The Unread Ledger: Why Transparency Isn't the Same as Accessibility

There’s a piece of received wisdom in our field that feels so self-evident we rarely stop to question it: that publishing data online makes it public, and that public means accessible. We champion the portal launch, celebrate the massive CSV dump, and herald the API as a victory for transparency. The ledger is open, we declare. The work is done. But I’ve come to believe this is a profound, and often damaging, misconception. Transparency is the act of revealing; accessibility is the capability of understanding. We are spending immense energy on the first while largely failing at the second.

Consider a typical municipal open data portal. It might contain years of building permit records—a treasure trove for historians, urbanists, or curious residents. But to the average person, it’s a cryptic spreadsheet with column headers like ‘PERMIT_TYPE_CD’, ‘ISSUE_DT’, and ‘FEE_CALC_432’. The data is technically public, but its meaning is locked behind a wall of bureaucratic jargon and assumed contextual knowledge. It is a ledger written in a dialect only city planners and specialized software can fluently read. This isn’t openness; it’s obfuscation by format.

The Cost of Presupposed Literacy

This gap between transparency and accessibility creates a two-tiered system of citizenship. On one tier are researchers, journalists, and tech-savvy advocates with the time, tools, and literacy to parse raw data. On the other is everyone else, for whom the promised transparency is an empty gesture. The ‘public’ in public records shrinks to a technically-literate few. We’ve digitized the town square but forgotten to build steps up to its platform.

The problem extends beyond jargon. It lives in the very architecture of our digital preservation. We archive websites and datasets with fanatical precision, ensuring every bit is perfectly conserved. Yet we often neglect to preserve the shared cultural knowledge—the ‘how’ and ‘why’—necessary to interpret them. A future historian examining today’s archived permit data might understand the syntax of the dates but have no clue what a ‘FEE_CALC_432’ entailed, because the internal manual linking that code to a fee schedule was never considered part of the ‘record’ worthy of preservation.

True accessibility requires a shift from a repository mindset to a translation mindset. It means that alongside the raw CSV, we provide a plain-language glossary. It means pairing the database with narrative context—a blog post from a city official explaining what the data shows and why it matters. It means designing interfaces that allow for human exploration, not just machine consumption. It is the hard, unglamorous work of building bridges between the data and the people for whom it was ostensibly made public.

An open ledger is only meaningful if its entries can be read. Our current model of digital transparency too often delivers the world’s information in a locked box, then pats itself on the back for providing the box. The next frontier isn’t about prying more data loose from institutional silos—though that work remains vital. It’s about ensuring that once freed, the data can speak a language its public understands. Otherwise, we’re not creating an open record; we’re just publishing a secret in plain sight.

Notes & further reading

A few pages I came back to while writing this: