Rural Broadband, National Audit Office (NAO) find 38% excess costs in BT’s financial models. Broadband Delivery UK concedes it has yet to find the most economic price!
Updated with an introduction on 11/2/2015, and further reason why the excess modelled cost
are likely to grow, and a summary statement.
This is note is intended for those with an interest in the
UK rural broadband programme,
particularly those interested in getting the most from the £1.7bn of
public monies invested. It does not
question the contract letting process or the awarding of the contract to
BT. It comments on the findings of the
two NAO reports done to date and suggests the next steps needed to ensure the
public monies get invested where they were intended. Feel free to use the information to further
the public interest. I would welcome comments by email to Mikekiely01@btconnect.com. I would be delighted to correct any errors
and improve the public record where evidence emerges to support any substantive changes.
This note alleges and attempts to prove that the £142m
excess modelled costs identified by the NAO to September 2014 have already been
paid to BT by Local Authorities using the BDUK Framework negotiated just before
the 2012 Olympics in the form of invoices for milestone payments. The invoices
for milestones payments expressed in terms of premises past are largely
unrelated to the actual costs for the components used now being reviewed by
Atkins/BDUK. The excess modelled costs
are likely to increase by an estimated £40m a quarter for the next 4 to 5
quarters and rest in BT accounts until plans for increased coverage in rural
areas can be revised well after 2017.
On 28th January 2015, the NAO issued a second
report to the Public Accounts Committee on the costs of the £1.7bn UK Rural
Phase 1 of the project provides £1.2bn of state aid to gap fund BT’s fibre to
the cabinet (FTTC) overlay network to some 5m premises in some 30,000 communities
in areas not covered by BT’s commercial investment. The Rural Broadband
programme is a 4 year programme where every week a further 40,000 premises can
connect to some 160-200 new green cabinets using 200-300km of fibre to connect to BT’s high
capacity core network.
The information provided by BT and BDUK to inform the NAO’s
first report in July 2013 looked and read like a joint plausibility exercise.
The NAO found it wanting, and the Public Accounts Committee meetings called to
discuss the findings were even more critical.
I should declare an interest; see;: http://www.theguardian.com/technology/2012/oct/03/whistleblower-sacked-bt-broadband.
So it was with some relief that the NAO’s second report, a year and half later (Jan
2015), found evidence that BT had inflated its cost models by 38%, and BDUK,
the government agency, is even now not
convinced it has BT’s best price.
This is likely to be the start of the difficult task of proving that BT has abused
its monopoly position, and that it has hidden the details of the rural
broadband project behind 44 separate and pernicious commercial confidentiality
agreements with local county councils.
The first NAO report in 2013 found subsidies that averaged some
£46k per VDSL cabinet and fibre path. The
subsidy, in the form of milestone payments, is much higher in Wales, in fact
more than £60k for each VDSL cabinet. The Welsh audit office is reporting separately
on this in April 2015.
These estimates included premiums for the universal service
commitment of 2Mbps. BT’s quarterly results
have reported state aid of £94m for each of the last two quarters. The latter
is consistent with milestone payments, driven by inflated cost models, of c£200
per premise passed. The NAO in its
latest report have identified actual costs closer to £104 per premise passed
and an emerging cabinet and fibre path costing an average of £21,000. This number in my opinion excludes the 23%
contribution from BT so the ‘should’ cost to the taxpayer ought to reduce to
£16,170 which correlates with the The
Bit Commons subsidy estimates presented to NAO for its first report. These were dismissed by DCMS but thankfully
not by the NAO.
The second report noted that these are the easy to do cabinets
and it will get more expensive. There
are at least four counters to this belief.
BT’s network is homogeneous to about 90% coverage at a
national level, so economies of scale will continue for some time. Therefore, the £142m the NAO identified as excess
modelled costs, which are included in the milestone payments on a job that is
26% complete, are, in my opinion, likely to grow at the rate of £40m a quarter
for another four to five quarters.
The second, more sensitive point is that BT was to provide
23% of the capital cost directly; this has not been mentioned in the NAO
comparisons between modelled costs and actuals.
BT’s contribution is self-certified, so this will need to be examined or
its status made clearer in the next NAO update. BT’s contribution could also be addressed in
the review of the state aid measure which takes place before June 2015. Furthermore, it could also be addressed in Ofcom’s
financial regulatory reporting consultation, should Ofcom wish to assist the
NAO, PAC, EU Commission and BDUK in calling BT to account.
Thirdly, BT did an excellent job in Northern Ireland, where the
subsidies amounted to about £12,500 per cabinet to reach some 88% of the premises. This included Belfast, but 70% of the money went
to areas outside Greater Belfast.
The fourth reason is the costing study is likely to have
been done in a rural area, long fibre spines and a low cabinet count. Why pick a costing exercise for a rural
programme in an urban area?
BT is not yet acknowledging that the milestone payments have
become divorced from the actual costs. BT is not acknowledging that the
reported state aid receipts cannot be based on actual costs because the
invoices become available only months after the milestone is achieved.
The accumulated excess modelled costs, sometimes referred to as ‘savings’, are resting in BT’s accounts and are likely to
do so through to 2020 as the BT run project is resource-constrained. Establishing a simple level of transparency
on what has been paid in exchange for what, is not helped by BDUK, which does not report on
the total subsidy paid, but only on central
government’s contribution to the subsidy. It neglects to include and report funding
from the local authority and EU. Perhaps
this will now change.
BDUK has added an extra layer of complexity by deciding to
contract for additional coverage without first reviewing the actual costs and BT
contributions made in phase 1.
The basic deliverables of the project read as follows, circa 160-200 cabinets installed a week,
200-300km of fibre laid each week, passing some 40,000 premises. You could call it good, but the public statements
are so misleading that one cannot tell how the 40,000 premises are counted or
how many will receive a more than 24Mbps service or 15Mbps, depending on the
There will also be gaps in coverage, whether areas outside
the reach of a cabinet, or parts of exchange areas which have not been
completed. Furthermore the funding available (and this is reflected in how BT
chose to present its financial models), suggests it could have employed another
1,500 apprentices when the Framework procurement agreement was signed in July
2012. The cost of employing these
apprentices would have been easily covered by the 38% excess costs identified by
NAO, and would have enabled a much more extensive delivery plan.
From a BT shareholder perspective, a deeper, more intense
fibre rollout would assist in long term cost removal. Delivering to the rural
economy the upgrade it needs would allow the government, central and local, justifiably to claim a ‘best in Europe’ for
this vital piece of infrastructure. It
would also create the environment for additional public investment provided
there was confidence in the cost models and capital contributions made. Trust, the basis of any functioning
relationship, will now be in short supply.
As infrastructure projects go this is technically as easy as
it gets. There is no decade of planning or consultation, no compensation and a
win-win for both the economy and BT. Or it would be if BT’s cost models were
not inflated to absorb the money available,
while failing to resource the project to match the funding being billed.
The overall numbers are easy enough to understand as the BT
network is a tree and branch architecture, with published quantities for
exchanges, cabinets, and handover points.
According to its guidance to analysts, BT has invested between £1.3 to
£1.6bn out of the £2.5bn BT thought it needed to upgrade some 19m premises in
its commercial roll-out. The £2.5bn was
not new money but an allocation from BT’s existing capital envelopes used to
maintain and extend the existing telephone network.
Dividing £1.3bn by 19 million premises passed gives you £68
invested per premise. This is an
interesting number; the Connect America rural broadband fund expects US operators
to pay the first $70 (£45.90) per premise passed before subsidies kick in. There is nothing in the second NAO report to show
that BT is paying its capital contribution up front, or how it is being accounted
for in the costs presented.
Estimating an average cost for a cabinet and fibre path is
not too difficult: BT has 90,000
cabinets (PCPs), of which it is reasonable to assume that statistically 50-55k
cabinets serve 19 million premises, leaving 25-30k cabinets to reach the more
remote 5m premises, helped by public money.
BT’s commercial cost
is thus £26k - £30k per cabinet (£1.3bn
/55-55k cabinets), fibre, handover point, core network and all the development
costs including operational support systems.
The Phase 1 publicly
funded £1.2bn BDUK project has a combined intervention areas of
approximately circa 5m premises and needing 25-30,000 cabinets. The NAO’s first
report, which relied on evidence from BT and BDUK, found the average subsidy
was estimated at £40-£48k per community serviced – of which £28.9k was for the actual
cabinet and fibre path. The £28.9k
average is important as it is a number BT acknowledged at PAC, while failing to acknowledge others, but the report (Table 11, page 33) shows it
only represents 36% of the total costs or c60% of the public subsidy.
The second NAO report reports the actual costs running at
£21k (excluding BT’s contribution in my opinion), not the £40-£48k subsidy budgeted.
But it looks from BT’s reported state aid receipts that BT is receiving the
higher number. Presumably this will be reconciled with the lower actuals costs
at some later date, provided local
authorities demand and pursue it. This
is separate and in addition to the clawback mechanism in their contracts and
should be pursued immediately.
All of this suggests the third NAO report will need to focus
on the following;
Verification of BT’s direct investment of £70 a premise or
£357m for Phase 1
Identifying and managing the excess costs built into the
existing milestone payment process,
Review of how the excess costs have been converted into
Estimating clawback amounts arising from take-up and
reinvesting the clawback.
Reporting on the use of the premiums charged for the
Universal Service Commitment or its conversion into additional coverage.
The third report will also need to confirm that the government
is actually gap-funding BT in a manner that is consistent with state aid
measures, in other words, not close to paying for the project in its entirety.
What a third report will not do is judge whether, given the financial resources the government
and local authorities have made available, BT is applying them effectively and whether BT’s
ambition is adequate. With £1.7bn in
taxpayers’ money, why will Britons not, as promised originally, have the best
broadband in Europe, rather than be only marginally better off than the French
or the Germans? Just a little
Churchillian drive would make a huge difference.
When this process began BT had a plan to make fibre to the
premise available to a substantial proportion of the population. BT decided to focus on cabinet based solution
which by their design leaves many excluded.
The NAO and BDUK have found 38% excess costs in BT financial
models. These monies now need converting
into a more ambitious plan which includes more direct
fibre rich access. Public monies should not be allowed to rest
as some huge contingency, while even more money is contracted to BT. BT should not be permitted to blag on this
matter but called to account using the combined pressure of our public
institutions working together to deliver an entirely achievable goal.
In the autumn of 2013 the then CEO of Openreach repeated on
BBC radio (Strike up Broadband, Dec
2013) that a FTTC cabinet cost £100,000
each. The July 2013 NAO report found BT
and BDUK supporting numbers of £80,000 per cabinet and fibre path which
included a public subsidy averaging £46,000 and an average cabinet and fibre
cost of £28.9k. The January 2015 NAO
report has found actual costs of £21,000 a cabinet and fibre path with BDUK
convinced it has yet to reach the most economic model. I am equally convinced the £21,000 excludes
any direct contribution from BT and so this number should drop to circa
£16,000. The third report should it be completed will find more of the modelled
excess costs. The pretence, blagging and
bluster is preventing a much deeper delivery of fibre rich access solutions.
With the second report, NAO and BDUK now agree and have
sight of the excess costs in BT’s financial models. They now need support from Ofcom, through the
financial regulatory reporting process and the EU Commission, through a revised
state aid measure to ensure the identified excess costs already paid of £142m,
and those that are likely to accumulate in the next 5 quarters at an
approximate rate of £40m a quarter actually reach the rural economy and are not
held up in BT’s accounts.
Finally, special thanks to those who peer reviewed this
note. I hope it is useful to anyone
contributing to the rural broadband programme,
especially those trying to manage the programme day to day. This includes many BT folk who make
informative material available online.