Reforming Openreach’s governance: Ofcom’s proposals

In my last post, I summarised Ofcom’s initial conclusions from its review of the UK digital communications market. One of Ofcom’s conclusions was that the functional separation of BT’s Openreach division (which provides last mile access)  from the rest of BT, and the associated regulation of the relationship between BT and Openreach (implemented by binding undertakings) was not sufficiently effective in constraining BT’s ability to  discriminate against its competitors. As a result Ofcom’s initial conclusion is that governance should be reformed to give Openreach increased strategic and operational autonomy, and although Ofcom very carefully avoid reaching a firm view, they seem to favour legal separation as a mechanism for reform, with lukewarm endorsement of (just) enhanced governance. If implementation of whatever is decided / agreed is ineffective to address the concerns identified, then Ofcom say that they will only then look at structural separation.

Separation options considered

Whilst many of BT’s competitors had asked Ofcom to consider structural separation of BT, Ofcom’s initial conclusion is that the issues that were raised could potentially be addressed through reform of  Openreach’s governance and that at this stage structural separation would be a disproportionate remedy. In reaching that conclusion Ofcom considered eight alternative structures:

8 models of separation









BT and Openreach currently operate under model 5, functional separation with local incentives. Ofcom’s initial paper discusses the merits of moving to model 6 (reformed governance), 7 (legal separation) and 8 (structural separation) as  ways of addressing the problems they identify.

Structural Separation (model 8)

As well as considering the implications for BT, Ofcom looked at international examples of structural separation including Australia, New Zealand and Singapore. However, all of these were largely driven by a different regulatory imperative – that of ensuring that government investment and subsidy for deployment of broadband networks was made into an entity that was not part of the incumbent. As previously discussed on this blog, the Singapore example serves more to highlight the challenges of achieving true separation and its benefits rather than acting as a blue-print.

Whilst Ofcom’s paper doesn’t rule out structural separation in the future, their decision to look at reformed governance and/or legal separation in the first instance is driven by their desire to be proportionate and choose the least intrusive method of regulation. However, the experience of competition regulators is that detailed ongoing behavioural remedies can be more intrusive than one-off structural remedies, and so the question arises as to whether the same view (which of course will require ongoing regulatory oversight by Ofcom) would be reached if looked at by a competition regulator.

I also had some interesting private comments to my last post, one of which suggested that the real question was whether the boundary between Openreach and BT was in the right place, and that it was perhaps better redrawn with Openreach managing passive assets only – something only touched upon by Ofcom in their consideration of the practicalities of separation.

Enhanced Governance Reform (model 6)

Ofcom flag that one of their main concerns is that “Openreach does not have sufficient strategic and operational autonomy to ensure the equal treatment of all downstream customers.

Ofcom go on to suggest that Openreach’s governance would need, at a minimum to be  to be reformed as follows:

“… solutions to the concerns identified would need to embed further specific behaviours by Openreach in a number of areas, including:

  • More independent governance, with a responsibility to serve all customers equally: Creating an Openreach Board with more independent governance within the current model of functional separation.
  • Increasing Openreach’s autonomy over budget and decision-making: This could address our concerns related to decision-making by giving Openreach increased financial autonomy to take strategic decisions on network investment, network maintenance and operational systems. One way of achieving this would be to increase the delegated authority given to the Openreach Chief Executive from the BT Board to make individual decisions on the allocation and use of funds. One outcome of this increased autonomy could be the ability for Openreach to reach co-investment or risk sharing agreements with operators other than BT.
  • Improving Openreach’s approach to consultation with customers: We want to ensure that Openreach listens and takes into account the views of all its customers in making decisions that could impact downstream operators. Specifically, we could establish obligations for Openreach to consult openly with downstream operators on substantial investment and innovation decisions. For example, commitments to transparency when considering new network investments, consideration of any alternative proposals and consultation at an early stage on any favoured proposals. The EAB, or another independent body, would need to check compliance with such obligations. 
  • Enhancing Openreach’s operational capability: Giving Openreach the ability to draw upon dedicated support services could address our concerns over its operational ability to deliver its priorities. This would ensure Openreach has sufficient internal capability to manage both its strategy and manage external supply arrangements. Such changes would have to be weighed up against any potential loss in efficiency or scale benefits. “

Ofcom then observes that these measures may be “insufficient on their own“. This is quite an interesting formulation from Ofcom – if they don’t think they are sufficient, then why raise as an option? Ofcom then go on (in a somewhat tentative way) to suggest that a stronger set of governance reforms would meet their concerns.

Legal separation of Openreach (model 7)

Ofcom then go on to suggest what appears to be their favoured option, the hive-down of the Openreach business to a wholly owned subsidiary within the BT group i.e. legal separation of Openreach, with the retention of economic ownership by BT, and restrictions on the influence and control by BT of Openreach. Ofcom summarise it as follows:

  • “Separate Openreach Board: Openreach would become a wholly owned subsidiary of BT Group, making the relationship between it and BT’s other divisions more transparent. The Openreach Board would be separate of the wider group and hold executive powers of decision-making over Openreach’s activities in the interests of the legally separate Openreach, rather than the wider interests of BT Group.
  • Responsibility to serve all customers equally: An explicit responsibility for Openreach to treat all downstream customers equally might be established through the objects and purposes of the company in its articles of association.
  • Autonomy over investments and decision-making: The new Openreach Board might be given more autonomy over capital investments and the broader use of cash within the business.
  • Ability to raise funds: There are different options for how a wholly owned subsidiary could raise funds. There may also be appropriate ways for BT Group to finance Openreach without directly influencing how the funds are spent. Alternatively, Openreach could raise funds directly from the market or fund network investments through contributions from downstream providers, secured by contract.
  • Statutory accounts: An important aspect of the model is that BT Group shareholders would retain full ownership of Openreach and continue to benefit from any associated profits. As a legally separate subsidiary Openreach would be required to file full statutory accounts, including a separate balance sheet, profit and loss statement, and cash flow. This would improve transparency of cost and asset allocations.”

Whilst Ofcom recognise that this approach creates some implementation challenges it remains to be seen if they have fully appreciated the costs and complexity of this option, which involves many of the downsides of structural separation, but without removing the incentive to discriminate, which means that it will still require ongoing intrusive behavioural regulation.