Civic
A new analysis from Harvard's Ash Center argues that publishing data is not the same as producing accountability, and the evidence from two decades of civic technology backs that up.
NewsOnScale Staff
September 4, 2026
For most of the past two decades, the civic technology movement operated on a foundational premise — that sunlight is the best disinfectant. Build the portal, publish the contracts, open the budget data, and democratic accountability will follow. A growing body of evidence, including a recent analysis from Harvard's Ash Center for Democratic Governance and Innovation, suggests that premise is, at best, incomplete.
The argument is not that transparency is worthless. It is that transparency, treated as an endpoint rather than a starting point, has become a way for institutions to perform accountability without actually submitting to it.
## The Gap Between Data and Consequence
Publishing a government contract online is not the same as enforcing it. Releasing a budget dataset is not the same as preventing funds from being diverted. The distinction sounds obvious, but the civic tech sector spent enormous energy — and attracted enormous philanthropic capital — building tools that optimized for the first half of each of those equations and quietly assumed the second half would take care of itself.
It largely did not. Open data portals launched with fanfare have gone years without meaningful updates. Budget transparency platforms have documented fiscal irregularities that produced zero prosecutions. E-procurement systems have been adopted by governments that simultaneously found new off-ledger mechanisms for corruption. In each case, the data existed. The accountability did not materialize.
The Ash Center's framing identifies a structural reason for this pattern. Transparency tools are typically designed by technologists and adopted by reformers inside institutions. But the actors most capable of converting disclosed information into consequences — investigative journalists, opposition politicians, civil society organizations, regulatory bodies, courts — are external to those institutions and often systematically under-resourced. The data pipeline gets built. The accountability infrastructure that would make the data matter does not.
## Who Benefits From the Current Model
This creates an uncomfortable dynamic worth naming directly. Governments that implement transparency portals receive significant reputational credit from international donors, development banks, and governance indexes. That credit is often extended before anyone has evaluated whether the disclosed information has changed any behavior. In some documented cases, the adoption of a transparency mechanism has functioned as a shield — allowing officials to point to the portal when questioned, rather than answer the underlying question about where the money went.
This is not a fringe concern. The open government movement's own internal critics have been raising it for years. What the Ash Center analysis contributes is a systematic look at the civic tech sector's own record, which makes the pattern harder to dismiss as cynicism.
## What the Next Generation of Tools Gets Wrong — and Right
The analysis lands at a moment when AI-assisted civic tools are attracting a new wave of investment and optimism. Automated contract analysis, anomaly detection in public financial data, large language model interfaces for navigating government records — each of these represents a genuine capability improvement over the static portals of the previous generation.
But capability is not consequence. If the structural gap identified in the analysis — between information produced and accountability delivered — is not addressed at the design stage, AI civic tools will reproduce the same pattern at higher speed and with more impressive demo videos.
The more honest actors in the civic tech space are beginning to build for the full chain: not just disclosure, but verification, routing to actors who can respond, documentation of whether response occurred, and feedback loops that create pressure when it does not. That is a harder product to build and a harder outcome to measure. It is also, the evidence suggests, the only version that works.
## The Accountability Debt
Transparency infrastructure is not nothing. It is a necessary condition. The sector's error was treating it as sufficient. Two decades of deployment data are now available to test that assumption, and the results argue for a significant reorientation — toward tools that close the loop between disclosure and consequence, and toward honest accounting of what the current generation of platforms has and has not achieved.
The civic tech sector's credibility depends on its willingness to apply to itself the same evidentiary standards it demands of the institutions it monitors.