The course's final move pulls the threads together around the idea that has recurred in every module: information is necessary but not sufficient for accountability. Transparency — audits, open budgets, freedom of information, the citizen's right to know — is the dominant governance reform of the past two decades. It works, but only when it completes a chain. This module sets out that chain and the institutions built around it.
The accountability chain
Information → attribution → sanction
For transparency to change behaviour, three links must hold. Information: the corruption or failure must be detected and disclosed. Attribution: citizens must be able to assign responsibility to a specific actor they can act against. Sanction: there must be a mechanism — an election, a court, an administrative consequence — to actually punish, and the will to use it. Break any link and disclosure achieves nothing: the audit no one reads, the leak no one can pin on anyone, the scandal that ends no career. Most transparency reforms fail at the second or third link, not the first. Designing transparency means designing the whole chain, not just the disclosure.
When information works
The successes from earlier modules all completed the chain. Reinikka-Svensson's newspaper campaign (information) let schools and parents see what they were owed (attribution) and demand it (sanction via local pressure), and leakage fell. Ferraz-Finan's audits (information) were broadcast on radio (attribution to the mayor) before an election (sanction), and corrupt mayors lost. The pattern is consistent: transparency bites when the information is specific, attributable, and lands where someone has both the means and the motive to act on it.
Supreme audit institutions and the audit-to-action gap
Most countries have a constitutional Auditor-General (a supreme audit institution) that audits public accounts and reports to parliament, usually through a Public Accounts Committee. In principle this is the information link, professionalised. In practice many countries exhibit a chronic audit-to-action gap: the Auditor-General produces detailed annual reports documenting large irregularities, the reports are tabled, and almost nothing happens — no recovery, no prosecution, no consequence. Kenya's Auditor-General, for instance, has for years flagged substantial unsupported and irregular expenditure across national and county government, with limited follow-through. The information link works; the sanction link is broken, so the disclosure does not deter.
The transparency toolkit
- Open budgets — the Open Budget Index scores how much budget information governments publish; transparency here lets legislatures, media, and citizens scrutinise allocation and execution (the fiscal-transparency theme of the Public Budgeting course).
- Open contracting — publishing procurement data (tenders, awards, prices, beneficial ownership) so over-pricing and favouritism become visible and benchmarkable — directly enabling the price-benchmarking measurement of module 4.
- Access-to-information laws — a legal right to request government information, shifting the default from secrecy to disclosure (strong de jure, but only as good as the de facto compliance, per module 1).
- Free media and a protected civil society — the transmission mechanism for the whole chain; without independent journalism and the freedom to publish, disclosed information never reaches citizens or attributes responsibility.
Why transparency sometimes fails — and the clientelism link
Information without an empowered, motivated citizen
Transparency fails predictably when the later links are missing: when there is no sanctioning mechanism (no functioning courts, captured prosecutors, no real electoral threat); when citizens cannot attribute responsibility amid diffuse, complex government; when the collective-action problem means no individual will act on diffuse information (the recurring Olson problem); and — importantly — when clientelism dampens the sanction. Where voters are tied to patrons by targeted benefits, or vote on ethnic lines (the previous course's road-building evidence), they may not punish a corrupt-but-co-ethnic or corrupt-but-generous incumbent even when fully informed. Information assumes a citizen who is free, able, and willing to act on it; supply the information without the empowered citizen and little changes.
Closing the loop
The constructive lesson of the whole course is now in view. Good governance is not produced by any single reform — not an anti-corruption agency, not a transparency law, not devolution, not a fiscal rule. It is produced by completing accountability chains: detection that reaches transmission that reaches a sanction that an empowered citizen or institution is willing to impose, all built on a state with the capacity to act and rules credible enough to bind. Each module supplied one link or one condition. The art of governance reform is assembling them — diagnosing which link is broken in a specific case (Grindle's selectivity) and fixing that one, rather than importing the whole apparatus and hoping function follows form.
Exercise
A country's Auditor-General publishes a damning annual report every year detailing billions in irregular spending. Media cover it briefly; little else happens, and the same irregularities recur. (1) Locate the failure precisely on the accountability chain. (2) Give three distinct reasons the sanction link might be broken, spanning institutions and politics. (3) The government proposes a new access-to-information law and an online open-budget portal as the fix. Evaluate whether these address the actual broken link. (4) Design a set of reforms that would complete the chain, and connect each to a specific module of this course.