Nearly two years after Mario Draghi handed Brussels a 383-item list of what Europe needed to do to stop falling behind, the question is no longer whether the diagnosis was right. It is whether the EU's legislative machine can deliver a list that long at all.
EPIC's Draghi Implementation Index, run by the European Policy Innovation Council at draghiwatch.eu, has been keeping score since September 2025. It counts how many of the report's recommendations have become binding EU law — not announcements, not strategies, not roadmaps. Its July 2026 preliminary update puts strict implementation at 15.7%: 60 of 383. Add partial implementation and the figure is 41.3%, or 158. In the January 2026 interim audit those numbers were 15.1% (58) and 38.9% (149).
Six months of legislating produced two extra fully implemented recommendations. That is the finding worth sitting with. The index assessed 34 legislative acts between February and June 2026, and the movers were real: the €90 billion Ukraine defence loan, the simplification of sustainability reporting, the Russian gas phase-out, the EU Talent Pool, rail capacity rules, Water Framework Directive flexibility, Global Gateway financing. None of it was trivial. It still moved the combined score by 2.4 points, against 7.5 points in the September-to-January stretch.
Deceleration after a fast start is not automatically a failure. The easy items go first, and the low-hanging recommendations in any 383-point list tend to be the ones that need a regulation rather than a treaty argument. But the slowdown lands at an awkward moment, with the report's one-year formal review due in September 2026 and the political case for the whole exercise resting on visible movement.
The July update also produced the index's first ranking by directorate-general, and the spread is the most quietly damning thing in it. DG TRADE sits top at 41.7% strict implementation. DG ENER is at 2.7%. DG EMPL is at 0.0%.
That is not a story about lazy departments. It is a story about which recommendations are politically cheap. Trade policy is an exclusive EU competence, executed by the Commission, largely insulated from national vetoes. Energy and employment run straight into member-state control over energy mixes, taxation and labour markets. The pattern EPIC has flagged repeatedly holds: the EU moves fastest when competitiveness wears a security badge — defence implementation jumped from 35.7% to 78.6% in the January audit — and slowest on the structural reforms that would force market outcomes.
EPIC's proposed remedy is institutional rather than technical: a European Competitiveness Act, a single delivery package modelled on Fit-for-55, which bundled climate legislation into one negotiable block with one deadline and one political price.
What the Commission has done instead is spread the work. The Competitiveness Compass, its 2025 roadmap, envisages up to 47 legislative and non-legislative proposals by the end of 2026. The 2026 work programme carries 70 major new initiatives, 44% of them under the competitiveness heading. The Commission also runs its own competitiveness implementation tracker, built on the 2025 and 2026 work programmes, listing what has been adopted and what is still in preparation.
The flagship of that approach arrived on 4 March 2026: the Industrial Accelerator Act, COM(2026) 100 final, now moving through Parliament and Council as procedure 2026/0068(COD). According to the European Parliamentary Research Service briefing by Agnieszka Widuto, published on 21 May 2026, the act would set a target of raising industrial manufacturing to 20% of EU GDP by 2035, from 14.3% in 2024. It covers energy-intensive industries, net-zero technologies and the automotive sector; introduces “Made in EU” and low-carbon preferences in public procurement; sets conditions on foreign direct investment in strategic sectors; creates industrial acceleration areas; and simplifies permitting.
It is a substantial piece of law. It is also, on EPIC's counting method, one file among dozens — and it will not register on the index until it is adopted.
The EU now has two competitiveness scoreboards running side by side, and they are not measuring the same thing. The Commission's tracker asks whether the Commission has done what the Commission said it would do. EPIC's index asks whether what Draghi recommended has become law. The first can look healthy while the second stalls, because tabling a proposal is an act the Commission controls and adopting one is not.
That gap is the argument for a single act in a sentence. Seventy initiatives means seventy negotiations, seventy chances for a capital to extract a concession or run out the clock. One package means one fight, at a level where heads of government have to show up and settle it.
The counter-case is serious. Fit-for-55 took three years and produced its own paralysis; bundling raises the stakes on every individual file and hands any blocking minority a bigger hostage. Nobody in Brussels wants to relive it, which is a large part of why the competitiveness agenda was deliberately built as a stream of separate proposals rather than one block. But that design choice has a cost, and the index is now measuring it: after nearly two years, five of every six Draghi recommendations are not law, and the rate at which that changes is falling. The September review will produce fresher numbers. The more useful thing to watch is whether the Industrial Accelerator Act clears its first reading intact — because if the single biggest competitiveness file of this mandate gets negotiated down over a year, the case for doing it all in one package gets considerably stronger.
