The most revealing statistic in this year’s debate over European democracy is not a measure of progress but of repetition. According to an independent assessment by the civil liberties network Liberties, 93 percent of the recommendations the Commission issued to member states in its rule of law cycle simply repeated earlier ones, frequently word for word, while the number of genuinely new recommendations was cut roughly in half compared with the previous year. A watchdog that says the same thing twice may be thorough. A watchdog that says it for the fifth time is describing a problem it has not been able to fix.

The Commission’s annual rule of law report was designed as a soft instrument, a yearly health check covering judicial independence, media freedom, anti-corruption measures and institutional checks across the Union and several enlargement candidates. Its theory of change was always gentle: name the weaknesses, repeat the naming, and let peer pressure and public attention do the rest. On the more technical questions the approach shows results. The accompanying Justice Scoreboard this year recorded an improved perception of judicial independence and efficiency in several states, the kind of incremental gain that quiet diplomacy can deliver.

The trouble is that the diagnoses most likely to be repeated are the ones that touch power directly. Stagnation, not backsliding or repair, has become the dominant pattern. Where reforms would require a government to loosen its grip on prosecutors, public broadcasters or the financing of political life, recommendations tend to be acknowledged, filed and ignored. The report can document this with rising precision, but documentation is not leverage. The instrument was built to nudge governments that wanted to be seen as cooperative, not to move those that have decided non-compliance carries no real cost.

That is where the Commission’s other tool comes in, and where its limits also show. The June infringements package, published this month, illustrated both the reach and the slowness of legal enforcement. Spain and Poland were referred to the Court of Justice for failing to transpose revised emissions trading rules, with requests for financial penalties attached. Procedures were opened against Greece, Luxembourg and Sweden over the directive on combating money laundering, and Spain drew a separate notice over the excessive collection of travellers’ personal data. Each step is the product of months, sometimes years, of formal notices and reasoned opinions. By the time a case reaches Luxembourg, the political moment that produced the breach has often passed.

A more uncomfortable theme has crept into this year’s assessments: the rule of law inside the Union’s own institutions. Liberties and other observers point to undemocratic practices in how decisions are taken in Brussels itself, from opaque trilogue negotiations to the handling of conflicts of interest at senior levels. The charge stings because the Union’s authority to lecture capitals rests entirely on the claim to be a standard-bearer. A body that asks others to be transparent and accountable cannot afford to be seen cutting its own corners, and the gap between the demand and the example is exactly the kind of thing that erodes the moral weight of a recommendation.

So the annual ritual arrives at an awkward crossroads. The reporting has matured into something detailed and credible; few now dispute the quality of the evidence it gathers. What remains unresolved is the question of consequence. Conditionality rules that tie EU funds to rule of law compliance have been used sparingly and cautiously, and infringement actions move at the pace of litigation rather than politics. Repetition without escalation teaches governments a lesson the authors never intended, which is that the warning and the penalty are not the same thing, and that one can be safely separated from the other.

For now the cycle will continue, more granular each year, more honest about institutions including the Union’s own. Whether honesty without enforcement amounts to protection of democracy, or merely to an unusually well-footnoted account of its slow erosion, is the question the next report will inherit unanswered.