The rules that guide how member nations participate in the multibillion-euro Framework programmes of research run by the European Union (EU) are complex, confusing and in urgent need of simplification — and nothing illustrates that better than the ongoing 're-audit' of the Sixth Framework Programme (FP6). The programme closed in 2006, but in recent months several national research agencies have been accused of cheating by the European Commission, the EU's executive arm, and have been asked to pay back grant money worth millions of euros.

The affair began last year, when pressure from the European Court of Auditors led the commission to re-audit sample files from those national research agencies that had handled the largest numbers of FP6 projects. The re-audit unearthed many cases of what the commission insists are irregularities, mainly in personnel costs and overheads. In those cases in which the same irregularities were found in each of the sampled files from an agency, the commission assumed them to be systematic and demanded that the money be returned. For a research agency such as France's CNRS, which handled nearly 900 FP6 projects, the sums being disputed amount to tens of millions of euros.

The agencies, however, are incensed. They maintain that they operated with the commission's approval, following an interpretation of the FP6 rules that allowed them to use their own customary national accounting and management practices. Now, they say, the commission has reinterpreted the rules after the fact, and has used the re-audit to impose penalties on the same customary practices it previously approved.

The exercise has become expensive for all concerned. The accused research agencies are having to allocate already-scarce resources to a check of all their individual files, trying to prove that they did nothing wrong. And the commission has had to hire extra staff to assist with the unanticipated re-audits — to the point that the enterprise is costing much more money than the commission could ever expect to gain though repayments.

Worse is the loss of good will. Some research organizations say that their scientists are becoming wary of involvement with Framework programmes for fear of future unforeseeable bombshells. The mood is not helped by the commission's proclamation that only errors in its own favour are subject to correction — those financial mistakes that would require a repayment to project scientists cannot be reimbursed because the FP6 is officially closed.

Both sides had hoped that the rules of participation would be clearer for the Seventh Framework Programme (FP7). The commission had promised to grant 'methodology certificates' to institutes whose customary practices it explicitly accepts. But as that seven-year programme reaches its halfway point, virtually no certificates have been awarded because the commission still lacks a confident understanding of its own rules — and that leaves the FP7 as vulnerable as the FP6.

The FP6 re-audit debacle will probably be concluded behind closed doors; neither the commission nor the research agencies seem to want a public scandal. In the meantime, there is — at last — broad political will to simplify the rules of the Eighth Framework Programme (FP8), set to launch in 2014.

This simplification must be accompanied by an infusion of institutional trust. On the whole, scientists are not potential criminals; they need to be attracted to, not repelled by, Framework programmes. Like all public funds, Framework money must be accounted for appropriately — but heavy-handed, after-the-fact meddling is not the best way to do it. Straightforward audits that are based on clear, consistent and agreed rules are a much better approach. It is too late for the FP6, and perhaps too late for the FP7. But the FP8 should treat researchers with the trust and dignity they need to do their best work.