026, Democrats and voting-rights organizations were preparing for the possibility of federal law-enforcement officers appearing around polling locations. The administration defended its election activities as enforcement of federal law and protection against illegal voting.

At this point, the distinction matters enormously: Something people fear might happen isn't the same thing as something that has happened. So this is an area where I'm going to watch what actually occurs rather than assume the worst beforehand.

This may ultimately be the most important pressure point in a close election. Votes have to be counted. Questions arise about provisional ballots. Signatures. Deadlines. Eligibility. Recounts. Challenges.

Then somebody has to certify the results. And sometimes courts get involved. Again, none of that is inherently sinister. Election disputes have existed almost as long as elections. The important safeguard is that no single person is supposed to control all of those decisions.

Which brings us right back to what happened on September 14, 2026.

The administration wanted to make a substantial change to election mail. States challenged it. Federal judges reviewed it. They stopped it. The administration appealed. The Supreme Court considered the request.

And the Court declined to let the rule take effect for the 2026 election.

That's actually a pretty good demonstration of something we talk about constantly but rarely watch operating in real time: Checks and balances. The president has power. Executive agencies have power. States have power. Federal judges have power. The Supreme Court has power.

But none of them possesses all of it.

That doesn't prove the system is invulnerable. It demonstrates something more useful: There really are institutional brakes in the system, and sometimes they work.

Most election manipulation wouldn't require somebody sneaking into a counting room and manufacturing 20,000 ballots. It could be much more mundane. A voter database. A registration deadline. A district boundary. A rejected envelope. A closed drop box. A legal challenge. A certification dispute.

Each decision might affect relatively few people. But American elections can be extremely close. That's why I'm beginning to think: “Was the election rigged?” is usually the wrong question. By the time we're asking it, we're probably already too late.

The better questions happen beforehand. Who is changing the rules? What problem are they trying to solve? What evidence shows that problem actually exists? Who benefits from the proposed solution? How many legitimate voters could be affected? And maybe most importantly: Who has the authority to say no?

On September 14, 2026, one answer was the Supreme Court. Another day it might be a state court. A federal district judge. A secretary of state. A county election official. Congress. Or voters themselves.

And maybe that's the thing I understand better now. The American election system isn't protected because everyone involved can be trusted to behave perfectly. It is protected, imperfectly, by dividing power among people and institutions that can challenge one another. That makes the system complicated. Sometimes frustrating. Sometimes frightening.

But perhaps that's also the point. Because the more power is divided, the harder it becomes for any one person, institution or political party to control the whole thing. And the more we understand where that power actually resides, the harder it becomes for anyone to manipulate us about what's happening.

That may be the most important safeguard of all.

The question I’m left with is this:

How do we protect elections from manipulation without creating a system in which every losing side simply believes the election was manipulated?

Because those are two different dangers.

One is a system that can actually be abused.

The other is a country that no longer trusts the system at all.

And democracy probably cannot survive either one for very long.

So where is the line between legitimate safeguards and partisan advantage?

How much discretion should elected officials have over the rules governing their own elections?

And perhaps most importantly:

How do we build an election system that people trust even when their side loses?

“I don’t know.”