How Americans Vote: From Paper Ballots to Voting Apps

Last updated 6 days ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.

The method by which Americans cast their votes has never been static. From public declarations in a town square to the tap of a touchscreen, the technologies of democracy have consistently evolved alongside broader advancements in society.

This change over time reflects a persistent tension between several considerations: the security of the ballot, the accessibility of the vote for all eligible citizens, public confidence in the outcome, and the cost of administering elections.

There’s now a powerful movement back toward paper-based systems to ensure auditability while simultaneously exploring the convenience and profound perils of a digital-first future, including internet and mobile app voting.

In This Article

• Voting evolution in the U.S.: From public voice votes → paper ballots → mechanical lever machines → punch cards → electronic voting machines (DREs) → ballot‐marking devices (BMDs) and scanners → emerging internet/mobile options.
• Paper ballots remain central: Most jurisdictions now use paper-based systems, often combined with electronic scanning, to ensure auditability.
• Security vs accessibility: Trade-offs exist between machine efficiency, voter accessibility (e.g., BMDs for disabilities), and security (risk of tampering, software errors).
• Help America Vote Act (HAVA, 2002): Spurred modernization after the 2000 election, encouraging DREs and voter-verifiable paper trails.
• Ballot-marking devices (BMDs): Increase accessibility but rely on voter verification; research shows some voters don’t always confirm printed votes.
• Internet/mobile voting: Pilot programs exist but cybersecurity experts caution against large-scale deployment due to privacy and auditability risks.
• Mail-in voting: Widely used and heavily regulated; fraud is extremely rare, but operational challenges like signature verification exist.
• Audit and verification: Paper trails and post-election audits are critical for election integrity.
• Variation by jurisdiction: U.S. elections are decentralized; technology adoption, auditing, and accessibility vary by state and county.

So What?

• Paper + audits = gold standard: Paper ballots combined with audits provide the most reliable balance of security, transparency, and voter confidence.
• Accessibility matters: BMDs and other assistive technologies are essential for voters with disabilities, but proper voter education and verification remain critical.
• Internet voting is high-risk: While convenient, large-scale internet voting is not yet considered safe or auditable, highlighting the need for continued research and cautious adoption.
• Decentralization creates variation: Voter experience, system security, and audit effectiveness differ widely across states and counties, emphasizing the need for localized best practices.
• Ongoing modernization needed: Election technology continues to evolve, requiring updates to security protocols, training, and funding to maintain integrity and public trust.

The Foundation: A History of Paper and Levers

The debates surrounding voting technology aren’t modern phenomena but are deeply rooted in a long American story of reform, crisis, and adaptation. The very concept of a private, secure ballot is the result of centuries of technological and political evolution.

From Public Voice to Private Paper

For the first 50 years of the United States, voting was a public spectacle. Eligible voters, who, at the time, were exclusively white men, would gather at the local courthouse and cast their vote aloud, a practice known as viva voce.

The shift toward privacy began in the early 19th century with the emergence of the first paper ballots. Initially, these were nothing more than scraps of paper on which a voter would write his preferred candidates’ names.

By the mid-19th century, political parties had co-opted this process, printing and distributing their own ballots, known as “tickets,” which listed only their slate of candidates. These pre-printed party tickets simplified straight-party voting but introduced new problems.

The distinct color and size of each party’s ticket often made a voter’s choice public, undermining secrecy and opening the door to intimidation and frequent accusations of voter fraud.

The Australian Ballot Revolution

Widespread concerns over fraud and coercion led to major reform: the adoption of the “Australian Ballot.” Pioneered in Australia in 1858 and first implemented in the United States by New York and Massachusetts in 1888, this system introduced two revolutionary concepts.

First, ballots were printed at government expense. Second, they listed all candidates for office, regardless of party. These standardized ballots were handed to voters at the polling place, marking the true birth of the secret ballot in America.

For decades, these official paper ballots were counted entirely by hand, a slow and labor-intensive process that was the norm until the age of automation.

The Age of Automation

The 20th century brought machines into the polling place, promising speed, efficiency, and enhanced integrity.

Mechanical Lever Machines

First used in the 1890s, these complex contraptions allowed voters to pull a lever next to their chosen candidate’s name. A large master lever would close a privacy curtain and, upon completion, increment a series of internal mechanical counters when pulled again.

Their “sheer physicality” inspired immense public confidence, and for much of the century, they were the dominant voting method east of the Mississippi River.

Punch Card Systems

Developed in the 1960s with the rise of the computer age, punch card systems allowed votes to be tabulated by machines for the first time, a major innovation for producing rapid election night results.

Voters used a stylus to punch out a small, pre-scored piece of paper—a “chad”—to indicate their choice. These systems became dominant in the western United States.

This history reveals a recurring cycle: a problem is identified (such as lack of privacy), a technological solution is adopted (the Australian ballot), which then reveals new problems (slow, error-prone hand counts), leading to another technological solution (automation).

This pattern of innovation, unforeseen consequences, and subsequent reform continues to define the landscape of election administration.

The 2000 Election Crisis and Federal Response

The era of levers and punch cards came to an abrupt and controversial end with the 2000 presidential election. The recount in Florida, with its infamous “hanging chads” and ambiguous voter intent, exposed the deep unreliability of these aging systems and triggered a national crisis of confidence.

This crisis spurred a rare moment of federal intervention in the typically state- and locally-run world of elections. Congress responded by passing the Help America Vote Act of 2002 (HAVA).

This landmark legislation banned the use of lever machines and punch cards in federal elections and, crucially, provided billions of dollars in federal funding for states to purchase new, modern voting equipment.

HAVA also established the U.S. Election Assistance Commission to provide guidance and set voluntary standards for these new systems. In doing so, HAVA fundamentally reshaped the technology of American elections, pushing the nation into the electronic age and setting the stage for the next generation of debates over security and verifiability.

Modern Voting Technology: Electronic Systems and Security Concerns

The post-HAVA era ushered in a new generation of voting technology. As states shed their old mechanical systems, they adopted a variety of electronic solutions, each with its own benefits and critical trade-offs.

The central debate that emerged revolves around a single, crucial question: can the electronic results be trusted and, if necessary, independently verified?

Direct-Recording Electronic Machines

In the wake of HAVA, many jurisdictions rushed to purchase Direct-Recording Electronic voting machines. These are fully computerized systems where voters make selections on a touchscreen or with buttons, and the vote is recorded directly into the machine’s electronic memory.

While seen as a modern solution, a decade of intense scrutiny by computer scientists revealed significant flaws.

Security Vulnerabilities Discovered

Security experts quickly discovered that DREs were plagued with vulnerabilities that could compromise an election’s integrity.

Software Errors: Like any complex computer, DREs run on software that can have bugs. In the context of an election, these errors can have catastrophic consequences. In a 2004 election in Carteret County, North Carolina, a DRE system lost over 4,400 electronic ballots that were never recovered. That same year in Broward County, Florida, 134 electronic ballots were found to be blank in a race decided by just 12 votes.

Vote Flipping: A widely reported issue with touchscreen DREs is “vote flipping,” where a voter selects one candidate, but the machine records a vote for another. This is often caused by aging hardware and miscalibrated touchscreens, and it has fueled viral videos and conspiracy theories about machines “stealing votes.”

Hacking and Tampering: Independent studies by researchers at institutions like Johns Hopkins and Rice University uncovered hundreds of security flaws in the software of major DRE manufacturers, concluding they were vulnerable to tampering by either a malicious insider or an external hacker. At one DEF CON hacking conference, an expert easily manipulated a paperless DRE to flip every vote cast on it.

The Paperless Problem

The most fundamental problem with early DREs was that they were paperless. This created what security experts consider an unacceptable risk: the impossibility of a meaningful audit.

If the electronic vote record is altered by a software bug or a malicious attack, there is no independent, physical record to check it against. A printout of the machine’s totals is not an audit; it is merely a reflection of the potentially compromised electronic data.

The VVPAT Solution

The solution to this problem is the Voter-Verified Paper Audit Trail (VVPAT). A VVPAT is a physical paper printout of the voter’s selections that is displayed to the voter for verification before the ballot is officially cast.

This paper record is then stored securely in the machine and can be used for a manual recount or post-election audit if needed. In many states, the VVPAT is legally considered the official ballot of record for recounts.

The value of VVPATs was demonstrated powerfully in the 2020 Georgia presidential election. Amid intense scrutiny and claims of fraud, election officials were able to conduct multiple hand recounts of the paper VVPATs, each of which confirmed the accuracy of the original electronic tally, ultimately increasing confidence in the result.

This shift from trusting the machine to verifying the machine represents a profound philosophical change in American election administration. By the 2024 election, over 98% of U.S. jurisdictions reported using voting systems that produce an auditable paper record.

The Resurgence of Paper: Optical Scanners

The most common type of paper-based system in use today is the optical scanner. With this method, voters mark a paper ballot by hand—for example, by filling in an oval—and then feed it into a scanner that tabulates the vote.

These systems, first introduced in the 1960s and inspired by standardized test-grading machines, are considered by many security experts to be the gold standard for verifiable elections.

Their key advantage is the unambiguous paper record created by the voters themselves. This allows for robust post-election audits, such as Risk-Limiting Audits, where a statistically significant sample of paper ballots is counted by hand to provide strong evidence that the machine-tabulated outcome is correct.

Ballot Marking Devices: A Hybrid Approach

Ballot Marking Devices are a hybrid technology designed to bridge the gap between electronic interfaces and paper records. Developed primarily to meet HAVA’s mandate that every polling place have an accessible voting option, BMDs allow voters to use a touchscreen, audio interface, or other assistive technology to make their selections.

The device then prints a physical paper ballot with those choices, which the voter can review before formally casting it (typically by inserting it into a separate scanner).

While essential for accessibility, BMDs have become the subject of a new security debate. As jurisdictions replace their old DREs, some have opted to use BMDs for all in-person voters, not just those who need assistance, for the sake of uniformity.

Critics argue this is a step backward in security compared to hand-marked paper ballots. They point out that a software bug or hack could cause the BMD to print a ballot that does not match the voter’s selections on the screen.

Since research shows many voters do not carefully verify the paper printout, such an error could go undetected. Furthermore, some BMDs encode votes in QR codes or barcodes that are unreadable to the voter, creating another point of potential failure that cannot be verified by the person casting the ballot.

Comparing Modern Voting Systems

To clarify these complex trade-offs, here’s how the key features of modern in-person voting systems compare:

Mail-In Voting: From Civil War Innovation to Pandemic Necessity

The option to vote by mail, also known as absentee voting, has transformed from a niche provision for soldiers and travelers into a mainstream method used by tens of millions of Americans. Its dramatic expansion, particularly during the COVID-19 pandemic, has placed it at the center of intense political and security debates.

Historical Development

The practice of voting by mail dates back to the U.S. Civil War, when it was established to allow Union soldiers to cast ballots from the front lines. For over a century, it remained a limited option, primarily for military and overseas citizens or those with a valid excuse, such as illness or disability, who could not vote in person.

The 2020 election marked a turning point. In response to the pandemic, states across the country expanded access to mail-in voting to ensure people could vote safely. This resulted in an unprecedented surge, making it a primary voting method for a large portion of the electorate.

While the use of mail ballots has since receded from its 2020 peak, it has become a firmly established feature of the American electoral system. According to the 2024 Election Administration and Voting Survey, about 30% of all ballots in the 2024 general election were cast by mail.

How Mail-In Voting Works

For a voter, the process is straightforward. It typically begins with an application, which can often be completed online or by mail. Once the application is verified, the local election office sends the voter a ballot packet.

This packet contains the ballot itself, a secrecy envelope or sleeve to protect the voter’s privacy, and a pre-addressed return envelope that includes a crucial section for the voter’s declaration and signature.

After marking the ballot, the voter places it in the secrecy envelope, seals that inside the return envelope, signs the declaration, and sends it back. Voters generally have multiple return options, including the U.S. Postal Service or secure, official ballot drop boxes, the use of which has grown significantly in recent years.

Multi-Layered Security Measures

Contrary to claims that mail-in voting is rife with fraud, the system is protected by a multi-layered series of security measures that make widespread tampering logistically difficult and highly detectable.

Signature Verification: This is a cornerstone of mail ballot security. Upon receiving a returned ballot, election officials meticulously compare the signature on the outer envelope with the signature on file in the voter’s registration record. If the signatures do not match, the ballot is set aside, and the voter is typically notified and given an opportunity to “cure” the discrepancy.

Secure Chain of Custody: From the moment they arrive at the election office, mail ballots are subject to a strict chain of custody. They are stored in secure locations, and all handling and processing are done according to documented procedures, often under the observation of representatives from political parties and the public.

Ballot Tracking: To increase transparency and voter confidence, many states have implemented online ballot tracking systems. These allow a voter to see when their ballot was mailed, when it was received by the election office, and whether it was accepted for counting.

Tamper-Evident Envelopes and Barcodes: Most states use specially designed return envelopes that make it obvious if they have been opened or tampered with. Many also use unique barcodes on ballot materials to track each ballot individually and prevent a voter from casting more than one.

Numerous studies, post-election audits, and the consensus of election officials have repeatedly shown that instances of mail ballot fraud are extremely rare. The 2020 election, which saw the highest levels of mail voting in U.S. history, was declared “the most secure in American history” by a coalition of federal and state election officials.

Accessibility Benefits

Beyond the security debate, mail-in voting is a critical tool for accessibility. It removes significant barriers for many voters, including those with disabilities, the elderly, rural voters, and those with inflexible work schedules.

Data consistently shows that voters with disabilities use mail-in voting at higher rates than the general population, and states that have made mail voting easier have seen significant increases in turnout among citizens with disabilities.

This underscores that “election security” is multifaceted; it includes not only preventing fraud but also ensuring the system is resilient and accessible enough for every eligible citizen to participate.

The Internet Voting Frontier: Promise and Peril

The idea of casting a ballot from a computer or smartphone holds undeniable appeal. It promises the ultimate in convenience, potentially boosting turnout by lowering the act of voting to a few simple clicks.

However, this vision clashes with stark reality: the overwhelming consensus among cybersecurity experts is that secure, private, and verifiable internet voting is not currently possible with today’s technology.

The Appeal of Online Voting

Proponents of internet voting argue it could dramatically increase participation by overcoming nearly every logistical barrier, from transportation issues to taking time off work. It is often framed as a vital tool for enfranchising hard-to-reach populations, such as military personnel serving overseas, citizens living abroad, and voters with severe disabilities who cannot easily mark a paper ballot or travel to a polling place.

The story of MacCene Grimmett, a 106-year-old homebound voter in Utah who was able to vote independently for the first time in years using a mobile app, powerfully illustrates this potential benefit.

Acknowledging these challenges, several states currently permit electronic ballot return—via email, fax, or an online portal—for these specific groups of voters.

Overwhelming Security Concerns

Despite the appeal, the technical and security hurdles are immense. Federal agencies, including the Department of Homeland Security and CISA, along with the National Academies of Sciences, have repeatedly warned that online voting systems introduce “significant security risks.”

The problems are not minor bugs but fundamental flaws in the concept itself. The core issue is a paradox between secrecy and verifiability. Other secure online transactions, like banking, work because they are fully traceable. The bank knows it’s you, and you get a detailed receipt of the transaction.

Elections demand the opposite: the system must verify your identity to issue you a ballot, but the cast ballot itself must be completely anonymous to prevent vote-selling and coercion. No current technology can reliably guarantee both of these contradictory requirements.

This fundamental conflict gives rise to severe risks:

Client-Side Malware: A voter’s own personal computer or smartphone can be infected with malicious software that could undetectably change their vote before it is even transmitted.

Server-Side Attacks: Election servers themselves are prime targets for hackers. A successful breach could allow an attacker to “observe, alter, and add votes” at will, with no way to discover the manipulation.

Denial-of-Service Attacks: An attacker could flood a voting system with traffic, causing it to crash or slow to a crawl, thereby preventing legitimate voters from casting their ballots.

Privacy Breaches: Online voting apps and platforms often handle immense amounts of sensitive personal data, including driver’s license information and ballot selections, which could be exposed to third-party vendors or stolen by hackers.

Lack of Meaningful Audit Trail: This is perhaps the most critical flaw. An electronic ballot transmitted over the internet cannot be meaningfully audited. Even if an election official prints a copy of the received ballot, the voter never saw or verified that specific piece of paper. It cannot serve as a trustworthy record of voter intent, making a true recount impossible.

Pilot Programs Reveal Persistent Problems

Over two decades of pilot programs in the U.S. have consistently highlighted these risks rather than solving them. Early Department of Defense pilots in 2000 and 2010 for overseas voters were deemed successful as small-scale proofs-of-concept but came with stark warnings about the insurmountable security risks and logistical challenges of scaling up.

More recent pilots have been even more alarming. In 2018, West Virginia used a mobile voting app called Voatz for some overseas and disabled voters. A subsequent security analysis by MIT researchers concluded the app was “not secure,” identifying vulnerabilities that could allow hackers to alter votes and even determine how a person voted simply by observing their network traffic.

This case became a prominent cautionary tale about the gap between a vendor’s security claims and the reality discovered by independent experts. This dynamic, where public demand for convenience creates a market for vendors to sell technology that has not been proven secure, is one of the greatest dangers in the push for online voting, especially in the absence of federal certification standards for such systems.

The Hidden Infrastructure: Funding, Accessibility, and Federal Oversight

The voting machines and methods that capture public attention are only the most visible parts of a vast and complex ecosystem. Underlying every ballot cast are critical issues of funding, accessibility for all citizens, and the role of government in setting standards and protecting the entire system from threats.

The Cost of Democracy

Administering elections is expensive, and the procurement of voting technology is a particularly complex and often opaque process. The market is highly concentrated, with three major vendors—Election Systems & Software, Dominion Voting Systems, and Hart InterCivic—controlling nearly 90% of the voting equipment used in the U.S.

A comprehensive 2021 analysis by Verified Voting, “The Price of Voting,” revealed that while the upfront unit price of a machine might be somewhat standard (around $5,750 for a common optical scanner), the total cost of ownership is driven by two other factors: arbitrary discounts and steep, long-term fees.

Vendors often provide idiosyncratic discounts to win contracts, but the real cost comes from mandatory annual fees for software licensing, maintenance, and support. These fees can amount to 40% or, in some cases, over 100% of the original purchase price over the system’s 10-year lifespan.

This “razors-and-blades” business model creates significant “vendor lock-in,” where a county, having made a massive initial investment, becomes dependent on a single company for parts and service, stifling competition and innovation.

The fact that election officials sometimes have to resort to finding replacement parts on eBay for decertified machines starkly illustrates this problem.

Regardless of the technology used, securing elections requires sustained investment. Many of the nation’s voting machines are aging and operating past their expected lifespans. Election officials have stressed that one-time federal funding appropriations are insufficient; they need consistent, predictable funding streams to properly maintain existing systems and plan for future upgrades.

Accessibility as a Civil Right

Ensuring that every eligible citizen can vote privately and independently is a core tenet of American democracy, enshrined in federal laws like HAVA, the Americans with Disabilities Act, and the National Voter Registration Act.

Since the passage of HAVA, significant progress has been made. The turnout gap between voters with and without disabilities has narrowed, and polling places have become more physically accessible.

However, a fundamental gap remains between the legal right to accessible voting and the on-the-ground reality for many. A 2022 survey by the EAC and Rutgers University found that voters with disabilities still reported experiencing difficulties at a rate more than three times higher than voters without disabilities (14% vs. 4%).

Persistent Barriers

The challenges are not primarily a lack of technology but failures in implementation, funding, and training. Voters report persistent barriers, including:

Physical Obstacles: Lack of accessible parking, steep ramps, crowded interiors, and long lines that are difficult for people with mobility impairments or chronic pain.

Ballot Marking: Difficulty reading or marking paper ballots for those with vision or dexterity impairments.

Invisible Disabilities: Crowded or disorganized polling places can cause overwhelming anxiety and confusion for people with cognitive or mental health disabilities.

Lack of Training and Awareness: Poll workers are often inadequately trained on how to operate accessible equipment or how to interact with voters with different types of disabilities. Many voters themselves are unaware of the accommodations available to them.

Closing this accessibility gap requires more than just purchasing the right equipment. It demands a commitment to better poll worker training, more thoughtful polling place design, robust mail-in and electronic ballot delivery options, and effective outreach to inform voters of their rights and options.

Federal Oversight in a Decentralized System

While elections are administered at the state and local level, the federal government plays a crucial role in providing support, setting standards, and ensuring security.

The U.S. Election Assistance Commission: Established by HAVA, the EAC is an independent, bipartisan commission that serves as a national clearinghouse for election administration information. Its two most critical functions are:

• Voluntary Voting System Guidelines: The VVSG is a set of federal standards for the security, functionality, and accessibility of voting systems. While their adoption is voluntary under federal law, most states require that their voting equipment be tested and certified to meet all or part of these federal benchmarks, making them a de facto national standard. The latest iteration, VVSG 2.0, incorporates modern cybersecurity principles like two-factor authentication, cryptographic protection, and support for post-election audits.
• Election Administration and Voting Survey: The EAVS is the most comprehensive source of data on U.S. elections. Conducted after every federal election, this survey collects state-by-state data on turnout, registration, voting methods, and technology use, providing the vital statistics that inform our understanding of how America votes.

The Cybersecurity and Infrastructure Security Agency: Following Russian interference in the 2016 election, U.S. election infrastructure was designated as “critical infrastructure.” CISA, part of the Department of Homeland Security, now works collaboratively with state and local election officials to protect this infrastructure.

CISA provides a wide range of no-cost services, including risk and vulnerability assessments, cybersecurity training, and information sharing to help officials defend against evolving threats.

Looking Forward: Balancing Innovation with Security

The evolution of American voting technology reflects ongoing tensions between competing values: accessibility, security, cost, and public confidence. Each technological advance has solved some problems while creating new ones.

As we look to the future, several trends are clear. The movement toward paper-based systems with electronic tabulation represents a consensus that auditability is essential for public confidence. The expansion of mail-in voting has permanently changed the electoral landscape, providing crucial accessibility benefits while requiring robust security protocols.

Internet and mobile voting remain tantalizing possibilities, but current technology cannot overcome fundamental security and privacy challenges. Until these can be resolved, such systems will likely remain limited to specific populations where the accessibility benefits outweigh the security risks.

The infrastructure supporting all these technologies—from vendor markets to poll worker training to federal oversight—requires sustained attention and investment. Elections are only as secure and accessible as their weakest components.

Our articles make government information more accessible. Please consult a qualified professional for financial, legal, or health advice specific to your circumstances.