by Sam McBride
Hepper later attempted to explain the sentence as meaning that a Northern Ireland RHI was the best option when compared to the alternative of simply replicating the GB RHI. But if that was her intent, her choice of language was grossly inadequate and misleading. The natural meaning of the words represented a far more sweeping statement – and that is how Foster said she read them. But Hepper’s submission to Foster was not just inaccurate – it was unusual in that it did not give the minister a recommendation. The various options were set out without a clear preference from the team of civil servants, who had been working on the policy for more than a year. A few days later, Hepper met Foster and Crawford in person to discuss the submission. It was at or immediately after this meeting, on 14 June, that the minister would decide to press ahead with the more expensive RHI scheme.
Hepper later told the inquiry that there had been ‘a fulsome discussion around the detail of the figure-work and the two different options’. Foster disputed that that was likely to have been the case – but said that she had no clear recollection of the discussion. Foster’s limited recall of events would prove to be a recurrent feature of her evidence to the inquiry. Foster said that going against her officials’ advice ‘would not have been my natural course’, and so she is likely to have gone along with at least a verbal steer from Hepper, even though the official’s submission had not made a recommendation. However, what transpired in that meeting will forever remain elusive because no records were kept of what was said. That would be particularly suspicious were it not for the fact that under the DUP and Sinn Féin Stormont had developed a widespread practice of deliberately not recording many meetings – and sometimes even important multi-million-pound decisions.
The man who was then the most senior civil servant in Foster’s department, David Sterling, told the inquiry that officials had consciously not recorded many internal discussions which ought to have been minuted because the DUP and Sinn Féin were ‘sensitive to criticism’. Sterling said that as a result of the desire for the public not to hear about what was going on it had become ‘common’ not to keep records because ‘it is safer sometimes not to have a record that might, for example, be released under Freedom of Information, which shows that things that might have been considered unpopular were being considered’.
In acquiescing in politicians’ desire for secrecy, civil servants broke their own rules which were clear that significant internal or external meetings must be minuted. In response to Sterling’s comments, both the DUP and Sinn Féin insisted that they had never asked for records not to be kept, something Sterling accepted. That suggested that canny officials did not need to be ordered to break their own rules but instead willingly did so after picking up on the desires of ministers and their advisers.
By the time the final CEPA report arrived on 28 June, the costs of RHI had risen further. Hepper admitted that the minister had not been told that the costs had changed after she made her decision, either evidence of sloppiness or an implied understanding from the officials that a leap in costs was not the sort of issue which would cause themselves or the minister to re-think – perhaps because of the belief, spoken or unspoken, that as much Treasury money should be secured for Northern Ireland. But two other critical errors had been made, one of which would wrongly lead at least one senior DUP figure to later work to keep RHI open when it was out of control, based on an erroneous belief that London was paying the full bill.
CHAPTER 4
TWO FATAL ERRORS
When Fiona Hepper’s June 2011 submission went to Arlene Foster, it contained three letters which would have meant nothing to most people – but to some within Stormont it made their eyes light up. The letters were AME, and to some politicians and civil servants they signalled that the funding was effectively ‘free money’.
As with all ministerial submissions, the document followed an established format whereby Hepper and her team filled in answers to a series of questions, one of which was on the nature of the funding. In that section, Hepper simply said it was ‘AME’ – the acronym for Annually Managed Expenditure, which mandarins and politicians pronounced ‘Amy’.
What Hepper had sent to Foster was a significantly inaccurate simplification of the true nature of how the scheme was to be paid for and one which years into the future would haunt Foster and her spad, Andrew Crawford. Two months earlier one of Hepper’s subordinates, Alison Clydesdale, had established that this was far from the full story and there was a major risk to the department if it spent beyond what the Treasury had allocated. Clydesdale had sought to clarify how the funding arrangements would work and had been told that the money would be coming directly to Stormont under AME arrangements. That is the form of spending used for government schemes where a budget is difficult or impossible to predict, such as for welfare claims.
Most devolved government spending is very different and involves the block grant – the annual lump sum the Treasury sends to Stormont – being divided up into multiple finite budgets, department by department, project by project. Each department is headed by a permanent secretary whose formal title of ‘accounting officer’ alludes to the significance of accounting for the department’s spending. It is a major crisis if a department spends beyond the budget democratically authorised for it by Parliament or the Assembly. But the amount required for demand-led expenditure, such as pensions, student loans or unemployment benefits, is difficult to predict, and there is a principle that the money should be paid if the individual meets the criteria set by Parliament. Therefore the funding is ‘managed’ and predictions are made as to the likely demand, but no budget is set.
Where RHI would break with long-established public sector accounting principles was that it would be described as AME – that is, demand-led and without a budget – but it would also have a budget. It would be on that rock of confusion – allied with the cynicism of some senior figures who saw the chance to squeeze more money out of the Treasury – that Stormont would perish.
In April 2011, Clydesdale spoke by phone to Jon Parker, the joint head of the Treasury’s energy branch. She followed that up with an email in which she asked him to set out in writing how the department could commit to 20-year payments based on funding, which was only for four years, as well as a request to set out the practicalities around how the money would come to DETI. Parker made clear that any commitments entered into during the initial years of the scheme would be honoured over a 20-year period and that Stormont’s funding would be based on a share – roughly proportionate to Northern Ireland’s size – of the GB scheme, with 2.98% of the GB RHI budget being made available to Stormont. He then went on to set out a critical warning, which ought to have disabused anyone in Stormont who read it, of the notion that RHI was a bottomless cash pit. Parker said:
The other key point it is necessary to let you know about is that the [Whitehall department] DECC RHI spending is not being treated as standard AME, where the Exchequer takes on all risks of overspend. Instead, there is a risk-sharing arrangement whereby should RHI spending in one year exceed the SR [spending review] profile, then DECC would need to repay this in future years.
Parker went on to say that not only would any overspend have to be recouped in future years – something which assumed the overspend was small enough to be recovered from the budget remaining for future years – but that on top of that there would be a penalty ‘likely to be of the order of 5%’ taken out of the department’s budget if it overspent. He said that ‘again, these rules would be applied in equivalent fashion to NI’.
The following month, Clydesdale received a second email warning in clear terms that the funding was not unlimited. Bernie Brankin, an experienced official in DETI’s finance division, told her that she had spoken to Stormont’s Department of Finance and ‘RHI spending is not being treated as standard AME’. Instead, she said, it was being treated as if it was allocated to the department out of the block grant, with a strict budget. She explained: ‘If you underspe
nd in any year, that part of your budget is lost to the department and, if you overspend in any year, DETI’s budget will be reduced by the amount of overspend in future years.’ The email – which referred to a verbal discussion between the two officials – went on to say: ‘You will need to take this treatment of AME into consideration when drawing up your proposals on how you will spend this allocation …’ Crucially, the email was copied to Hepper.
Hepper was copied into another email later that day from Clydesdale in which she said to another DETI official that she had spoken ‘at length’ to Brankin about the situation and that it ‘presents a significant challenge’. Stormont’s finance department would require evidence of DETI’s ability to control the number of RHI applications. Perhaps influenced by the problems with the recent Reconnect grant scheme, she said that grants would be ‘the riskiest route financially as it is hard to control the number of applications especially at the end of the programme’. Clydesdale went on to say that CEPA would have to be asked ‘to factor this in as a risk factor’ because ‘I didn’t see any evidence of this in the draft that we have already received … it will need to be fully addressed’.
Almost seven years later, Clydesdale told the public inquiry that in May 2011 ‘I recognised the need for cost controls to control over/under spending and the associated risk’. That claim is supported by her emails from the time. Yet almost immediately after her engagement with Parker, Clydesdale was moved to another role as part of the constant churn of civil servants moving around the system.
The following month, just days before Hepper’s crucial June meeting with Foster, another Whitehall civil servant, Akhil Patel in DECC, the department running the GB scheme, gave DETI another warning about overspending. Patel emailed Peter Hutchinson – the main individual under Hepper who was working on designing RHI – and told him explicitly that if they overspent, the money had to come out of DETI’s budget. If that wasn’t sufficiently explicit, Patel added: ‘Clearly, this represents a large financial risk on the department so the policy team is currently looking to develop a system of tariff degressions [further cost controls] … to ensure (among other things) that we manage the risk of overspending against our budget.’ Essentially, Patel was telling DETI that although the GB scheme had been set up with a basic cost control of tiering, that was not enough, and so it was looking to add an additional cost control of degression.
Tiering was a simple mechanism whereby there were two tariffs which every boiler owner could claim. The first high tariff was only available for 1,314 hours each year – allowing high payments for 15% of the year. After that, the second-tier tariff was all that could be claimed and it was far less lucrative. The intention was to encourage energy efficiency, discourage fraud and shield taxpayers from having to pay huge sums to those who legitimately needed to run their boilers for most of the year.
Patel’s message did nothing to prompt a sudden change of course within DETI. Days later, Hepper, who was aware of the nature of the funding, gave Foster the first of many submissions which simply described the funding as being AME. Hepper later told the inquiry that she was ‘pretty sure’ she had verbally explained to Foster that the funding could threaten the department’s budget if it overspent. Whatever was said in conversation between Hepper and Foster, the civil servants would go on to continually misrepresent the funding as simply ‘AME’ for years, with one civil servant suggesting that the inaccurate information was simply copied and pasted from one ministerial submission to the next.
Foster herself would tell the inquiry that her belief that the bill was being paid by London influenced how she viewed the issue of cost controls. David Scoffield QC for the inquiry put it to her that in a later 2013 submission to her from Hepper there was ‘a pretty stark warning’, because it referred to a ‘finite budget’ which ‘can’t be breached’, and explained that therefore ‘we must have this facility [an emergency brake of cost controls] available to us.’ Tellingly, Foster replied: ‘Yes, but I was also aware that the AME budget [meaning the Treasury would pay the bill, rather than Stormont] was there [in the submission].’
At the inquiry, Hepper played down the significance of the point, arguing that the department never intended to overspend anyway and would ‘try our best to make sure that that did not happen’. Although she knew that limits on the scheme were necessary to prevent overspending, she thought that ‘we’d plenty of time to do [cost controls]’.
Sam Connolly, the DETI economist assigned to assist energy division, accepted at the inquiry that Foster had made her decision based on ‘untrue’ information in the submission from Hepper. Connolly would himself admit to the inquiry that he had failed to follow the rules for public sector economists and argued that RHI was a policy ‘involving concepts and sums of money that were beyond me’.
***********
Days before Foster had even formally decided that there should be an RHI scheme and before the final CEPA report had been received, Hepper took a decision of her own which indicated that she was confident the minister would back that option. Hepper authorised the spending of public money on an outside law firm, Arthur Cox, to write the legislation that would set up RHI. It was a remarkably bold act which further suggested Hepper was convinced that she knew the minister’s mind. But what was to follow was more baffling.
Alan Bissett, a partner at Arthur Cox, was a specialist energy lawyer whose services did not come cheaply. When DETI asked him to work on drafting the RHI legislation, he expected a substantive assignment. But it gradually became clear that something else was going on. Bissett was ultimately paid to largely copy and paste the Westminster legislation, and in doing so he replicated the GB scheme’s system of tiered tariffs. But that was stripped out by DETI officials – not once, but on multiple occasions. Having hired lawyers to draft the regulations, civil servants now disregarded Bissett’s work and produced their own draft of the regulations.
Speaking of the bulk of the work, which he was asked to undertake, Bissett told the inquiry: ‘It didn’t seem clear to me why it couldn’t have been done by the Departmental Solicitor’s Office. We weren’t advising anything to do with policy; we were just drafting and were changing references and correcting terminology from GB to NI terminology.’ One DETI official suggested Bisset capitalise some letters and queried his use of semicolons. Meanwhile, cost controls were being stripped out without alarm. Ultimately, it became clear to Bissett that his company was one of three sets of lawyers involved in drafting the legislation, with Stormont’s own lawyers and those from Ofgem also contributing. ‘Objectively it seems odd,’ he told the inquiry.
Bissett also told how DETI later commissioned his firm to advise it on what ought to have been a basic function of the civil service – asking their Whitehall counterparts what they were planning to do in future with their RHI scheme. Inquiry counsel David Scoffield QC asked him: ‘Why on earth were lawyers being paid to do that?’ Bissett replied:
That did seem odd … I had assumed that relationships between DECC and DETI were not good and we were being asked to do that … I would have felt that DETI would have been in a better position to tell me what DECC was doing. We had no contact with DECC – they wouldn’t speak to us. One of my colleagues tried to find out timelines and what was happening and they wouldn’t speak to us.
But while Bissett wasn’t kept in the loop, Hepper and her team believed they had good reason to disregard the regulations drafted by him and instead draw up their own legislation without tiering.
The explanation lay with CEPA. In its report, the consultancy had set out proposed tariffs for a host of green technologies – from ground source heat pumps to solar thermal units. It drew attention to the fact that the subsidies proposed included two tiers for some technologies, in line with the situation in GB. The consultants explained how tiering operated and went on to say: ‘However, when setting the NI recommended levels for this report, the incremental fuel cost was higher than the subsidy rates in all c
ases. Therefore no tiering is provided in the rates in this report.’ With that decision, the consultants were taking an enormous risk. If the cost of fuel was ever less than the rate of subsidy, the scheme would be fundamentally flawed. Even at that point, it should have been obvious that the cost of fuel was constantly fluctuating while the proposal was for the tariffs to remain set in stone for 20 years, with guaranteed inflationary increases.
But the problem was to get much worse. DETI launched a public consultation on its proposals and received responses which complained that the tariffs proposed for Northern Ireland were lower than for GB. At a glance, the Northern Ireland scheme did indeed seem far less generous. Not only were the initial tariffs for the smallest biomass boilers lower – 4.5 p/kWh in Northern Ireland as opposed to 7.9p in GB – but the most lucrative GB tariff was available for far larger boilers. In GB, it was possible to install a 199 kW boiler, capable of pumping out more than four times the heat of the 45 kW boilers which were to get Stormont’s top subsidy.
But the small print showed that unlike the situation in GB, where tiering meant that after running the most lucrative boiler for 1,314 hours in the year the subsidy dropped to 2p, Stormont proposed to pay the top rate for the entire year – with no heat limit.
