THE BOTTOM LINE
- Data integrity is paramount: Regulatory fines based on statistical performance metrics can be successfully challenged if the underlying data contains significant uncertainties. Ambiguous data points may not be sufficient to prove a violation.
- The regulator bears the burden of proof: A company’s own conservative internal reporting is not a blank check for regulators. The authorities must still independently prove an infringement beyond a reasonable doubt before imposing a penalty.
- Scrutinize your metrics: This ruling highlights the critical need for businesses to rigorously define how anomalies and missing data are handled in both regulatory reporting and commercial service level agreements (SLAs).
THE DETAILS
The Netherlands Authority for Consumers and Markets (ACM) imposed a €2 million fine on PostNL, the designated universal postal service provider, for failing to meet a key quality standard. Under Dutch law, PostNL is required to deliver at least 95% of all standard letters by the next working day. Based on a quality survey conducted for 2019, PostNL itself reported a performance rate of 94.34%, triggering the substantial penalty. However, the legal challenge hinged on a small but crucial detail within that survey: 0.6% of the “test letters” used for the measurement were never reported as received by the survey participants. PostNL had conservatively counted these letters as “delayed” in its report but argued in court that their actual fate was unknown and they should be excluded from the final calculation.
The Dutch Trade and Industry Appeals Tribunal sided with PostNL, focusing on the fundamental legal principle of the burden of proof. The court established that the ACM, as the fining authority, is responsible for proving the violation beyond a reasonable doubt. Simply relying on PostNL’s self-reported figures was insufficient, especially since PostNL had flagged the uncertainty around the unreported letters. The court found that there was no binding rule or prior agreement that dictated how these ambiguous data points should be treated, meaning PostNL was not trapped by its own cautious accounting.
Ultimately, the court determined that the €2 million fine could not stand on uncertain ground. Since the reason for the non-reporting of the 0.6% of letters was unknown—it could have been a postal delay, but it could also have been an error by the survey’s senders or recipients—they could not be definitively counted as a failure on PostNL’s part. To impose a penalty, the regulator needed certainty. By excluding this ambiguous dataset, PostNL’s performance rate rose to a level where, within the survey’s statistical margin of error, the 95% target was met. The legal basis for the fine evaporated, and the penalty was annulled.
SOURCE
Source: College van Beroep voor het bedrijfsleven
