Why Public Sector Reporting Needs an "Evidence Infrastructure"

A recurring theme in recent commentary on New Zealand's public sector - including concerns raised by the outgoing public spending watchdog - is that reporting has become disconnected from the delivery it's supposed to reflect.

The structural reason is straightforward: most institutions still treat reporting as a document exercise rather than an evidence system. Updates are gathered manually, spreadsheets are reconciled after the fact, suppliers are chased for status reports, and metrics are interpreted inconsistently across teams. The result is enormous operational friction - and a accountability picture that is often incomplete by the time it reaches the people who need it.

When reporting is this disconnected from delivery operations, several things follow. Risks stay hidden too long. Underperformance becomes difficult to prove. Contracts quietly diverge from original commitments without anyone formally registering the gap. Broader outcomes collapse into vague narratives rather than measurable indicators. And executives lose confidence in the underlying data - often with good reason.

The outgoing watchdog's concerns reflect a wider shift that is playing out globally: governments are entering an era where "trust me" reporting is no longer sufficient. What's needed instead is what might be called evidence infrastructure — systems capable of continuously collecting delivery evidence, validating milestones, tracking obligations in real time, linking spending to measurable outcomes, and surfacing emerging risks before they become failures.

This is the gap that a new generation of contract performance platforms is designed to fill. Rather than treating contracts as static records, these platforms treat them as live operational tools — using AI to extract obligations, KPIs, deadlines, and risk indicators at signing, then tracking delivery against those commitments continuously. Audit trails are built automatically. Supplier performance is measured against evidence, not self-reporting. Social and environmental outcomes are tracked within the same system rather than managed separately.

dharta was built around exactly this model. Its architecture is designed not as a contract repository but as a governance layer — one that turns procurement documents into real-time delivery intelligence and makes that intelligence accessible to the people responsible for oversight.

The practical implications are significant. Consider what becomes possible when every major public contract has live delivery visibility, supplier performance is continuously measured, and auditors can access structured evidence rather than reconstructed narratives. Ministers and boards can see emerging delivery risk as it develops, not six months later in an annual report.

None of this is futuristic. The underlying technology exists. What remains is the institutional decision to adopt it.

Public sector accountability is evolving from "did you produce a report?" to "can you prove delivery with evidence?" Those are fundamentally different standards - and they require fundamentally different tools.