Summary

  • BEREC’s draft report finds that only nine of 22 national regulators supplied information on actual wholesale use of state-funded fibre products; five reported dedicated effectiveness monitoring.
  • The figures do not establish that European subsidies are failing. They show why coverage, wholesale access, take-up and retail connections must be measured separately.

A fibre network can satisfy its construction target and still leave the hardest question unanswered: who uses it? That is the problem sitting behind the new draft from the Body of European Regulators for Electronic Communications (BEREC). Adopted for consultation after the regulators’ 1–2 October meeting, the report is open for comments until 13 November.

The rulebook already has a clear starting point. The EU’s broadband State-aid guidelines and General Block Exemption Regulation require subsidised networks to offer wholesale access on fair, non-discriminatory terms. But a published reference offer is a statement of availability. It is not a receipt showing that another provider bought access, reached the handover point and sold service to a household.

An incomplete view of use

Twenty-two national regulatory authorities described the wholesale products that beneficiaries must offer. Nine also had information about actual use. Five reported dedicated mechanisms for tracking whether State aid achieved its purpose. BEREC warns that missing information does not necessarily mean zero use: ministries or funding agencies may hold the data, and an authority may know about bitstream take-up without having records for ducts or dark fibre.

The limited sample still reveals why a single “open access” label is too blunt. Among eight jurisdictions with bitstream-use data, four reported low take-up and four significant take-up. Duct and pole access was generally unused or lightly used in the reporting countries outside France, where use was significant. These are counts for particular products and countries, not a European failure rate. BEREC also cautions that newer networks may need time to attract access seekers.

Poland’s 2025 POPC programme supplies one useful denominator. BEREC’s review of the Polish regulator’s reporting says 52.5% of addresses covered by the scheme had no retail service; 32.5% were served over co-financed infrastructure and 15% over other infrastructure. “Connected” and “connected using the funded asset” are different outcomes. A programme can pass premises while retail service arrives over a competing network—or does not arrive at all.

Spain offers a counterweight, not a rebuttal. The CNMC’s evaluation estimated that aided municipalities had fibre penetration about 10 percentage points above comparable municipalities by 2020, while the leading operator’s share was 7 points lower overall. The regulator says its method estimates an average comparative effect; it cannot isolate the marginal effect of each subsidy or show whether the total aid was optimally sized. The two examples make the same point: outcomes need to be measured, and comparisons need their limits attached.

The handover is an economic boundary

A wholesale product may exist on paper and remain commercially unreachable. BEREC points to the size of funded areas, the number of beneficiaries, network topology and the price of access. A small project area may leave a competing operator with too few customers to justify the backhaul and equipment needed to reach the handover point. Price controls cannot repair every such mismatch.

The report’s own tables also mix programme units: premises, households, population, address points and, in one case, socioeconomic locations. That makes a raw cross-country connection rate misleading unless the denominator and service period travel with it. Regulators need to know which homes were funded, which were passed, which use the subsidised network, which buy a wholesale product, and when each count was taken.

BEREC does not argue that every network should carry every possible access obligation. The practical question is which products match a network’s architecture and the demand of potential access seekers. A leaner obligation can be more useful than a long menu nobody can economically use—but only if subsequent reporting reveals what happened.

The draft is open for consultation through 13 November. The useful response is not another universal access-product checklist. It is a request for comparable, time-stamped evidence from public money to premises served, with the funded network’s contribution kept distinct from service delivered over other infrastructure.

Sources