States are flocking to buy the new “universal use” touchscreen ballot markers, which have all the disadvantages of existing touchscreen voting machines. As an added “bonus,” the new touchscreens print unverifiable barcodes that are then counted as our votes!

By Jennifer Cohn, May 13, 2018

About fifteen years ago, in the wake of the 2000 “hanging chad” debacle, many states bought touchscreen voting machines, using the billions of dollars allocated to states under the Help America Vote Act of 2002. Quite a few of these machines are still in use today.

Touchscreen voting machines — with or without so-called Voter Verifiable Paper Audit Trails (“VVPATs”) — have been a disaster for election integrity because voters cannot know if their vote as recorded inside the machine — where the actual counting occurs — matches their intention or even the wording on the VVPAT. (1) Thus, when election officials and others describe these paper printouts as “voter verifiable,” they lull the country into a false sense of security because the VVPAT itself isn’t actually counted as your vote.

Unlike hand marked paper ballots, VVPATs are counted only if included in a manual audit or manual recount. Given the dismal status of state recount and audit laws, this means that most VVPATs (unlike hand marked paper ballots) are never counted at all. (2)

Even when VVPATs are used in hand recounts or manual audits, there is no way to know which VVPATs, if any, were actually reviewed for accuracy by the voters. Depending on the length of the lines and whether election officials provide instructions, voters may not review them at all. (3) Studies suggest that most voters won’t notice discrepancies even if they undertake such review (which is unlikely) (3,4), and that they won’t start over even if they find a problem (5).

It has also proven difficult to catch errors when reviewing VVPATs in a post-election recount or audit. A study conducted by CalTech/MIT found that “[o]ut of 108 elections that contained errors, … no errors were reported in the VVPAT audit.” (Id.)

Meanwhile, jurisdictions that use touchscreen voting machines generally have longer lines than those that count hand marked paper ballots on optical scanners. (6) This is especially true during peak voting hours because many people can hand mark their ballots at the same time, whereas with touchscreen voting machines, you are limited by the number of machines distributed to the polling place.(7) “Only one optical scanner is required in each polling place to serve the same number of voters as ten to twelve electronic voting machines.” (8) Optical scanners are also less expensive than touchscreens, which means localities can buy more machines. (9)

Long lines are not merely inconvenient. They can also disenfranchise voters who are unable to wait due to health issues, old age, or work and family commitments.

On a related point, if equipment actually breaks down — due to a denial-of-service attack or unintentional “glitch” — the potential for voter disenfranchisement is much greater with touchscreen voting machines than with hand marked paper ballots counted on optical scanners (where electronic failures stop only the counting, not the actual voting). In 2008, voters in Horry County, South Carolina were forced to “vote” on “scraps of paper” when “human error” caused the touchscreen voting machines to malfunction in 80% of the county’s precincts in 2008. (10) “State Election Commission spokesperson Chris Whitmire was widely quoted as telling people to vote on ‘paper towels’ if necessary.” (11) In 2016, “Improperly coded memory cards” led 3/4 of all the machines in Washington County, Utah to break down. Poll sites offered backup paper ballots — “until some of them ran out and told voters to come back later.” (12)

Fortunately, most touchscreen voting machines have finally reached the end of their lives, and Congress has recently allocated more than $300 million to help states “upgrade.” (13) Unfortunately, rather than upgrade to hand marked paper ballots, many localities are embracing yet another supposedly “verifiable” but not really verifiable touchscreen device: touchscreen ballot markers.

Touchscreen ballot markers have been used for years to assist voters with disabilities. But in the past few years, vendors have marketed them for universal use. The two most popular “universal” touchscreen ballot markers are the ES&S ExpressVote Universal Voting System and the Dominion ImageCast Democracy Suite.

These touchscreen ballot markers generate something that some vendors, election officials, and the media misleadingly call a “paper ballot.”

What they don’t mention is that the “paper ballot” includes both text and a barcode, and the barcode (which humans can’t read) is the only part of the ballot counted as your vote. This specific concern was highlighted by Computer Science Professor Duncan Buell (University of South Carolina) in a Voting Technologies Task Force report submitted to the South Carolina League of Women Voters:

“The new ES&S voting machine, the ExpressVote, has major problems, beginning with the fact that the voter cannot verify the ballot information that will be counted by examining the ballot… The voter may think that s/he is seeing a list of names that will be counted, but it is the barcode, not the list, that is read by the scanner that counts the vote.” (14)

As explained by a recent panel of election security security experts, this is problematic because barcodes present an opportunity for hackers:

“[B]arcodes on ballots…could give hackers a chance to rewrite results in ways that could not be traceable…” (15)

Thus, when the Georgia legislature recently considered a bill that would have allowed the state to replace its paperless touchscreens with these touchscreen ballot markers, Computer Science Professor Richard A. DeMillo (Georgia Institute of Technology) spoke out against it:

Despite the recent fascination with electronic ballot markers that print bar codes for scanning ballots, bar codes have no place in Georgia’s election system. They introduce a whole new class of vulnerabilities. (16)

As Professor DeMillo explained to CBS46 news in Atlanta, “The difficulty with that [barcodes on ballots] is that you and I can’t read barcodes.” (Id.) If a hacker got to the barcode, DeMillo says they could manipulate the counting: “For example, telling the barcode reader to flip votes on demand or at a certain time.” (Id.)

But even without the barcodes, as explained by Verified Voting, ballot marking systems “require programming, servicing and software licensing fees. They are also susceptible to breakdown and hacking.” (17)

Moreover, similar to touchscreen voting systems, touchscreen ballot marking systems cost more than twice as much as optical scan systems. (18) Thus, for the same economic, logistical, and common sense reasons discussed above as to touchscreen voting machines, touchscreen ballot marking systems are more likely than optical scan systems to cause long lines and potentially disenfranchise voters. After Maryland tried the ES&S ExpressVote in 2016, “all but one county opposed widespread use” because they took voters longer to use than hand marked paper ballots:

“[V]oters took far longer to vote using the [ExpressVote] ballot marking device than to mark a paper ballot by hand. This caused lines of people waiting to use the ballot marking device…” — Rebecca Wilson, Chief Election Judge, Prince George’s County Precinct 17–01, Maryland (19)

In recent testimony, Verified Voting agreed that touchscreen ballot markers would likely cause longer lines than hand marked paper ballots counted on optical scanners. (20)

Making matters worse, the particular optical scanner that comes with the ES&S ExpressVote — the DS200 — was recently discovered to include cellular modems. (21) According to Computer Science Professor Andrew Appel (Princeton), the cellular modems in the DS200 scanners make it easy for a “man-in-the-middle” hacker to alter votes. (22)

And that’s not all. The New York Times recently published an explosive piece on ES&S, which revealed that the vendor has sold systems with remote access software. (23) “Voting machine vendor ES&S offered a remote access option in 2006 and in 2011, according to The New York Times.” (24) In 2017, ES&S signed a ten year contract with the state of Michigan, which also referenced a “remote access” option. (25) When confronted by reporters, however, ES&S claimed that the remote access option in the Michigan contract pertained only to “print on demand” devices used to print blank ballots. (26)

A Florida Department of State report shows that Dominion has also sold systems with “remote access” software. (27) When asked to explain, Dominion said it “does not remote into any Florida customer site,” and that “this is not a method by which we provide customer service.” (28) As of 2009, its ballot markers apparently came “equipped with a convenient slotted hole that allows anyone to stuff ballots directly into the locked ballot box.” (29) They also reportedly came “equipped with USB ports” that could “facilitate network, internet and wireless access.” (30)

Alarmingly, the ES&S and Dominion touchscreen ballot markers are starting to spread throughout the country like a virus. Quick searches on Google and Twitter reveal that counties in Wisconsin, Texas, Louisiana, Kentucky, West Virginia, Tennessee, New Jersey, Indiana, and Arkansas have already purchased the ExpressVote. And Colorado uses the ImageCast for election-day voting. Georgia Secretary of State Brian Kemp is also enamored of the ExpressVote, and Pennsylvania appears poised to choose it as the state’s new voting system in the next few weeks. That is, unless we can stop this train wreck ourselves.

Unfortunately, the Election Assistance Commission (“EAC”) has not, and most likely will not, lift a finger to discourage the spread of these touchscreen ballot markers. On the contrary, it has already certified both the ExpressVote and the ImageCast systems. (31) This is not surprising. In 2006, the National Institute of Standards and Technology recommended that the EAC not certify further purchases of paperless touchscreen voting machines, but the recommendation was rebuffed. (32)

Meanwhile, the recently appointed EAC chairman, Thomas Hicks, has lulled the public and perhaps some election officials into a false sense of security by spreading the myth that voting machines can’t be hacked because they don’t connect to the internet. (33) In reality, computer science experts agree that all voting machines can be attacked through the internet (and otherwise) — even if they do not directly connect to it and lack remote access software, cellular modems and USB drives — because (among other reasons) they all receive programming before each election from centralized computers that often do connect to the internet. (34) We must therefore strive to minimize our use of electronic election equipment, not double up on it with the addition of unnecessary touchscreen ballot markers.

You can check whether your own county and state use the ES&S ExpressVote or Dominion ImageCast systems with the tool linked here, although I’m not sure how often this tool is updated. Thus, it would be best to contact your state and county election boards directly to see if they use or plan to use these systems. Don’t panic if your county has already bought the ExpressVote or ImageCast. The question is whether they plan to use these touchscreen ballot markers for all voters (which would be irresponsible) or limit their use to voters who are unable to hand mark their ballots (appropriate).

When discussing election equipment with state and local election officials, your message should be clear. Please tell them:

  • No to touchscreen voting machines
  • No to touchscreen ballot markers (except for voters who are unable to hand mark )
  • No to barcodes on ballots
  • Yes to hand marked paper ballots.

Please convey this same message to your Members of Congress as well.

It might also help to give them examples of jurisdictions that are getting it right. Minnesota, Massachusetts, North Dakota, New Mexico, and New Hampshire use hand marked paper ballots throughout the state (although with the exception of New Mexico, they lack laws requiring meaningful post-election audits, a problem plaguing almost every state in the nation). (35) So does Virginia, which recently decertified touchscreen voting machines in favor of hand marked paper ballots counted on optical scanners. (36) Missouri lawmakers have introduced legislation — currently pending in the state senate — requiring hand marked paper ballots counted on optical scanners as well. (37)

Thus, there is hope. We must not allow profit-motivated vendors and complicit election officials to snuff it out with half-truths and outright lies about the supposed “verifiability,” “security,” and non-existent “cost saving” features of touchscreen ballot markers and barcodes on ballots.



I forgot to mention that vendors and their minions will sometimes try to fool people with the claim that hand marked paper ballots must be “verified” too. But in reality, “voter verification” refers to the need for voters to review the paper printout from a touchscreen voting machine or touchscreen ballot marker to confirm that the touchscreen hasn’t flipped their selections. This is not an issue with hand marked paper ballots because they are software independent!

Vendors will also sometimes to try to fool people with the claim that hand marked paper ballots have a higher error rate than paper printouts from touchscreens. But there are no studies to support this. And explained by Professor Stark (UC Berkeley) who invented post-election Risk Limiting Audits, “[t]he percentage of truly ambiguously hand-marked ballots is microscopic, as statewide recounts in Minnesota have shown. With rare exceptions, voter intent is usually quite clear to humans reading ballots. Examples:;


Here is another recently reported problem with the ExpressVote. If you don’t hit the “More” button, you apparently won’t see all the candidates!

In addition, Sedgwick County, Kansas recently reported that ExpressVote touchscreen barcode ballot markers have a “design flaw” that caused problems in the special election to fill Mike Pompeo’s seat! [referencing “design flaw” with the County’s new touchscreens]; [“Sedgwick County, Kansas … is seeking a firm or firms to provide one or more Electronic Poll Books,” which “must be compatible with the ExpressVote Voting machines that the county will be using for elections beginning in 2017.”]


This 10-minute video by election integrity advocate Lulu Friesdat — who is also an Emmy award winning documentarian — includes footage showing that voters typically do not review VVPATs. It also includes a discussion with Computer Science Professor Richard DeMillo (Georgia Tech) about the security concerns regarding barcodes on ballots from touchscreen ballot markers like the ExpressVote and ImageCast. The video is a “must watch” for anyone interested in election security.

On a related point, even without the barcodes, the “summary cards” from touchscreen ballot markers like the ES&S ExpressVote and Dominion ImageCast are extremely problematic. Looking at the “summary card” mock-up below, without a cheat sheet, would you notice if a few of your selections were deleted? Also, note how there is no reference to party affiliation. Again, without a cheat sheet, would you expect most voters to remember the last names of who they selected for judge, clerk, Tax Commissioner, etc.? Would you remember the content of each and every Constitutional Amendment or Proposition as cursorily described in the summary card?

Remarkably, these “summary cards” have not been subjected to human usability studies! In fact, the EAC advisory board recently voted down a proposed resolution by UC Berkeley Professor of Statistics Philip Stark to withhold certification unless and until such studies are conducted. (38)

One of the “no” votes came from Jim Dickson whose organization has accepted donations from Diebold, which was acquired by ES&S in 2009. (39) (When Diebold dissolved in 2010, its assets were divided between ES&S and Dominion Voting.)

Another important “no” vote came from Linda LaMone, Maryland’s state elections administrator. Ms. LaMone is known for having presided over the state’s purchase of paperless Diebold voting machines in 2002 — after testers were “able to remotely upload, download, and execute files with full system administrator privileges. Results could be modified at will, including changing votes from precincts.” (40) When the Maryland Election Board voted in 2006 to switch to paper ballots, LaMone used the funds to instead buy Diebold e-polling books. (41) The e-polling books malfunctioned, but LaMone nonetheless allowed herself to appear in Diebold promotional materials, for which she was later censured by the governor. (Id.)

If this sounds familiar, you may be thinking of Georgia’s former Secretary of State Cathy Cox who similarly appeared in Diebold promotional materials after signing a $54 million contract with the company in 2002.

Cox’s former election director, Kathy Rogers, likewise raised eyebrows when she moved to Diebold after presiding over Georgia’s deployment of Diebold’s paperless machines and defeating a paper trail bill. (42) Cox moved to ES&S when it acquired Diebold in 2009. Her online bio states that she was responsible for implementing Georgia’s paperless voting machines in 2002.

This tenacious trio — Cox, Rogers, and LaMone — has consistently defended touchscreen voting machines, disregarding the grave concerns raised again and again by independent IT experts:

Maryland’s Linda Lamone, [and] Georgia’s Kathy Rogers and Cathy Cox … were among the state election officials who consistently blasted computer scientists for our criticism of their beloved touch screens. These people behaved as if they were the vendors whose products were being attacked, when in fact they were customers who had been sold inadequate products. I could easily imagine what motivated them. The DRE voting machines unquestionably made elections much easier to administer. They conveyed an element of progress as well. Officials who brought in these machines could feel proud about keeping pace with the “state of the art.” (43)

Alarmingly, Rogers is now encouraging counties throughout the United States to replace their aging touchscreen voting machines with the touchscreen barcode ballot markers from ES&S, i.e., the ExpressVote. Here is a link to a video of her promoting the ExpressVote in Maryland.

Maryland fell for the pitch. Despite grave concerns expressed by Maryland’s Department of Information Technology, the state deployed the ExpressVote for universal use during early voting in 2016. (45) ES&S declared Maryland’s use of the ExpressVote to be a huge success. (46)

But the Maryland Elections Board was so unimpressed that it decided to limit the use of the ExpressVote — originally leased by the state for universal use— to voters who are unable to hand mark their ballots. (47) In fact, “all but one [Maryland] county opposed widespread use [of the ExpressVote machines]” because (among other reasons) they took voters longer to use than hand marked ballots! (48)

Although many individual counties throughout the U.S. have fallen for the Expressvote, Maryland’s decision to ditch the ExpressVote means that no state is currently committed to the ExpressVote for universal statewide use. Rogers hopes to make Georgia the first — just as Georgia was the first to deploy touchscreen voting machines statewide in 2002. Earlier this year, she persuaded state legislators to include language in a voting system bill that would have allowed touchscreen barcode systems like the ExpressVote. The bill sailed through the Georgia senate and failed in the state House only after on-the-ground activists caught wind of the problematic language and protested. Rogers has nonetheless won over Georgia’s current Secretary of State, Brian Kemp, who has already test piloted the system. It is therefore up to concerned Georgia voters to sound the alarm and demand that the newly established Georgia Voting Commission reject touchscreen barcode balot markers once and for all.


Background: Jennifer Cohn is an attorney and election integrity advocate in the San Francisco Bay Area who graduated from UCLA and Hastings College of the Law. As an attorney, her areas of practice included insurance coverage and appellate law. She practiced law for more than twenty years, including seven years as a partner with Nielsen Haley & Abbott, LLP in Marin County, California. Since 2016, she has devoted her professional efforts full time toward investigating our insecure election system and potential solutions. She can be contacted through her Twitter account, @jennycohn1.

End Notes


2 [“‘Most states never look at the paper,’ Bernhard said in their CCC presentation. ‘You have a great way to defend against an attack, but you never use it.’”] [compilation of sources re: the difficulty of obtaining meaningful hand audits and hand recounts in the U.S.]

3. [Las Vegas survey found that “fewer than 40 percent of voters actually checked the paper record of their vote before leaving the polling place.”] [“In addition, a study done in Nevada showed that only 31% of voters actually compared the audit trail to the screen upon which they voted.”]

4., pp. 2–3 [a Rice University study of voting machine “review screens” showed that “over 60% of voters do not notice if their votes as shown on the review screen are different than how they were selected. Entire races can be added or removed from ballots and voter’s candidate selections can be flipped and the majority of users do not notice.”]

5., p. 10 [Professor Ted Selker of MIT reports that, “In watching 500 voters casting ballots, I saw less than one in 10 people who, when they were told they had a problem with their ballot, were actually willing to take a new ballot and vote again.”] [It has also proven difficult to catch errors when reviewing VVPATs in a post-election recount or audit. A study conducted by CalTech/MIT found that “[o]ut of 108 elections that contained errors, … no errors were reported in the VVPAT audit.”]

6. [Academic paper: “Touchscreen voting machines cause long lines and disenfranchise voters”] [f“Voting on paper ballots helps to prevent long lines, since voters don’t have to wait for an available machine before they can mark their ballots.”]

7. [Academic paper: “Touchscreen voting machines cause long lines and disenfranchise voters”] [Using hand marked paper ballots counted on optical scanners “allows many more people to simultaneously vote than would be the case with fully computerized voting.”]

8. [“Only one optical scanner is required in each polling place to serve the same number of voters as ten to twelve electronic voting machines.”]

9. [“Optical scans are also far less expensive than touchscreens. That means localities can buy more machines, keeping lines at the polls shorter.”]


11. [“State Election Commission spokesperson Chris Whitmire was widely quoted as telling people to vote on ‘paper towels’ if necessary.”]

12. [“Improperly coded memory cards” led 3/4 of all the machines in Washington County, Utah to break down. Poll sites offered backup paper ballots — ”until some of them ran out and told voters to come back later.”] [“Election officials scrambled to issue paper ballots in some locations, which quickly ran out, according to reports online. Paper ballots were not offered in other locations, such as St. George and Hurricane, local newspaper The Spectrum reported.”]


14. …


16. [Professor Richard DeMillo of Georgia Tech: “Despite the recent fascination with electronic ballot markers that print bar codes for scanning ballots, bar codes have no place in Georgia’s election system. They introduce a whole new class of vulnerabilities.”] [Professor DeMillo interview with CBS46 News in Atlanta, Georgia]

17. [“Additionally, these systems would also require programming, servicing and software licensing fees. They are also susceptible to breakdown and hacking”]


19. …



22. …


24. [“Voting machine vendor ES&S offered a remote access option in 2006 and in 2011, according to The New York Times.”]

25. Id.

26. Id.


28. Id.


30. Id.



33. [transcription below]

34. Id; see also [transcription below]

See also [Article by Computer Science Professor Andrew Appel (Princeton)]

See also [article compiling sources on this subject]




38. file:///C:/Users/jenni/AppData/Local/Packages/Microsoft.MicrosoftEdge_8wekyb3d8bbwe/TempState/Downloads/EAC_Board_of_Advisors_Verbatim_2018–04–24.pdf

39. …

40. [“Md. testers cast a vote: Election boxes easy to mess with”]; [“She isn’t worried about machines, top state elections official testifies”]; …

41. [In 2006, “Maryland’s House of Delegates” overrode LaMone and “voted 137-to-0 to drop its [Diebold] machines and switch to paper ballots.”]; [The Voting News, a Service of Verified Voting: “Then [Maryland] Governor Ehrlich allocated $28.5 million for optical-scanners in March 2006. But State Election Chief Linda Lamone spent the money on Diebold electronic poll books instead, causing nothing but problems for judges when the poll books malfunctioned. LaMone was later censured by the next Governor O’Malley for allowing Diebold to use her picture and endorsement in sales literature featuring the poll books.”]. [“Elections Chief Stars in Diebold promotion”].

42. [“In January of 2006, Rep. Harry Geisinger introduced HB790, which … proposed to … “Provide a voter verifiable permanent paper record as the official ballot of votes recorded for each voter,” but “the Committee voted against it after Elections Director, Kathy Rogers, adamantly opposed it.”]

43. … … [LaMone, Cox, and Rogers testify in favor of paperless voting in 2004] …

44. …

45. … … … …


47. …

48. …

Attorney and Election Integrity Advocate #ProtectOurVotes #PaperBallotsNow @jennycohn1