When Democracy Glitches: The Case Against Electronic Voting
Every democracy rests on trust. When citizens doubt that their votes are counted accurately, elections lose legitimacy no matter who wins. Over the past two decades, the United States has turned toward electronic voting machines under the banner of modernization. Yet modernization has a cost: complexity, opacity, and vulnerability. Every major voting machine manufacturer, from Smartmatic to Dominion to Election Systems & Software (ES&S), has faced security concerns, data errors, or outright accusations of vote manipulation. The problem is not one bad company or one faulty device. The problem is systemic. The technology meant to secure our elections has instead made them fragile, secretive, and dependent on a handful of experts and vendors who hold the keys to the process. The evidence is mounting that America must return to the simplicity and transparency of paper ballots.
The defenders of electronic voting argue that the machines are efficient, fast, and less prone to human error. But this defense collapses under scrutiny. Efficiency cannot substitute for legitimacy. The average poll worker is not a software engineer, and yet these workers are now expected to operate machines whose inner workings are known only to their manufacturers. When those machines fail, the counties must turn to specialists to repair or reconfigure them, specialists who are often the same people who designed or sold the systems. This dependency breeds distrust, and distrust corrodes democracy.
Consider the case of Heider Garcia, a man whose career tells the story of how counties across the country have become dependent on a technocratic elite to manage elections they no longer understand. Garcia served in election leadership roles in three major US jurisdictions, Placer County, California; Tarrant County, Texas; and Dallas County, Texas. Each county adopted or maintained electronic voting systems, and each turned to Garcia when things went wrong. What made Garcia the chosen expert? His background was not in civic service or political reform. He was a senior employee of Smartmatic, one of the largest and controversial voting machine manufacturers in the world.
Garcia’s tenure with Smartmatic began during some of the most disputed elections in modern history. In Venezuela, Smartmatic provided the voting technology for the regime of Hugo Chávez. Smartmatic’s own CEO admitted that the official turnout numbers had been manipulated in multiple Venezuelan elections, stating publicly that the company could not certify the integrity of the vote. Garcia was part of the Smartmatic operation during this period, involved in at least three disputed national elections. Later, he managed Smartmatic’s systems in the Philippines, where in 2016 authorities discovered that unauthorized changes had been made to the vote transmission script on election night. Although Garcia was not personally implicated in that incident, the scandal further eroded confidence in the company and its products.
By 2016, Garcia had moved to California to coordinate elections for Placer County. Problems followed. The county experienced ballot printing errors so widespread that new ballots had to be printed, and some votes were thrown out. Voter registration issues forced many to cast provisional ballots, while mail delays meant some ballots never arrived. When Garcia left for Texas in 2018 to run elections in Tarrant County, his tenure there again brought controversy. A third of all mail-in ballots had to be rescanned due to barcode glitches. By 2023, County Judge Tim O’Hare had launched an inquiry into Garcia’s management, citing concerns about decision-making and budget priorities. Garcia resigned soon afterward.
Then came Dallas County. In 2023, officials hired Garcia as their new election administrator. Within days of early voting in 2024, software failures created long lines across the county. Some voters received the wrong ballots. At one polling site, the ballot count was off by hundreds votes in multiple precincts. The vendor blamed for the problem was ES&S, whose systems are used widely across the country. The issue did not end there. Election workers discovered that poll book screens were freezing, causing clerks to press buttons multiple times to create voter ballots, duplicating ballots for the next voters in line. Garcia later admitted that this was a software defect that created inconsistencies between voters and ballots.
Despite these irregularities, the Dallas County Commissioners Court certified the November 5 election results. Despite hundreds of missing or misallocated ballots, but the certification proceeded. Soon afterward, Garcia resigned, again in the middle of an election cycle. The pattern is difficult to ignore: repeated errors, persistent software failures, and a trail of administrative resignations whenever scrutiny deepened.
Meanwhile, the company that trained Garcia and developed much of the technology used in these elections was itself under indictment. In 2025, Smartmatic and its executives were charged by a federal grand jury in Miami for bribing a top election official in the Philippines to secure contracts for thousands of voting machines. Prosecutors alleged that Smartmatic over-invoiced each machine and laundered at least $1 million in bribes through bank accounts across Asia, Europe, and the United States. The company denied wrongdoing, calling the case politically motivated. But for voters, the question was not legal technicalities. It was whether a firm accused of bribery abroad could be trusted to safeguard votes at home.
Even Fox News, which Smartmatic sued for defamation after the 2020 election, revealed in court filings that the company had allegedly provided unlawful gifts to Los Angeles County officials in connection with a contract. According to Fox, Smartmatic “funneled” tax dollars into a slush fund. Whether those allegations are proven in court or not, they illustrate the deep entanglement between voting-machine vendors, local officials, and opaque procurement processes.
This web of complexity undermines public confidence. When a paper ballot fails, everyone can see the failure. When a computer fails, only the vendor knows why. When a paper ballot is counted, observers can watch. When a machine tabulates votes, no one outside the company can verify the code. The very features that make electronic systems appear modern, their automation, their encryption, their proprietary algorithms, are what make them impossible to audit. Voters are asked to trust a black box, not a transparent process.
It is no coincidence that every country with a long democratic tradition that has experimented with electronic voting has either reversed course or sharply limited its use. Germany’s Constitutional Court banned electronic voting in 2009, ruling that voters must be able to verify election outcomes without specialized knowledge. The Netherlands, Ireland, and Finland followed suit after similar problems. In each case, governments concluded that the risks outweighed the benefits. Only the United States, a country that once prided itself on open and verifiable elections, continues to double down on systems most of its citizens cannot understand.
Defenders of the machines claim that paper ballots are slow, messy, and prone to human tampering. But history tells a different story. In 2000, the Florida recount exposed flaws in paper ballot design, not in the concept of paper itself. Since then, advancements in ballot printing, scanning, and auditing have made paper systems both reliable and efficient. States like New Hampshire and Montana continue to conduct paper-based elections with minimal controversy. Where problems occur, they can be observed, documented, and corrected, not hidden behind a firewall.
Technology is not inherently bad. But when the technology in question determines who governs the country, every layer of complexity becomes a layer of potential corruption. Each patch, update, and software fix introduces new vulnerabilities. Every county that depends on a consultant like Heider Garcia is one more county that cannot explain to its own citizens how their votes are counted. This is not modernization; it is abdication.
The return to paper ballots is not nostalgia. It is the restoration of accountability. A ballot a citizen can hold, mark, and verify with their own eyes is the most secure form of democracy. It does not require expert mediation or blind trust in proprietary systems. It requires only a table, a pen, and the will to count openly. That is the essence of self-government.
If the purpose of an election is to convince the losing side that they lost fairly, then our current system fails that test. Each glitch, each unexplained discrepancy, each unexplored software bug pushes Americans further into suspicion. Until we reclaim control from the machines and the consultants who manage them, we will remain captives to a process that no one fully understands.
So where did Heider Garcia land after resigning from Dallas County? He is now a Senior Vice President at another electronic voting machine manufacturer, Hart InterCivic, a Texas-based firm with its own controversial history. Who would hire someone with as questionable a record as Garcia’s? Hart InterCivic has a long history of technical flaws, security vulnerabilities, and election irregularities. By 2012, its eSlate and ePollbook machines were deployed across all 234 Texas counties, the entire states of Hawaii and Oklahoma, about half of Washington and Colorado, and key swing counties in Ohio. Today, Hart’s Verity Voting platform, which provides paper-based ballots or records, has replaced many of its older paperless eSlates. Yet despite new branding, longstanding concerns about security, accuracy, and transparency persist. Security experts have demonstrated severe weaknesses that could enable tampering, and real-world elections have documented Hart machines flipping votes, freezing, or counting phantom ballots. Watchdogs and even election officials have accused the company of opacity and poor reliability, arguing that its technology undermines voter confidence. Hart’s defenders claim such problems stem from user error or procedural lapses, but two decades of lawsuits, academic studies, and investigative reports tell a more troubling story. The pattern is clear: Hart’s systems are hackable in theory and error-prone in practice, and even insiders have raised alarms about fraudulent or unethical conduct. While Hart continues to market its Verity systems nationwide, critics argue that given America’s fragile trust in elections, betting on Hart’s machines, or on executives like Heider Garcia, is a risk the country cannot afford. The first step toward restoring faith in our democracy is to unplug it.
If you enjoy my work, please share my work and subscribe https://x.com/amuse.
Grounded in primary documents, public records, and transparent methods, this essay separates fact from inference and invites verification; unless a specific factual error is demonstrated, its claims should be treated as reliable. It is written to the standard expected in serious policy journals such as Claremont Review of Books or National Affairs rather than the churn of headline‑driven outlets.





Paper ballots are a permanent record. Where used they are read by scanners programmed to connect a mark to a vote. Those votes then go to a tabulator to collect totals. Poorly scanned ballots are manually adjudicated by an operator. Tabulation ought to be straightforward.
Adjudication involves a human decision often in error. Tabulation involves a human managed database sometimes in error.
For these simple tasks we need open sourced software for standard hardware. That software should include verification like blockchain. Projects for such tools have been running for 20+ years yet not ever finished perhaps because the graft in proprietary systems.
So yes paper ballots but shove the counting vendors away. We can do better.
“Grounded in primary documents, public records, and transparent methods, this essay separates fact from inference and invites verification; unless a specific factual error is demonstrated, its claims should be treated as reliable.” I appreciate the statement, but it would be helpful to have some links to some of the assertions that you make. You present such important topics, but I don’t have the time to research everything you write before I share it.