"Most Secure Election in American History."

Ok. What Does That Mean? What Should We Do Now?


There are two competing narratives about the 2020 election -- it was rigged and it was the most secure election ever. Both can’t be correct unless all of our presidential elections have been rigged to varying degrees. This reality tells me that one of the two assertions is closer to the truth than the other but still misses the mark. So I set out to fill in the gap. I went online to do some reading. I also contacted a nonpartisan, nonprofit organization that is at the forefront of creating a modern, more secure voting machine system to replace obsolete software dating back to the time Bill Clinton was president.

We need more information, I think, because many Americans, for reasons sound and not-so-sound, doubt the credibility of vote counts and election results.

To start with, I wanted to learn the meaning of the phrase “the most secure election in American history”. A founder and Chief Operating Officer of OSET Institute gave me a detailed answer in an email. Gregory Miller helped start OSET in November 2006 to create an open-source voting machine system that manufacturers and voting jurisdictions can use to replace obsolete digital machines across the country. I am reprinting his answer in full because it clarifies what the government meant when it said the 2020 election was the most secure in American history. I have put in bold portions of the statement that seem significant and new to me.

Miller said, “First, it is very important to bear in mind that the statement released from CISA regarding the 2020 election was not of their making alone, but of a collective including the members of the Election Infrastructure Government Coordinating Council (EIGCC), NASS, NASED, the Election Infrastructure Security Coordinating Committee (EISCC which includes vendors). That's important because it means this assertion is a consensus opinion or a "joint statement."  

  1. Second, "ever" as the term was applied here covers the period of time that election administration activity has been actively monitored by the Feds (DHS) for nefarious interference--primarily disruption and subversion (but not disinformation).  So, for sake of argument, we consider this to be the period 2008-forward (although issues in 2002, '04, and '06 increasingly drew the attention of experts and began to engage DHS).

“Now let's look at how we (OSET Institute) parse that based on our own interactions.

“PART 1: The DHS/CISA VIEW

“From the standpoint of the federal government agencies charged with monitoring and assisting in the protection of critical infrastructure, this election was -- in their professional opinion -- the "most secure ever" based on what they were monitoring which is in near totality an array of mechanisms, services, monitors, and devices surveilling for external network intrusion attempts.  The simplest, but not only, example is the deployment of Albert Sensors.  In short, from the DHS/CISA view of the universe, this election witnessed the lowest amount of mendacious or even suspicious network activity.  But obviously, that is only a small part of the story.  What about the balance of that consensus assessment and their viewpoints?

“PART 2: The Election Experts of the Stakeholder Community (NASS, NASED, EIGCC, and EISCC)

“For these groups of experts 2020 was the most secure election in terms of eight factors:

  1. “The 2020 election relied on the lowest use of paperless voting machines.

  2. The 2020 election produced the highest portion of voters using paper ballots uniformly counted by optical scanning devices.

  3. The 2020 election had the most jurisdictions conducting ballots audits, that are or will become a regular process (admittedly still a small fraction, but growing).

  4. The 2020 election produced the greatest number of state and local election staff receiving basic training for cybersecurity (we were involved in such efforts).

  5. The 2020 election deployed the greatest amount of cyber defenses by states to monitor and protect voter records system (see "PART 1" above).

  6. The 2020 election experienced the greatest amount of cyber assistance provided to states by DHS (this "detente" between the Feds and States was a monumental achievement).

  7. The 2020 election produced the broadest participation by states and localities in cybersecurity information sharing organizations.

  8. The 2020 election had the greatest resiliency planning to date for state and local levels (i.e., contingency planning; mostly driven by COVID-19).”

Miller’s comments demonstrate the relative nature of the government declaration on security. 2020 was better than 2016, 2012, and 2008. It was better in specific ways as viewed through the lens of election experts in the stakeholder community. His remarks also tell us that important parts of our system were not covered in the security declaration.

What follows are links to information that take aim at some of those ignored components. The reports linked below question the ethical and technical underpinnings of our voting system; affirming clearly that we have not reached Security Nirvana yet.

  • A 2020 DC Report investigation examined Election Systems & Software, the largest maker of electronic voting machine systems for US voting jurisdictions. DC Report discussed ethical questions about ES&S, which is owned by a private equity company. I sent an email to Mike McCarthy, a power behind both companies. I invited him to comment. He did not respond.

  • A 2020 Who. What. Why. article and interview by Scrutineers founder Emily Levy. She interviewed Bennie Smith, who discovered, several elections earlier, that some voting machines in Memphis changed whole votes into fractions in mostly black precincts. According to Levy, “When losers of a local election came to him claiming their races had been stolen, he set out to prove them wrong. 

    “Instead, he discovered a high-risk tampering mechanism built into US election systems.” Smith is now an elections commissioner in Shelby County, which includes Memphis. He calls the digital anomaly he discovered “Fraction Magic.”

    According to Miller, “The OSET Institute's TrustTheVote Project is building ElectOS (the open-source voting software you mention), which among other aspects of election administration, does indeed, include ballot casting and counting mechanisms.  However, our technology framework does not provide for, or support any fractional or pro-rata voting.”

  • A 2020 article in The Markup asked, “Why Do Voting Machines Break on Election Day?” It recounted recent problems, saying, “Take the 2020 primaries…“Coding mishaps” in Iowa; “broken machines” in Dallas and New York; “glitches” in L.A.; in parts of Georgia, voters waited as long as eight hours during the June 9 primary as poll workers wrestled with new voting equipment.” The Markup also reported, “In 2010, one precinct in the Bronx threw out an unusually high percentage of ballots due to voters marking more than one choice. State election officials investigated and found that the machines had added “phantom votes” in part because they were overheating.”

  • A National Election Defense Coalition report said, “Like all computers, these systems are susceptible to software bugs and programming errors, as well as miscalibrations that “flip” votes, and machine breakdowns that cause long lines, particularly in minority communities where older machines are often placed.”

  • A 2018 Scientific American article revealed how problems with electronic machines cause confusion. “Some early voters in Texas have already reported votes they cast for Democratic U.S. Senate challenger Beto O’Rourke were switched on-screen to incumbent Republican Sen. Ted Cruz. There’s no evidence of hacking, and the particular machines in question are known to have software bugs, which could account for the errors.”

The reports I referenced above are not meant to be a comprehensive accounting of issues related to ethics and old software and aging machines that lack replacement parts. However, in an article published on the OSET website about two weeks after the 2020 election, Miller wrote that “we must not believe that ‘the most secure election’ means that there is no further need for government funding or innovation in the existing election technology infrastructure, or that we can just let the status quo remain. We cannot.”

The nonprofit is working to improve the status quo. OSET Institute is doing this by creating an entire voting machine system that is “ based on a durable paper ballot of record,” according to the OSET website. They are devoting 85% of their effort to technology R&D, 10% for cybersecurity advisory, and 5% to public policy and government relations work.

In an email, Eddie Perez explained the goal of the project. He is the Global Director of Technology Development & Open Standards at OSET Institute. Previously he directed product management at Hart InterCivic, one of the three major voting systems vendors in the US.

“We are developing and will publish as public technology (i.e., "open source" and under an OSI compliant open source license -- either our OPL or the GPL) ElectOS, an election technology framework freely available for any vendor or election jurisdiction to adopt, adapt (to local regs) and deploy.

I asked Perez if the three companies that dominate the voting machine system business in the US would maintain their “monopoly” if they used OSET software. His extended answer follows:

  • Any vendor--including the three largest--can implement systems using ElectOS, and we sincerely hope they do (and that everyone does; someday worldwide and as prevalent in this sector of Gov I.T. as open-source Linux dominates in the enterprise computing and global Internet infrastructure.)

  • NO, their collective "oligopolistic behavior" will fail to continue. (NOTEit has not been legally established that there is a monopolist among them, nor even that there exists an oligopoly, despite apparent behaviors. Our common sense tells us otherwise, but we defer to a judicial proceeding to eventually make that call)

    • The reason for this is that the underlying technology on which they build their new systems will be based on open data, open standards, and open-source.  

    • That means they can no longer institute proprietary practices that would restrict customers’ ability to change vendors (switching costs). 

    • And the existence of the ElectOS technology will lower barriers for new commercial entrants as the business model will become one of "systems integration" and not a manufacturer of proprietary black-box systems. The basic reason is that the technology will already be developed, certified, and available, which will "jumpstart" any vendor seeking entry into this marketplace.

    • And in fact, it will broaden the market in another vector: those election jurisdictions that desire to will be able to procure a system integration vendor to implement a system on their behalf (this was the business model for the LA County VSAP system; LAC had the technology and hired a vendor (Smartmatic) to produce the finished system. Bear in mind that some other nuances and complexity of that business arrangement derailed their open-source intentions, so as an example, it is limited in reference.

Perez continued, “More generally, and as we've discussed in the past, a "public technology option" for voting systems means that states can allow localities more options in terms of which companies it chooses to do business with to provide and support voting systems. Yes, this is all handled through procurement but bear in mind states maintain "schedules" of pre-qualified vendors.  

He added, “So again, regardless of whether the three largest systems vendors decided to adopt ElectOS (similar to IBM's decision, for example, to adopt Linux over their own proprietary OS/2), the public technology option now opens a wide array of options for building, delivery, deployment, and support of election technology.  That makes for a better performing and well-behaved market.”

Many of the approximately 10,000 voting jurisdictions in the US will have to spend hundreds of millions of dollars combined to replace their systems which are up to two decades old now. 

Will OSET offer an alternative that can help increase trust in the system and make the voting experience work better for voters?  We will start to learn the answer, probably within the next decade.

Officials connected with Elections Systems & Software and Dominion Voting – the two largest companies in their field -- did not respond to requests for comment.


Sidebar: “Vegan Voting” or “Luddite Lite Voting” Anyone?

Digital technology is everywhere. But that doesn’t mean that everyone appreciates its virtues and vices. In fact, at least one American believes we should scrap computerized voting systems for paper ballots and manual counting of all votes. A fellow American explains:

"I'm always reporting on voting machines, the better versions, the companies involved, the lobbying, and any related news but I personally don't understand why we have them at all,” said Adrian Tawfik, the 30-something who created Democracy Chronicles a decade ago. He edits the online publication, running articles on voting, election reform, and pro-democracy movements at home and around the world. (Disclosure: he has published some of my articles.)

Tawfik continued, “Perfectly functioning, well-designed voting machines may exist but it is just so easy for the average voter to see them as suspect as they would any tech. Modern people are rightly suspicious about tech. It's everywhere, it's powerful, and it is recording you, especially in big cities. I think voting machines just add unneeded complexity to elections and lend to the appearance of corruption if not corruption itself. Paper ballots have worked fine for years. Let's just take the time to count manually and have no mechanical or electric device involved at all."