Are American Elections Secure? Exploring Voter Fraud and Mail Voting

Estimated reading time: 14 minutes

Introduction

This post explores the integrity of American elections, focusing on the distinction between secure voting machinery and the vulnerabilities of the electoral environment. As of September 2026, evidence suggests minimal fraudulent voting, but administrative disruptions and political delegitimization of legitimate votes pose significant risks. Trump’s Executive Order claims to enhance election integrity but could infringe on state authority. The Supreme Court’s recent decision did not affirm his regulatory power over mail voting. This research proposes examining the evolving dynamics of election security, legitimate voting procedures, and the implications of executive intervention. 

---

As of September 10, 2026, the evidence points toward a significant distinction: the machinery of American elections is generally quite secure, while the institutional and political environment surrounding elections may be becoming more vulnerable. That distinction is the key organizing idea for the post. 

How safe are American elections? 

On the narrow question—Are votes being fabricated, voting machines changing results, noncitizens voting in numbers capable of deciding elections, or mail ballots producing widespread fraud?—the available evidence is reassuring. 

Ever since Donald Trump ran for president in 2015, he has previewed his opinion of American elections. In my opinion, Trump has created an atmosphere of dishonesty and misinformation about American elections. His criticism of national elections has changed over the years, as shown in the chart below. If you read down the column of Trump election statements, you will see how his opinion has changed.

The U.S. Election Assistance Commission reports that in 2024 more than 98% of election jurisdictions used equipment in which voters either marked a paper ballot or the machine produced an auditable paper record. Nearly 93% of states, territories and D.C. required testing and certification of voting systems. Elections also use bipartisan handling procedures, chain-of-custody controls, pre-election testing and post-election audits. 

Cybersecurity & Infrastructure Security Agency (CISA) reported after the 2024 presidential election that it had “no evidence of any malicious activity that had a material impact” on election infrastructure. 

Mail voting deserves a little more nuance. Voting away from a supervised polling place does introduce vulnerabilities—in particular, possible ballot interception, coercion and improper ballot collection.

There have been genuine absentee-ballot fraud cases. But documented cases are rare, and the MIT Election Data + Science Lab says research provides no reason to conclude that fraud is more common in vote-by-mail states.

States also verify returned ballots through signatures, identification numbers and other procedures. A particularly revealing study of roughly 4.5 million Washington voters found only 14 ballots potentially cast suspiciously after the voter’s death—about 0.0003%—and even those cases could include matching errors rather than fraud. 

The empirical problem in American elections is not widespread fraudulent voting. The more important question may be whether legitimate votes can be prevented, discarded, administratively disrupted, or politically delegitimized. 

Disenfranchisement and Vote Suppression Before Trump

Long before Trump’s entry into presidential politics in 2015, the United States had an extensive history of disenfranchisement and vote suppression. Some methods prevented people from registering or casting a ballot; others allowed voting but reduced the political effect of those votes. The historical record is particularly extensive regarding Black voters after Reconstruction.

Here are seven major examples:

1. Poll taxes. Several states required citizens to pay a fee before voting.
2. Literacy and “understanding” tests. Prospective voters could be required to read or interpret constitutional passages.
3. Grandfather clauses. Some states exempted a person from restrictive registration requirements if he—or an ancestor—had been eligible to vote before the Civil War/Reconstruction era.
4. White primaries. In parts of the one-party South, political parties excluded Black citizens from primary elections. Because winning the Democratic primary often effectively determined the general-election winner, exclusion could make a person’s formal right to vote largely meaningless.
5. Intimidation, violence, and economic retaliation. Voting suppression wasn’t limited to statutes. Black citizens attempting to register or vote faced threats, job loss, economic reprisals and physical violence.
6. Registration restrictions and selective administration. Local officials could use registration deadlines, residency requirements, “good character” requirements, complicated paperwork, and discretionary decisions to keep otherwise eligible citizens off the voter rolls.
7. Gerrymandering and vote dilution. This differs from preventing someone from voting in person. Governments could redraw boundaries, use at-large elections, or alter election structures so that minority voters could cast ballots but have far less ability to elect representatives.

There’s an important distinction in the history being developed here: “preventing a vote” and “diluting a vote” aren’t identical phenomena. American history contains both. A useful historical progression would be:

Who may vote ? Who may register ? Who can reach/cast a ballot ? Whether the ballot is counted ? How much political power the counted ballot carries.

That framework gives us a useful historical baseline for comparing pre-2015 voting restrictions with more recent disputes over mail ballots and election administration—without assuming the mechanisms or motivations are equivalent. So…

That brings us to Trump. 

What did Trump actually order?

Trump’s March 31, 2026 Executive Order 14399, Ensuring Citizenship Verification and Integrity in Federal Elections, goes substantially beyond conventional cybersecurity measures.  President Trump signed Executive Order 14399 which would require DHS and SSA to send a list of U.S. citizens eligible to vote in federal elections to state elections officials. The EO also directs the U.S. Postal Service to start a NPRM regarding mail-in ballot criteria and handling.

To help us understand the sequence of legal actions since EO 14399 was issued, I created a table tracking the legal actions from March 31, 2026, to September 10, 2026.

Date	Legal Action
March 31, 2026	Executive Order 14399 issued. Trump directed USPS to develop new federal-election ballot-mail procedures, including requirements involving ballot-mail design, state voter lists, USPS enrollment, and transmission of ballots. This became the foundation for the subsequent mail-voting litigation. Brennan Center analysis of EO 14399
April 2	League of Women Voters v. Trump filed. The League of Women Voters and allied organizations sued in the District of Massachusetts challenging EO 14399 and actions contemplated under it. This became one of the two principal tracks of litigation.
April 3	California v. Trump filed. Twenty-three states and the District of Columbia brought a related challenge to EO 14399. The states argued, among other things, that the administration was attempting to exercise election-regulation authority that Congress had not delegated to the executive branch
June 25	District court rules against parts of EO 14399. Judge Indira Talwani concluded that important provisions of the order, including the USPS directive, exceeded executive authority and enjoined implementation as to the plaintiff states. This was the first major judicial barrier to Trump’s mail-ballot program.
August 24	Supreme Court stays the earlier injunction. The Supreme Court concluded that the challenge to the USPS portion was premature because USPS had not yet issued its final rule. Importantly, the Court did not hold Trump’s mail-voting policy constitutional. It expressly left open a challenge to the eventual final USPS rule and said its decision did not mean that the resulting measure would necessarily be lawful.
August 26	USPS issues its final rule. The Postal Service promulgated Ballot Mail for Federal Elections, creating requirements involving approved ballot-envelope designs, voter-specific Intelligent Mail barcodes, voter information uploaded through a Federal Ballot Mail Portal, and USPS verification of outbound ballot mail.
Late August	New challenges target the final USPS rule. The League plaintiffs and states returned to court, this time challenging an actual final agency rule rather than merely Trump’s executive-order directive. This addressed the justiciability problem identified by the Supreme Court.
August 28	District court temporarily blocks implementation. Judge Talwani imposed temporary emergency relief preventing USPS from enforcing the mandatory provisions while the court considered the new challenge.
September 3	Administration seeks Supreme Court intervention, No. 26A297. USPS asked the Supreme Court to stay the temporary order. Justice Ketanji Brown Jackson requested a response.
September 4	Judge Talwani issues a preliminary injunction. In a 49-page opinion, she blocked USPS from enforcing the mandatory portions of the final rule for the November 3, 2026 midterm election. She concluded that the plaintiffs were likely to succeed in showing that USPS lacked congressional authorization for the program.
September 6	Administration withdraws No. 26A297. Because the temporary order had been superseded by the September 4 preliminary injunction, USPS withdrew its original Supreme Court application.
September 6	USPS files new Supreme Court application, No. 26A305. The administration immediately asked the Supreme Court to stay the preliminary injunction itself, seeking authority to implement the final rule during the 2026 election. Justice Jackson ordered responses by September 9.
September 8-9	Major amici and challengers oppose the stay. The states, League plaintiffs, election officials, members of Congress, election-law scholars and other organizations submitted briefs. A whistleblower-related filing also raised operational concerns about the Federal Ballot Mail Portal and implementation of the USPS verification system.
September 10	First Circuit refuses to lift the injunction. The First Circuit issued orders in both League of Women Voters of Massachusetts v. Trump, No. 26-2029, and California v. USPS, No. 26-2031. That left Judge Talwani’s preliminary injunction in force while the litigation continued.
September 10	First Circuit ruling is reported to the Supreme Court. California and the League plaintiffs filed letters informing the justices of the appellate decision.
September 10	Administration files its Supreme Court reply. USPS submitted its reply in support of emergency relief in USPS v. California, No. 26A305, completing the principal emergency briefing.
September 14	On September 14, 2026, the Supreme Court of the United States denied the Trump administration’s emergency request in United States Postal Service v. California, No. 26A305 to stay Judge Indira Talwani’s preliminary injunction. As a result, USPS cannot enforce the challenged mandatory portions of its August 26 ballot-mail rule for the November 3 election

The EO, among other things, it directs DHS, working with the Social Security Administration, to create a State Citizenship List of adult citizens; directs the Justice Department to prioritize investigation and possible prosecution of officials who distribute federal ballots to ineligible people; and directs USPS toward a system involving approved mail-voter lists, special ballot envelopes and unique tracking barcodes. 

That matters constitutionally because American elections have traditionally been highly decentralized. Article I’s Elections Clause gives state legislatures initial responsibility for the “Times, Places and Manner” of congressional elections and gives Congress, not the president, authority to alter those regulations. 

Trump’s order therefore raises a question larger than mail ballots: 

Can the executive branch use control of DHS, DOJ, and the U.S. Postal Service to acquire regulatory leverage over elections administered by the states? 

What the Supreme Court just did—and did not do 

In Trump v. California, the Court lifted a lower-court injunction. But its reasoning was largely procedural. The majority characterized the executive order as instructions from the president to executive agencies rather than a direct command imposed upon the states. The Court did not decide that Trump’s proposed election system is constitutional. 

And the practical picture remains tangled. A separate injunction arising from litigation brought by the League of Women Voters and others continues to prevent important USPS provisions from taking effect. Litigation was still active on August 25. 

A unanimous three-judge panel of the United States Court of Appeals for the First Circuit denied the government’s motions to stay Judge Indira Talwani’s preliminary injunction, leaving the mandatory portions of the USPS mail-ballot rule blocked for the November 3 midterms. The court also denied USPS’s request for an administrative stay as moot. 

What appears to be Trump’s goal? 

We can’t establish someone’s private motive simply from outcomes. But Trump’s stated goals and observable actions let us identify several increasingly ambitious objectives. 

The narrowest interpretation—and the administration’s position—is election integrity: verify citizenship, improve ballot tracking, deter fraud and increase public confidence. The White House explicitly presents the order in those terms. 

Trump’s statements about elections began in 2015, and have extended to today. He’s very angry with the three justices that he appointed. The voted to against him in the final straw about mail voting.

But Trump’s own statements establish a considerably broader objective. On August 18, 2025, he said Republicans were going to do everything possible to “get rid of mail-in ballots”, explicitly connecting the issue to partisan competition and saying Democrats wanted mail ballots because, in his view, they helped Democrats win. And in his July 16, 2026 national address, he advocated eliminating ordinary mail voting while retaining exceptions for illness, disability, military service and travel. 

That gives us stronger evidence than speculation about motive. 

We should investigate four nested goals: 

1. Restrict ordinary mail voting. This is an explicitly stated Trump objective, not an inference. 

1. Move election eligibility verification toward federal executive control. The citizenship database, DOJ enforcement, and USPS requirements establish federal executive machinery that could potentially affect who receives a ballot. 
2. Change the allocation of authority over elections. The deeper constitutional struggle may concern whether a president can exercise powers over election administration that historically belong primarily to the states and Congress. 
3. Shape the legitimacy environment surrounding elections. Trump’s repeated claims that mail ballots are “inherently corrupt,” despite the lack of evidence for widespread mail-ballot fraud, create a political framework in which particular categories of ballots can be regarded as suspect before they are counted. FactCheck.org’s review of his July election-security address found that he offered no evidence demonstrating widespread fraud. 

The fourth possibility deserves particular attention because of what happened in 2020. 

A potentially more revealing research hypothesis: 

Is the Trump administration constructing a federal apparatus capable of determining which voting procedures—and ultimately which ballots—it regards as legitimate? 

That is empirically researchable. 

Here is the chain:

claims of widespread fraud > delegitimization of mail voting > citizenship database > federal mail-ballot eligibility lists > USPS control over ballot transmission > DOJ investigations of election officials > litigation over presidential power > Supreme Court decisions defining how quickly states can challenge executive intervention. Figure 1 is a Trump election statements archive, 2016-2026. It’s valuable to go down the “verbatim excerpt” column, and note how his reasons to question election outcomes have changed from millions of people voting illegally in 2016, to mail in ballots will lead to massive electoral fraud. These allegations have been proven wrong. And in 2026, Trump and associates can’t get away from voter fraud (its really non extent), and mail voting is sure to be fraudulent (it isn’t).

That is a much more consequential story than simply “Trump opposes mail voting.” 

And there is another fascinating complication: mail voting isn’t inherently Democratic. About 30% of ballots nationwide were cast by mail in 2024, and Republicans use it too. Trump’s campaign against mail voting therefore isn’t simply reducible to a mechanically predictable partisan advantage. The politics depends upon which voters use mail ballots, in which states, under which rules, and which ballots encounter new eligibility or administrative barriers. 

How Safe is an American Election?

There are three different election safety levels:

• Ballot security (fraud, machines, mail voting, audits);
• Administrative security (whether eligible voters can vote and their ballots are counted);
• Constitutional security (whether officials accept the results and whether the rules can be altered by executive power immediately before an election). 

The first category looks remarkably strong. The second is where the 2026 mail-ballot controversy belongs. And the third may ultimately be the most important. 

That produces a compelling inversion of the familiar election-integrity debate: 

Perhaps the central threat to an American election is no longer someone secretly inserting fraudulent ballots into the system. It is the possibility that government power can change the rules governing legitimate ballots, disqualify voters or ballots through administrative mechanisms, and persuade millions of Americans beforehand that votes cast by the opposing side are illegitimate. 

That is a hypothesis, not yet a conclusion. But I think it is the hypothesis worth testing. 

I can next build a research dossier from 2020–2026, including Trump’s statements about voting, the March 2026 executive order, the lower-court decisions, the August 24 Supreme Court opinion, empirical studies of mail-vote fraud, and the constitutional history of presidential power over elections. That would give us the evidentiary foundation for an article or book chapter. 

Because this situation is changing rapidly before the November 3 midterms, I can also monitor the litigation and USPS implementation for consequential developments 

Questions

• Do you think Trump will try other methods to effect the November election?
• What methods will be use?
• Will the FBI be involved, such as being at polling stations?
• Will he call a national emergency and then cancel the election?
• Will he use ICE-type tactics using violence at polling sites?