Written evidence submitted by Mike Kiely


Thank you for the opportunity to submit evidence to this recall session on Broadband.  

My name is Mike Kiely.  I provided written evidence to the inquiry in November 2020[1]. That evidence remains valid. This short submission provides an update on the status of BT Capital Deferral intended to complete rural upgrades. There is an opportunity to complete rural broadband and achieve a greater than 99% coverage provided the monies owed by BT (£568m untriggered clawback) are re-invested as was the original plan confirmed to this Committee several times. The first re-investment promise was made in the PAC 2013 inquiry on rural Broadband. There is an immediate danger that this long-standing promise is to be reneged upon and the opportunity for an exceptional but late outcome is lost.  This will mean England and Wales will have less complete rural Broadband coverage than that already contracted in Northern Ireland and Scotland.

Why am I responding?

There is a danger that the hard work done by three[2] Select Committees over the last 9 years will be undermined if the current rural Broadband programme is brought to a close where BT’s accounting treatment of at least £558m of untriggered clawback is not used to complete the rural broadband upgrades.  The failure to complete is coming at a cost where BT has already capitalised the profit share it owes and including that profit share in cost recovery processes for work which may now never be completed.

This inquiry marks some nine years of attempts by three select committees, PAC, DCMS and EFRA to squeeze out the very high levels of normalised deviancy imposed on the original 2012 BDUK Framework by BT and reported upon by this committee during its initial July 2013 hearings.  Excess costs in BT bids of 38% were eventually confirmed by the NAO in 2015, and the BT gainshare was reported by the NAO in 2019 as £900m.  BT’s actual capital contributions have never been reported upon and much of the sub-optimality in planning less full fibre is a product of gaming this capital liability.  The recent NI Audit on Broadband could not confirm value for money and failed to determine what BT had or had not invested.[3]

The success of getting to 97% coverage with a further 1% under contract yet to be delivered needs to be celebrated.  If the information outlined in this submission is used, then greater than 99% coverage can be achieved using money already capitalised in BT’s accounts.  However, there is a real danger the 99%+ will be missed if the ‘gigabit-broadband (Boris Broadband) is pursued at the expense of completing the rural work.  Despite the oral promises to this committee the Departmental consultations on Gigabit Broadband do not prioritise those rural premises so far excluded.  The term Boris Broadband is used not in an unkind way but more to highlight that detailed rationale, reasoning and business case for an election whim has not yet to made for this proposed intervention.

Completing Rural Upgrades versus Gigabit-Broadband (Boris Broadband)

There is a distinct, once in a generation opportunity to complete rural broadband upgrades using the monies owed and already capitalised in BT’s accounts[4].  If pursued, PAC can demand DCMS use the monies owed to achieve 99%+ coverage where 1.5m[5] of the most rural customers can order full fibre services. All other customers can already order 30-80Mbps services. This can be achieved if monies owed in the form of BT’s Capital Deferral is converted into coverage by contracting for a further ~300,000 premises in rural England and a further 30,000 in Wales.  Scotland and Northern Ireland have contractual cover for 99%+ coverage.

The PAC Inquiry of November 2020 confused, I think, two very different topics and these need to be treated separately.  The quest to complete rural as above, needs to kept separate from the election promise of ‘Gigabit Broadband’.  The Committee might be better calling this latter proposal ‘Boris Broadband’ at least until there is a Treasury approved business case.  The latter is highly problematic as it involves re-subsidising and overbuilding the already subsidised 30-80Mbps services to 5m premises. 

Overbuilding with new subsidy already subsidised services is problematic for several reasons.  Demand for >100Mbps services is proven where no network upgrades have occurred.   Demand for full-fibre where 30-80Mbps services are available is not proven.  For example, subscribers for BT’s G.Fast (up to 300mbps service) has so far only achieved 6.8%[6] of the reported premises passed.  G.Fast is rolled out in the same locations as commercial FTTC-30-80Mbps services. The benefits accruing to the public purse to re-subsidise those on 30-80Mbps to take a greater than 100Mbps service is not clear, neither is the customer demand. The rationale has not been outlined in any of the many consultations DCMS have run so far. The benefits where they accrue will be in the form of cost savings to the network operator.  Unless the incremental benefits to the taxpayer are clear why re-subsidise that which has so recently been subsidised?  Furthermore, BT have committed to upgrading 6.2m rural premises to full fibre, where the EFRA select committee[7] confirm there are only 6m premises in their rural and very rural classifications. Furthermore, BT have confirmed to their shareholders that cost of the rural roll out is no different to the commercial rollout[8].  So why subsidise or even re-subsidise? Finally, the DCMS Gigabit consultation process for individual areas does not prioritise the completion of premises so far excluded and outstanding from the contracts originating in 2012.  Accommodating an unsubstantiated election promise of £5bn is in danger of being executed at the expense of completing rural Broadband upgrades.

Current Status of BT’s capitalised gainshare or profit share.

The NAO in 2019 referenced £900m of clawback.  BT’s published accounts had historically referenced a total of £825m where a further £37m was added in 2020/21.

In August 2021 BT investor relations confirmed the following.

‘The £558m refers to the existing untriggered Gainshare provision, IE it has not been through a formal take up review as yet, but based on the take up projections is the amount of future Gainshare that would be generated.

‘The £267m (£825m less £558m) is triggered Gainshare provision, some of which has been handed back or utilised for additional build , and a portion of which is being held in escrow until the local authorities determine how they wish to utilise it (be that refunded back or reinvested for additional superfast build roll out).

It is fair to conclude that there is at least £558m available plus the £37m added in 2020/21 already capitalised in BT’s accounts and being recovered as if the money has been spent on the rural network.  But the work is outstanding and yet to contracted.

The accounting treatment that leads to this is of a similar ethical nature to those practices identified in Pandora papers but BT will have some discretion on these matters. Ofcom has been notified several times, including in a recent response to the B-USO inquiry but Ofcom to date have wished to avoid the matter.  The notion that monies set aside for rural are capitalised for years[9] and used in cost recovery before the decision to invest the funds in rural is another layer of accepted normalised deviancy in a long running process.  It is evidence of long term state capture in a sub-optimal procurement process where commercial confidentiality was used to hide the embedded original normalised deviancy of heavily inflated costs in a gap funding model for subsidy.  It has taken 9 years of inquiries and audits to unpick the initial reported wrongdoing.  Additional steps are needed to make the most of the progress made.

In early September 2021 the matter was made worse when a member of the Ministerial team wrote on the 6th of September the following in response to a query of on the untriggered clawback.

Clawback yet to be realised from Openreach contracts cannot be used to fund these new gigabit-capable infrastructure contracts with other suppliers, hence the need for new funding under a new programme with new contracts.’

I have queried why such funds cannot be used, given the re-allocation of clawback funds featured in EFRA select committees in 2017 and 2019 and where the DUP/Conservative election agreement included £150m for Broadband. The latter originated request in a discussion on re-allocating clawback[10] at 2017 EFRA select committee meeting.  The response received fails to address the point raised.

It would seem appropriate if the PAC committee is minded to ask the following questions?

  1. Why can untriggered Clawback gainshare funds (£558m) not be used to conclude rural upgrades?
  2. Can BDUK confirm BT’s total capital contribution to date to allowable costs annually since 2013?  This is the amount paid, not contracted, presumed, incurred, owed.
  3. How does clawback vary where BT’s capital was not subsequently used because of the presence of inflated costs in BT’s models?
  4. Can you confirm the number of premises under contract but yet to be delivered?  (>~420k premises[11])
  5. Can you confirm a monthly delivery run rate for premises under contract but outstanding?  (current estimate is 2-3k premises a month reported by Thinkbroadband[12] – suggesting Superfast work will continue at a slow pace for another decade.  This can be challenged and  some number of ~10k premises a month delivery agreed.)
  6. Can you confirm the number of incremental premises that need to be contracted to meet 99% and 99.5% coverage where full fibre is used? (Estimate of ~300k -400k premises in England including 30k more premises in Wales.  Scotland and NI are already fully contracted.)

In relation to Gigabit Broadband (Boris Broadband) it may be necessary to get additional assurances that the completion of rural broadband is not sacrificed while the proposals for Gigabit Broadband continue to be developed.

It is not unreasonable for the Committee to seek detail on the rationale that justifies re-subsidising those who can access 30-80Mbps services, where the incremental demand for >100Mbps service is not clear and the incremental benefit to the taxpayer of subsidising this incremental upgrade of £5bn is equally unclear.

End. Mike Kiely October 2020.

[1] https://committees.parliament.uk/writtenevidence/13108/pdf/

[2] PAV,  DCMS and EFRA select committees.

[3] https://www.niauditoffice.gov.uk/publications/broadband-investment-northern-ireland

[4] Footnote 18 as per BT published accounts 2020/21 available on BTplc website.

[5] The authors estimate arising from 580k full fibre already subsidised, ~430k premises under contract and yet to be delivered and the opportunity to contract another ~400k premises using the funds already capitalised for that purpose in BT’s accounts.

[6] BT reported KPI’s for q1 2021/22 reported G.Fast passing 2.823m premises but only 192,000 subscribers.

[7] EFRA Select Committee report 2019, 6m includes Scotland and NI,  6 classifications from ‘Rural Hamlet and isolated dwellings (sparse setting)’ to ‘Rural Town and fringe’ in England Wales.  Separate classifications for NI and Scotland used in EFRA report and reflected in the 6m number.

[8] The presentation can be on BT website. The confirmation of the cost is on page 5 of BT investor Bulletin for Q1 and confirms the rollout of 6.2m rollout to be in the range of £300-£400. Text available.

[9] BT first reported clawback as a capital Deferral in 2015 in their accounts.

[10] BT and DCMS avoided answering questions on the re-use of clawback during two EFRA hearings e.g. Ritchie q.427 Oral evidence: Rural Tourism (including Broadband), HC 428 Wednesday 29 March 2017 Ordered by the House of Commons to be published on Wednesday 29 March 2017.  The inclusion of the Broadband in the DUP/Conservatives election followed these attempts.

[11] Calculated from the last BDUK report – https://docs.google.com/spreadsheets/d/1Hs00bNsyRV1WoOt-fow3rsNXzpcKg26AsOWvk1bvJRk/edit#gid=1411146266 compare contracted 5.76m premises and delivered 5.34m premises.

[12] https://www.thinkbroadband.com/news/9072-openreach-reaches-5-million-premises-passed-in-our-tracking-of-roll-out - Here Thinkbroadband reported 2,500 additional properties completed in the month using BDUK subsidy.