When it comes to the Commission’s Annual Rule of Law Report (“ARoLR”), first launched in 2020, the years go by and they all look the same.

In its latest edition published last July, the Commission referred to a “broadly positive trajectory” regarding the rule of law in the EU. The year before, the Commission similarly asserted a “positive reform dynamic” and “positive trajectory” in the EU. This phrasing was virtually identical to the one used in 2024 (“positive reform dynamic”) and 2023 (“positive trend”). Simultaneously, impressive implementation/progress claims continue to be made in relation to the country specific recommendations (“CSRs”) which the Commission started including in each country report in 2022. The following year, the Commission claimed that almost 65% of the 2022 CSRs were followed up. This figure increased to 68% in 2024 before decreasing to 57% in 2025, with 47% of the CSRs allegedly fully or partially implemented in 2026.

Spin v. reality

These claims and figures do not, however, survive close scrutiny. For instance, while the Commission’s 2025 ARoLR speaks of a “positive trajectory”, the 2025 World Justice Project’s Rule of Law Index shows a decline in the rule of law score of 18 Member States out of 27. Denying the reality of a negative trajectory is all the more surprising when one considers the findings of another authoritative annual democracy report published a few months before the Commission’s 2026 ARoLR. In short, according to the 2026 V-DEM Democracy Report, the EU is “greatly affected” by autocratisation processes, with four additional EU Member States (Croatia, Italy, Slovakia, and Slovenia) joining Greece, Hungary and Romania as countries experiencing backsliding. (Un)surprisingly, the Commission neither refers to the WJP rule of law index, nor the V-DEM democracy report.

In addition to omitting inconvenient sources, the Commission is continuing to assert CSR implementation figures that beggar belief. For instance, one may contrast the 2025 Commission figure of 57% with the finding of a study published in November 2025 showing an opposite result with 61% of the CSRs showing little or no progress. Similarly, while the Commission claimed an implementation figure of 47% in 2026, an independent assessment found that no more than 2% of the CSRs have been fully implemented, with most remaining “in a stagnant state of partial implementation”.

The Potemkin-like nature of the Commission’s implementation figures is facilitated by a number of methodological issues the Commission continues to ignore by repeating that the ARoLR “is based on a solid, transparent and objective methodology”. Meanwhile in the real world, it remains unclear how the Commission understands – and measures – full “implementation” or “progress”. The Commission further distinguishes between multiple and rather inscrutable levels of progress: no (further) progress, limited (further) progress, some (further) progress, significant (further) progress. This suggests a desperation to find any sort of progress no matter what, with the Commission unwilling to contemplate a situation where a Member State backslides. To make matters worse, the Commission tends to embrace a tick-boxing approach when it comes to identifying progress, and the mere existence of draft laws – or even action plans – has been found to constitute “progress” regardless of whether they enter into force, are not followed up or are offset by parallel developments (a similar criticism has been made by the European Court of Auditors regarding the Commission’s application of rule of law conditionality mechanisms).

Additional problems undermine both the credibility and usefulness of the Commission’s CSRs. In short, it only makes sense to pay attention to implementation/progress data if the CSRs actually address the most important rule of law issues faced by the relevant Member State. And relevant CSRs can only be adopted if the country report itself correctly identifies these issues. At this stage, there is ample evidence of failure on both accounts, with some country reports omitting serious issues, or identifying them without making them the subject of CSRs. In this context, it is worth noting that the Commission has decided not to report on a core rule of law issue: (non)compliance with rule of law-related rulings of the CJEU is neither systematically examined, nor the subject of CSRs.

Bottom line: By equating all issues and treating all CSRs the same regardless of their variable foundational systemic importance, or excluding crucial issues such as (non)compliance with CJEU judgments, one cannot exclude seeing a country report alleging progress across the board in respect of a Member State experiencing a steep autocratisation process. For instance, Greece had a 100% ARoLR progress rate in 2024, with the Commission finding that the country had fully implemented one CSR and made “some further progress” and “some progress” in respect of the other CSRs. A few months later, the V-DEM Institute’s annual democracy report referred to Greece “as a statistically confirmed autocratizer” whose “decline is steep”.

Beyond these damaging aspects for the credibility of the ARoLR as an (alleged) early warning system, the ARoLR has done more harm than good due to a number of persistent flaws the Commission has been unable – or rather unwilling – to remedy over the years.

Doing More Harm Than Good

The Commission’s ARoLR was initially criticised as a toothless tool. Seven editions later, the ARoLR is not merely toothless, it has become harmful as its main impact has been to help disguise or minimise backsliding and normalise the most serious threats to/violations of the rule of law. This follows from a number of defects.

To begin with, the Commission’s ARoLR horizon remains limited to an overview of significant developments – from the Commission’s point of view – over a period of twelve months. Some limited references may be made to previous reports but a major weakness remains: The ARoLR does not outline a country’s rule of law adherence over a sufficiently long period of time. This means that the ARoLR cannot help detect backsliding, not that the Commission seems keen to do so. Presenting it as an early warning mechanism is, therefore, false advertising. At best, the ARoLR amounts to a 12-month stock-taking exercise offering details on developments of variable importance which drown out the big picture.

This defect is compounded by a second one: A selective use of sources so as to avoid engaging with findings contradicting its (rose-tinted glasses) assessments. This has led the Commission not only to ignore the most directly relevant sources regarding autocratisation processes and rule of law trends, but also to disregard inconvenient findings originating from other EU institutions. For instance, the 2026 ARoLR does not refer to the 2025 WJP Rule of Law Index. As a result, the Commission neither addresses the existence of a global rule of law recession, nor the Index’s finding that Hungary and Slovakia were amongst the top rule of law “decliners” in the world in 2024-25. Speaking of Slovakia, the 2026 country report similarly fails to mention the European Parliament’s prior call on the Commission to act in respect of the deteriorating situation in this country.

To make the situation worse, in the name of ensuring “equal treatment for all Member States”, the best and the worst rule of law performers are treated the same. This leads to some odd outcomes. For instance, the EU’s top rule of law performers (Denmark, Finland and Sweden) received a total of 3, 4 and 3 CSRs respectively in July 2026, while the EU’s worst rule of law performers (Greece, Bulgaria and Hungary) received a total of 4, 6 and 5 CSRs. In other words, Finland, a country ranked 3^rd in the world in the WJP Index, has received the same number of CSRs as Greece, a country ranked 48^th and one of the world’s top autocratising countries by magnitude of decline in 2026. To date, the Commission has failed to heed the Parliament’s advice to pay more attention to the Member States “in which the Commission has identified serious and systemic violations, and calls on the Commission to clearly distinguish such cases in its recommendations”. In practice, however, the Commission has never sought to clearly identify serious or systemic violations. In the 2026 transversal report, for example, one may only find a couple of brief references to “serious concerns” and “serious challenges” regarding length of proceedings and undue influence over the media in a couple of countries.

This brings us to yet another flaw reinforcing the minimisation/normalisation effects of the defects outlined above: The ARoLR continues to use the most euphemistic language, even in the face of manifest violations of EU rule of law requirements. It is not unusual, for instance, to see a number of measures (or practices) straightforwardly incompatible with EU law being mentioned in a purely descriptive manner, in addition to not being the subject of any CSRs. In the 2026 ARoLR, for example, the Commission nonchalantly mentioned that in Romania, “judges were recused by the High Court of Cassation and Justice because of their intention to refer questions to the Court of Justice of the European Union”. A non-expert could reasonably conclude that this must be fine considering the lack of any criticism on this basis and more broadly, the Commission’s silence on a widely discussed documentary exposing high-level justice system corruption. Would the Commission offer a clearer, more assertive assessment in a situation where evidence of unlawful and large-scale wiretapping of magistrates has emerged? The answer is negative, with the Commission merely noting in Bulgaria’s country report that a case, “filed by one of the magistrates subject to intimidation”, is now pending before the ECtHR. The Parliament did encourage the Commission “to use clearer language when assessing compliance with the values enshrined in Article 2 TEU”. The 2026 edition of the ARoLR shows this advice is yet to be heeded.

One final, indirect, negative consequence of the ARoLR flaw may be noted. Writing with Professor Bárd in 2022, we warned against the ARoLR ending up “cannibalising the focus, energy and limited resources of the Commission to the detriment” of its enforcement duties as Guardian of the Treaties. Fast forwarding to the situation in 2026, the ARoLR has indeed become a particularly resource-intensive cycle. In addition to the annual production of 32 rule of law reports (1 transversal communication, 27 country chapters for EU Member States, and 4 country chapters for selected EU candidate countries), the Commission undertakes a plethora of country visits (for instance, more than 600 meetings took place between January and March 2026) in addition to gathering a large volume of input from national authorities and stakeholders and participating in multiple debates.

While an increasing number of documents are being produced every year, the Commission has simultaneously failed to deliver on its initial promise, made at the time of the launch of the ARoLR, to “pursue a strategic approach to infringement proceedings related to the rule of law”.  Worse, the Commission has been overselling the ARoLR’s effectiveness and successes in a broader context where political considerations have led the Commission – but also the Council – to underenforce compliance with EU rule of law requirements (and reward cosmetic compliance or promises of compliance on top of this). One may therefore expect the Commission to continue ignoring the recommendations that aim, inter alia, to better align the ARoLR with enforcement mechanisms.

Until then, the ARoLR will continue to be no more than a defective and overrated “stock-taking exercise” doing more harm than good.