The Double-Edged Screen: How Digital Tools Can Support or Undermine Trust in Elections

I spend a lot of time thinking about the architecture of participation. That phrase comes from my years at Facebook, where I worked on civic engagement products, and it’s become my shorthand for how the design of a digital tool shapes what people do with it. Elections, at their core, are a massive act of collective participation, and increasingly, they run on digital rails. From voter registration databases to the social media feeds where we argue about candidates, these systems are not neutral conduits. They can shore up the foundations of democratic trust, or they can chip away at them—one line of code, one algorithmic decision at a time.

The question isn’t whether digital tools are good or bad for elections. They are both, simultaneously, often within the same platform. The real work is in understanding the mechanisms: how a tool’s specific features interact with human psychology and institutional fragility to produce an outcome. When I analyze a new election technology, I look for three things: the transparency of its process, the clarity of its error-handling, and the distribution of its power. Get those wrong, and you’ve built a machine for sowing doubt.

The Promise: When Digital Tools Build Confidence

Let’s start with the positive case, because it’s real and it’s powerful. The most straightforward example is voter registration. In many places, the shift from paper forms to online portals has been a quiet revolution. I still remember my first registration—hunched over a card table in a public library, squinting at a tiny form with boxes the size of my pinky nail. The digital alternative is often a clear, step-by-step process with immediate validation. Mistype your driver’s license number? The system tells you right then, not weeks later in a rejection letter. This kind of real-time error correction does something subtle but deep: it treats the voter as a partner in a process, not a supplicant subject to an opaque bureaucracy. When you can see your registration status, confirm your polling place, and even track a mail-in ballot online, the abstract machinery of election administration becomes tangible. It feels less like a black box and more like a service you can verify.

Another area of promise is the rapid dissemination of official information. Election officials, often under-resourced and overlooked, can use websites and verified social media accounts to directly counter falsehoods. During a recent election, I watched a county clerk in a small jurisdiction methodically debunk a rumor about polling place hours that was spreading on a local Facebook group. She posted a photo of the official notice—a simple, time-stamped document. That act of radical transparency, enabled by a platform that often does the opposite, cut the rumor’s spread substantially in her community. Digital tools can also make the post-election audit process vastly more accessible. When jurisdictions publish cast vote records—anonymized digital representations of every ballot—they invite the public to run the math themselves. This isn’t just about trust; it’s about verifiability. It shifts the basis of confidence from “I trust the officials” to “I can see the evidence.”

The Peril: How Design Erodes Faith in the Process

Yet, for every tool that clarifies, there’s another that obscures. The most potent threat isn’t a dramatic hacking of vote totals, though that risk is real. It’s the slow, ambient erosion of trust generated by platforms optimized for engagement rather than accuracy. Social media’s core business model is a fundamental challenge to electoral integrity. The algorithm that serves you content isn’t designed to inform you; it’s designed to hold your attention. And nothing holds attention like outrage, fear, and partisan identity. An election, which should be a deliberative civic act, is processed by these systems as a multi-week, high-stakes spectacle—a perfect storm of emotional engagement.

Consider the simple act of seeing an election result. On a well-designed official site, you see a table of numbers, clearly sourced and updated with a time-stamp. On a social feed, you might see a decontextualized graphic, a video of someone claiming a “dump” of ballots, or a post from a political operative framing a perfectly normal counting process as a sinister anomaly. The platform’s design—the infinite scroll, the share button without friction, the algorithmically sorted feed—strips away the procedural context that gives the numbers meaning. It trains users to expect instant, emotionally satisfying narratives, not the slow, meticulous, and often boring reality of counting ballots over several days. This mismatch in tempo is a structural attack on trust. When the process doesn’t meet the platform’s artificial deadline, the platform doesn’t apologize; it fills the void with conspiracy.

Another digital pitfall is the use of electronic voting machines that produce no voter-verified paper record. From a product design perspective, this is a failure of basic feedback loops. A voter interacts with a screen, the machine stores a record internally, and the voter walks away with no independent evidence of what was recorded. The system asks for total faith in a piece of complex software and hardware that the voter cannot inspect. When something goes wrong—a calibration error on a touchscreen that flips a vote, a software bug that misrecords a selection—there is no way for the voter to have caught it, and often no independent way to audit it. This design choice, often made in the name of convenience or accessibility, bakes a fundamental unverifiability into the heart of the election. It transforms a public act of record into a private, invisible transaction within a machine. The tool, intended to modernize, instead undermines the very public, observable nature that gives elections their legitimacy.

The Transparency Paradox

We also have to wrestle with a paradox. More data doesn’t automatically mean more trust. The push for radical transparency in elections can backfire when the raw data is complex and easily misread. Releasing a massive CSV file of raw vote counts without context, clear metadata, and a user-friendly front-end is an invitation for bad-faith actors to find “anomalies” that are actually just normal statistical noise. An unrounded fraction in a spreadsheet becomes proof of a rigged algorithm. A batch of votes that heavily favors one candidate because it comes from a specific precinct becomes a “suspicious vote dump.” The tool—the database—is neutral. The design failure is in the interface, the lack of a narrative wrapper that explains what the data represents. Transparency without interpretability is just noise, and noise in a polarized environment gets weaponized.

The Architecture of Participation: Principles for a Healthier System

So, how do we build digital tools that support trust rather than undermine it? I come back to my three-part test: transparency of process, clarity of error-handling, and distribution of power.

Transparency of process means that the tool reveals its own workings. A ballot-tracking system should show not just the current status, but the journey: “Ballot mailed on Oct 12. Received by postal service Oct 14. Arrived at county facility Oct 18. Signature verified Oct 20. Ballot counted.” Each step is a data point that a voter can independently confirm against their own experience, creating a chain of verifiable actions. Clarity of error-handling means that when something breaks, the system admits it plainly and offers a remedy. An error message shouldn’t just say “Transaction failed.” It should say, “We couldn’t match your signature. Here’s exactly what you need to do to cure your ballot, and here is the deadline.” This treats the voter not as a problem, but as an essential participant in a fixable process. Distribution of power is the hardest principle. It asks: who can see, who can verify, who can act? A system that only allows a small group of technicians to audit results is fragile. A system that publishes encrypted, anonymized ballot records for anyone to download and analyze with free tools distributes the power to verify. It moves us from a model of trust-me authority to a model of check-me proof.

None of this is easy. It requires election officials to think like product managers, and it requires technology companies to accept that their platforms are de facto public infrastructure for democracy—a role they have often been reluctant to embrace. There’s a deep tension between the culture of “move fast and break things” and the culture of election administration, which rightly values stability, predictability, and meticulous procedure above all else. Bridging that gap is the central challenge of our field.

Where the Responsibility Lies

This isn’t a problem we can solve with a single piece of legislation or a new app. It requires a layered approach. Election officials need sustained funding not just for equipment, but for communication staff, user-experience designers, and security auditors who can make transparency functional. Platforms need to fundamentally re-architect how they treat election-related content during defined sensitive periods, prioritizing authoritative sources and slowing down the velocity of group-sharing for unvetted claims. And we, as voters and citizens, need to cultivate a new kind of digital literacy—not just the ability to spot a fake news article, but an understanding of how real election procedures work. The best defense against a conspiracy theory about vote counting is a population that knows a legitimate count often takes days, involves bipartisan observers, and is done according to public, verifiable rules. That knowledge is the bedrock upon which all the digital tools must rest. Without it, even the best-designed system is just another object of suspicion.

Frequently Asked Questions

Why is a paper trail so important for electronic voting machines?

A voter-verified paper audit trail provides an independent, human-readable record of each vote that the voter can verify before casting. This record is not dependent on software, which can have bugs or be manipulated. It serves as the ultimate source of truth for a recount or post-election audit, allowing officials to check the machine’s electronic totals against a physical, tangible record. Without it, the public must place blind trust in complex, proprietary software that is difficult to inspect fully, creating a critical vulnerability for election integrity and public confidence.

How can I tell if a piece of election information on social media is trustworthy?

First, check the source. Is it from an official election authority, like a state or county board of elections, or a verified, reputable news organization? Be wary of content from anonymous accounts, partisan influencers, or groups designed to look like news outlets. Second, look for procedural details. Trustworthy information often explains the “how”—how ballots are processed, how results are tabulated, what security measures are in place. Misinformation tends to be emotionally charged, vague on process, and focused on shocking allegations. Finally, if a claim makes you feel an urgent, angry reaction, pause before sharing. That emotional pull is often a feature of the content’s design, not a sign of its truth.

What is a “risk-limiting audit” and how does it use digital tools?

A risk-limiting audit is a modern procedure that uses statistical methods to check whether a reported election outcome is correct. It involves manually examining a random sample of paper ballots. Digital tools are essential: they generate the statistically sound random seed, help select the sample of ballots to retrieve, and compute the audit’s progress in real time. When done well, these tools make the audit transparent—the process is documented, the math is open for review, and the results are published. This contrasts with traditional recounts that only count a fixed percentage and stop. A risk-limiting audit provides strong statistical confidence that the declared winner really won, or it escalates to a full recount if anomalies are found, efficiently combining human verification with digital precision.