Table of Contents
References & Edit History Related Topics

Voting Machines

Are Electronic Voting Machines the Best Method for Voting? 
print Print
Please select which sections you would like to print:
verifiedCite
While every effort has been made to follow citation style rules, there may be some discrepancies. Please refer to the appropriate style manual or other sources if you have any questions.
Select Citation Style
Feedback
Corrections? Updates? Omissions? Let us know if you have suggestions to improve this article (requires login).
Thank you for your feedback

Our editors will review what you’ve submitted and determine whether to revise the article.

External Websites

The earliest forms of voting were fairly simple but public, so everyone knew your preferred candidate. In ancient Greece, a small ball or token was deposited in a ballot box for a particular candidate. The use of balls was carried forward into voting in secret societies, where a white ball was selected to vote someone into the organization and a black ball to keep the individual out (which is the origin of the phrase “to blackball” someone). The earliest voting machines also used small balls, including a machine proposed by the Chartists in 1839. In another method, voters might choose a colorful “ticket” from the party they wished to vote for and drop it in a clear jar. In each instance, voting was a public act.

Similar methods of voting continue in some areas of the world. In Gambia, for example, citizens vote by dropping marbles in vividly colored drums marked with the photo and logo of the party of their choice; a chime is heard when the marble is dropped, indicating that a vote has been cast. These audible and visual cues allow illiterate constituents to vote with ease and confidence. [3]

(This article first appeared on ProCon.org and was last updated on Dec. 15, 2022.)

Non-Mechanical Voting: Paper Ballots and Viva Voce

While we may think of voting as a fairly orderly event in which we cast a private vote at our local polling place or drop an absentee ballot in a mailbox, voting prior to 1856 was a sometimes chaotic and, again, public event. Men might publicly declare their vote viva voce (by voice) to the election officials one-by-one either from a line or from among the crowd of voters and spectators. Voting viva voce dates to ancient Greece, in which “an early form of applaudometer” was used wherein the citizenry shouted or otherwise made noise for their chosen candidate. A group of people in a closed room would decide which candidate had received the loudest approval. While Aristotle deemed the method “childish,” the method was refined to each voter simply voicing their vote aloud. Versions of both the early “applaudometer” and more restrained viva voce vote are still used today from TV show contests to the UK Parliament and U.S. Congress[2][4]

Paper ballots on which voters indicate the candidate of their choice were used as early as 139 BCE in Rome. Signed scraps of paper were often used as ballots in the United States prior to the 20th century, raising questions of uniformity, privacy, and fraud. (Separate ballots were often cast for separate candidates meaning one man would have to vote multiple times with multiple scraps of paper, multiplying the possibility of lost ballots and fraud). Moreover, public forms of voting were frequently subject to vote-buying and other influences from candidates and parties at the polling place.[1]

Non-Mechanical Voting: Secret Paper Ballots

Private paper ballot voting was not popularized until the 1850s when Victoria and South Australia, Australia, adopted what would later be called the “Australian Secret Ballot.” These ballots, printed by the government, listed each candidate’s name in a fixed order and were counted by hand. This type of ballot was adopted by the United Kingdom in 1874 and the United States in 1888 (after accusations of fraud during the 1884 U.S. presidential election). [1][5]

Mechanical Voting: Lever Machines

Votes were mostly cast by Australian Secret Ballot until the advent of the locked mechanical lever machine, patented by Jacob H. Meyers (patent US415549A) in Nov. 1889, that discouraged tampering. Men were offered privacy while voting (thus continuing the secrecy of ballots), while the machines kept a running tally of votes, allowing for faster reporting of results. The lever machine was first used in 1892 in Lockport, New York. Meyers stated the machine was built to “protect mechanically the voter from rascaldom, and make the process of casting the ballot perfectly plain, simple and secret.” Alfred J. Gillespie would patent another version of the mechanical lever machine on July 11, 1899 (patent US628905A).[1][6][7]7]

The machines were widely praised. In 1898, the Brooklyn Eagle boasted that  “where other cities were hours and even days in counting their votes, Rochester [New York] knew the complete result in the city on every office—State, County, Assembly, Senate and Congressional—in just thirty-seven minutes. There was not a mistake, not a hitch.” The voting booths, which cost $550 each in 1898 and were ”built like bank vaults [and] weigh about as much,” were not manufactured after 1982 but were still used by voters into the 21st century. [1][7][8]

Mechanical Voting: Punch Card Machines

While lever machines continued to be used, other technologies were developed that would later be used in elections. Punch cards were invented by Herman Hollerith and first used by the Baltimore Board of Health and then for the 1890 U.S. census. In 1958, IBM created the Port-A-Punch, which adapted the punch card to include pre-scored holes that aided the accuracy and completion of the punch.[1][8][9][10]

Punch cards were not used for voting until the 1960s, when University of California at Berkeley professors Joseph P. Harris and William Rouverol improved upon IBM’s technology to adapt the punch card to machine voting. They patented their invention, called  the Votomatic, in 1966 (patent US3240409A). The machine sold for $185 in 1965, was highly portable (weighing only six pounds), and even allowed for a mechanical vote recount, supposedly eliminating the subjectivity of a human recount.[1][8][9][10]

However, the 2000 U.S. presidential election put punch card machines in an unflattering spotlight due to the controversy over “chads.”. A “chad” is the circle of paper punched from the ballot when a vote is cast. A “hanging” chad is when the paper circle is still partially attached to the ballot. A “dimpled” or “pregnant” chad results when an indentation can be seen where the chad should have been punched but remains completely attached. The punch card machine could not read the vote if a chad were not completely punched out, leaving humans to ponder and decide voter intent. These punch-card problems had existed since the 1960s but seemed inconsequential until the presidential election in 2000, when hanging and pregnant chads and how to interpret them became so critical to the vote count in Florida that the election was ultimately decided by the U.S. Supreme Court (in Bush v. Gore), with George W. Bush declared the winner.[1][11]

The controversy had consequences. The Help America Vote Act of 2002 (HAVA) effectively banned the use of pre-scored punch cards for American elections after 2006, though the machines were not completely phased out in the United States until their last use in the 2014 general election in two counties in Idaho.[1][9][10][12][13][14]

Electronic Voting: Mark-Sense, Optical, and Digital Scanners

Mark-sense scanners were familiar to many school children via standardized tests. The student filled in a circle with a #2 pencil to mark multiple-choice answers before the teacher fed the answer sheet through a mark-sense machine that scored the test by detecting the graphite in markings. Mark-sense scanners were first developed by IBM in 1937. They were first used in an election in 1962 in Kern City, California, and the machines stayed in use for voting for about a decade. [1]

E. F. Lindquist, a professor at the University of Iowa, improved upon the mark-sense machine with his creation of an optical scanner that could detect simple marks on the ballots rather than the specific graphite, allowing a variety of marking tools to be used. Optical scan machines were introduced to elections in 1964, in San Diego, California. The machines are still widely used today. [1]

Digital scanners, considered the next generation of scanner technology, work similarly to home office scanners. These machines take a digital image of each ballot to count votes and store the images in a database in case a recount is needed. [13][14]

Many jurisdictions with older voting machines switched to optical or digital scanners in response to HAVA, which required that outdated equipment be replaced. [12]

Electronic Voting: Direct-Recording Electronic Machines (DRE

Direct-recording electronic (DRE) machines date to 1850 when Albert Henderson patented an electrochemical legislative roll call recorder (patent US2177525A) that allowed legislators to vote using “aye” and “nay” telegraph keys at their desks. Thomas Edison improved on the machine (patent US90646A) in 1869, adding electromechanical counters to tally the votes, and in 1898 Frank S. Wood invented a push-button paperless version of the machine for elections (patent US616174A).[1]

Contemporary direct-recording electronic (DRE) machines were invented in 1974 with the advent of the Video Voter (patent US3793505A). This machine was first used in elections in 1975 in Streamwood and Woodstock, Illinois. DRE machines with touch screens soon followed, though the core idea remains the same: DRE machines record votes directly to the machine’s memory, without a paper ballot, after the voter uses a touch screen, wheel, or other device to mark their votes. Some DRE machines are equipped with an add-on: the Voter-Verified Paper Audit Trail (VVPAT), which prints a paper record of votes that the voter may review before casting the ballot. [1][13][14]

In response to HAVA, most jurisdictions replaced old lever and punch card machines with DRE machines.[1][9][10][12][13][14]

Electronic Voting:  Ballot Marking Devices

Ballot marking devices (BMD) are used in jurisdictions with older DRE machines and in concert with scanners so people with disabilities can vote independently, a requirement set forth by HAVA. The devices, frequently a tablet, allow voters to mark a virtual ballot by selecting choices on a screen or tablet. The machine then prints the marked ballot, which can then either be fed into an optical scanner for electronic counting or hand-counted.[1][9][10][12][13][14]

BMD’s help those with disabilities in several ways. First, the ballot can be seen on the device or listened to via headphones. Second, there are four ways to mark the ballot: via touch screen, a keypad with Braille, a sip and puff device (for people with limited or no motor capability), and rocker paddles (for people with limited or no motor capability). Third, the voter can adjust the ballot for better visibility with zoom in and out and high contrast functionality, as well as better audio with repeat and speed controls.[15]

Online Voting

Internet-based voting platforms have been widely debated. Of particular concern is the protection of ballots from domestic and international hacking and from malware on personal devices. After all, even current voting machines are susceptible to hackers. Nevertheless, some jurisdictions have gone forward with Internet voting.[16]

Estonia, for example, has used Internet voting nationally since 2005 for an early election period. Bulgaria not only uses Internet voting domestically but has also installed e-voting machines in international embassies for their citizens living abroad. [17][18]

Seven US states allow ballot return via an online portal for UOCAVA (Uniformed and Overseas Citizens Absentee Voting Act) voters: Alabama, Arizona, Colorado, Massachusetts, North Carolina, North Dakota, and West Virginia. Four of those states also allow voters with disabilities to vote via an online portal: Colorado, Massachusetts, North Carolina, and West Virginia, the last state allowing certain first responders to vote online as well. [13][16]

In light of the security risks that accompany online voting, are electronic voting machines still the best technology for voting in the 21st century? Or should we return to the time-honored, hand-counted, paper-marked ballot?

For more on the history of voting technology, see ProCon’s Historical Timeline: Electronic Voting Machines and Related Voting Technology.

PROSCONS
Pro 1: Current electronic voting machines with voter-verified paper audit trails are secure voting methods. Read More.Con 1: Current voting machines are outdated, and their software is vulnerable to domestic and foreign attacks. Read More.
Pro 2: History proves the reliability of electronic voting machines. Read More.Con 2: Current voting machines are mechanically obsolete and extraordinarily expensive to repair or replace. Read More.
Pro 3: Using the Internet for elections is an incredible risk the country does not need to take. Read More.Con 3: Voting machines should be replaced with Internet voting. Read More.

Pro Arguments

 (Go to Con Arguments)

Pro 1: Current electronic voting machines with voter-verified paper audit trails are secure voting methods.

Paper trails, whether from a direct-recording electronic (DRE) machine with Voter-Verified Paper Audit Trail (VVPAT) or an optical scan machine, provide a safe method of voting and protect against hacking by facilitating a manual recount.[19]

Douglas Jones, retired University of Iowa computer science professor, noted that he has never seen evidence of election machine hacking outside of controlled tests. [19]

Senator Ron Wyden (D-OR) explained the value of such tests: “White hat hackers [“a hacker who tests computer systems for possible vulnerabilities so that they can be fixed”] do an invaluable public service in this technologic age by identifying security holes and, if necessary, shaming the government or the companies responsible into fixing them. The success of the Voting Village — in which public demonstrations of voting machine flaws by hackers at Defcon quickly convinced officials in Virginia to promptly move to paper-based voting systems — is a prime example of how the computer security community has positively impacted public policy and protected our national security.” [20][21]

Even in the 2016 U.S. presidential election during which many blamed Russian agents for infiltrating the election, only Illinois’ voter registration system was breached despite 21 states receiving notifications from the Department of Homeland Security that their election systems were targeted by hackers. Attacks in the 20 other states were unsuccessful, proving the security of those machines and the policies governing them. [22]

Vulnerabilities in certain types of voting machines were suspected of allowing people to manipulate votes in the 2020 U.S. election. However, the machines were investigated by the Cybersecurity and Infrastructure Security Agency (a department of Homeland Security) and found to be safe. As Brandon Wales, Executive Director of CISA, stated, “We have no evidence that these vulnerabilities have been exploited and no evidence that they have affected any election results. Of note, states’ standard election security procedures would detect exploitation of these vulnerabilities and in many cases would prevent attempts entirely. This makes it very unlikely that these vulnerabilities could affect an election.” [23]

Further, every system has vulnerabilities that can be exploited. However, optical scan and direct-recording electronic (DRE) machines have paper trails and legislative protections to avoid such manipulations. According to Gabriel Sterling, a top aide to the Georgia Secretary of State, “reports show what reasonable people already know — if bad actors are given full and unfettered access to any system, they can manipulate that system. That is why procedural, operational, and legal election integrity measures are crucial.”[23]

Pro 2: History proves the reliability of electronic voting machines.

Optical scan and direct-recording electronic machines have been used for some 50 years with no evidence of significant errors or voter fraud during an election. The machines rely on paper ballots to supplement the technology, allowing for a good marriage of speed and the ability to verify votes. [1]

“The point of using paper is to have a chain of evidence you can use to test the correctness of the count. The point of using scanners is to mechanize the count so you avoid as many clerical errors as possible,” explained Douglas Jones, retired University of Iowa computer science professor. [19]

For example, the 2020 US presidential election results were widely challenged. In Georgia, where ballot-marking devices were used, officials were able to hand-count the paper ballots, confirming now-President Biden as the winner of the state. If other technology without a paper trail were used, the recount could have been nearly impossible and much more contentious. [19]

In 2020, fewer than 9% of American voters lived in jurisdictions that did not provide paper trails for votes cast.[19]

Additionally, the machines are equipped with or allow for add-on ballot-marking devices that create accessible, independent voting for people with disabilities and for whom English is not their first language. [13][14]

Where electronic voting machines are concerned, there is no need to reinvent the wheel. With continued maintenance, software updates, and evolving legislative security measures, the current electronic voting machine system will continue to serve voters well.

Pro 3: Using the Internet for elections is an incredible risk the country does not need to take.

While online banking and other sensitive tasks are secure online, elections are entirely different because votes must be kept anonymous. “All the same mechanisms that you might use to detect that something went wrong are all the mechanisms that we’re not allowed to give you because of the requirement for anonymity,” explains Dan Wallach, a professor of computer science at Rice University. [24]

Democracy Live’s OmniBallot is an Internet ballot-marking platform. A review found the platform “vulnerable to vote manipulation by malware on the voter’s device and by insiders or other attackers who can compromise Democracy Live, Amazon, Google, or Cloudflare.” Further, the company did not have a privacy policy and collected voters’ identity, ballot selections, and browser fingerprints, which are unique hardware and software configurations including everything from which fonts are installed on a device to if cookies are enabled.[25][26][27]

In 2018, West Virginia allowed some voters to use an Internet voter platform called Voatz. A review of the platform by MIT computer scientists concluded, “Our analysis has shown that this application is not secure. A passive network adversary can discover a user’s vote, and an active one can disrupt transmission in response. An attacker that controls a user’s device also controls their vote, easily brushing aside the app’s built-in countermeasures. And our analysis of the protocol shows that one who controls the server likely has full power to observe, alter, and add votes as they please.” [28]

Not only are the voting platforms insecure and vulnerable to attack, but the devices used to vote are as well. Internet voting requires individuals to “bring your own device.” With millions of individual computers, tablets, and phones come innumerable security risks as each device may carry malware, have been hacked, or be otherwise compromised. An attack on a single device could endanger the entire election system. [29]

Internet voting is a “nonstarter” according to Aviel Rubin, Technical Director of the Johns Hopkins University Information Security Institute. “Internet voting requires trusting the software developers who create the system, trusting the network providers, trusting the network protocols, and trusting the end systems in people’s homes. Over the last decade, we have seen security incidents on the Internet increase exponentially. As attackers grow in sophistication, the systems they are exploiting are growing in complexity. It is the perfect storm for mayhem.” [30]

Con Arguments

 (Go to Pro Arguments)

Con 1: Current voting machines are outdated, and their software is vulnerable to domestic and foreign attacks.

According to a Brennan Center for Justice March 1, 2022 report, “Like any computerized system, voting machines age into obsolescence. For electronic voting machines purchased since 2000, experts agree that the expected lifespan for the core components is between 10 and 20 years. For most systems, however, it is probably closer to 10 than 20…. Today, 24 states, home to over 41 million registered voters, use machines first fielded more than a decade ago as their principal voting equipment. While jurisdictions in many states have replaced older voting machines in the last eight years, the need to replace equipment as it ages continues.” [31]

Voters’ confidence in the security, accuracy, and fairness of voting machines is low. A 2018 national survey found about 80% of Americans believed the current voting system may be “vulnerable to hackers.”[32]

That feeling is supported by experts. “[F]rom a security perspective, old software is riskier, because new methods of attack are constantly being developed, and older software is likely to be vulnerable,” explained Jeremy Epstein, Lead Program Office of the Secure and Trustworthy Cyberspace (SaTC) program for the National Science Foundation. [31]

Experts have long warned that computers are “inherently vulnerable” to hacking, making the use of digital technology in elections especially dangerous. Efforts to secure digital elections systems are often piecemeal and not uniformly implemented, making a patchwork of security efforts that create multiple vulnerabilities that are easily taken advantage of by domestic and foreign forces. [33]

“These vulnerabilities, for the most part, are not ones that could be easily exploited by someone who walks in off the street, but they are things that we should worry could be exploited by sophisticated attackers, such as hostile nation states, or by election insiders, and they would carry very serious consequences,” according J. Alex Halderman, Professor of Computer Science and Engineering and Director of the Center for Computer Security and Society at the University of Michigan. [33]

The vulnerabilities could allow for a malicious virus to spread throughout a jurisdiction’s voting machines, which could be done by a hacker with physical access, via the internet, by USB fobs used by election workers, or by the cards used by technicians who service the machines. The virus could disable the system or allow a hacker to take control of the machine(s). [33]

Halderman said “attackers could then mark ballots inconsistently with voters’ intent, alter recorded votes or even identify voters’ secret ballots.” [33]

Con 2: Current voting machines are mechanically obsolete and extraordinarily expensive to repair or replace.

As of March 1, 2022, voting machines used by 23 states are no longer manufactured—they are especially vulnerable to breakdowns, errors, and fraud. These problems not only create long waits at polling places but can also shake the public’s confidence in elections in general. Those states are home to 21 million registered voters. [31]

Because the machines are no longer manufactured, the concern becomes whether parts can be found to repair them. As Douglas Jones, retired University of Iowa computer science professor, explains, “The trouble is, the laptop computer marketplace from which many of the components for these machines came is predicated on the idea that people will throw away their old laptops fairly frequently. What you end up doing is cannibalizing parts from working equipment to keep other equipment working.”[19]

Further, the older voting machines do not have to meet the same accessibility and security regulations as newer machines. This obsolescence also creates voting problems for people with disabilities and whose first language is not English among the approximately 40 million voters who live in the 26 states and two territories in which assistive voting equipment, such as ballot marking devices (BMDs), are no longer manufactured. [19][31]

The Elections Infrastructure Initiative estimated in Dec. 2021 that replacing “antiquated” voting machines would cost $1.8 billion. That cost does not include the costs of updating and bolstering administration and operation, registration systems costs, audit systems, or cybersecurity, which would cost an additional $51.5 billion.[34]

Why, then, spend over 53.3 billion to patch a failing system when we could create a new system without aging, vulnerable machines?

Con 3: Voting machines should be replaced with Internet voting.

We conduct most of our lives online. From banking to medical records, some of our most sensitive, private information is stored on and accessed via the Internet. This fact was reinforced during the COVID-19 pandemic, with the dramatic rise of online jobs, services, and classes. We have crossed the Rubicon in terms of using the Internet for social and personal needs and tasks.

In 2022, 91.8% of Americans had Internet access, a percentage that is expected to grow to 95% by 2027. 99% of Americans over 18 years old used the Internet in 2021, accounting for nearly all voting-aged Americans. [35][36]

Moreover, Internet usage is largely an equal-opportunity tool, cutting fairly across racial demographics: 95% of Hispanic people, 93% of white people, and 91% of Black people used the Internet in 2021. The rates are similarly equal and high when urbanity is considered: 95% of urban city dwellers, 94% of suburban residents, and 90% of rural inhabitants use the Internet. [37][38]

Internet use only drops below 90% when income is considered: 87% of people who make less than $30,000 a year use the Internet.[39]

Because the online world is already accessible to the vast majority of voters, Internet voting could increase voter turnout and overcome obstacles such as transportation to polling places or taking time off work to vote. [40] Voting online is also as secure as existing voting machine software, Internet banking, and medical records. The Internet already boasts incredibly accurate tracking mechanisms including click and open rates as well as privacy protocols. [40]

In fact, online voting is already happening in U.S. elections. Seven U.S. states allow ballot return via an online portal for UOCAVA (Uniformed and Overseas Citizens Absentee Voting Act) voters and four of those states also allow voters with disabilities. Moreover, almost three million Americans who are eligible to vote live abroad and 35 million people living in the U.S. are disabled. Online voting would facilitate their participation in democracy. “Consider MacCene Grimmett, who is, at 106, Utah’s oldest voter. When she was born in 1913, women did not have the right to vote. Homebound since she broke her ankle two years ago and unable to hold a pen steadily, she was able to cast her ballot… [in 2019] thanks to an app on a mobile device. The technology empowered her, helping her execute — independently, anonymously, securely and with dignity — her most basic duty as a citizen,” said election officials Amelia Powers-Gardner and Chris Walker.[13][16][41]