Holed government accounts

31 December 2011
In theory the Whole Governments Accounts ought to be a pinnacle of transparency. Instead it is a cop-out which shows political motivation. I will focus on two areas: pension liability since it is by far the largest item, and private finance imitative liability as its presentation borders on accounting fraud. I am not an accountant, but seemingly inconsistent treatment of financing costs looks rather spurious considering the extent of unfunded liabilities. Much of the remainder is of little interest for those who just wants numbers, and in any case as stated on page 54 of the report, it concentrated more on working how much was spent rather than analysing where.

Remuneration Report

A fairly short & dry section, but it does have one nugget: Comparable public & private median salaries. The table in paragraph 6.14 is pretty explicit in showing the public sector salaries are higher. Driving it home is the graph in 6.16 showing public sector salaries having a bigger rise from what was already a better base-line. I am a bit suspicious of the implication that private sector salaries actually rose in 2010, as the actual figures are not quoted. One thing that also springs up is whether the annualised figures in 6.15 and the weekly figures in 6.16 refer to the same cohort. Assuming they are, it implies that the median full-time public sector worker works 42.3 weeks a year, or in other words gets 46 days off per year.

How much?

It is not a good when paragraph 5.12 states that figures are materially misstated by delayed/absent reporting, unclear reporting of intra-organisation transfers, and outright inaccurate reporting. Nevertheless paragraph 3.2 on Key Figures spells out the damage. If this was a company the receivers would have been called in yesterday.

Total revenue of £582bn (of which £485bn is tax) against expenditure of £666bn (excluding financing costs), which represents a 22% (£165bn) overspend. If finance costs are included, as they are in paragraph 7.34, then total government spending is £746.5bn, which is a 28% overspend. Normalised to total spending, 22% of spending is serviced by borrowing. Assets only cover 50% (£1.2trillion) of liabilities, with the total liability (paragraph 3.3) of £2.4trn being 170% of GDP.

Pension liability

If Chart 3.2 (page 12) were in an annual company report, the finance director would be in prison. Anet (i.e. unfunded) pension liability is simply illegal in the private sector. To make matters worse, it is £1,132bn. Leaving aside the issue of this being 85% of GDP, or 151% of annual government expenditure, this liability cannot be inflated away due to the way the underlying pensions are index-linked.

While the report is pretty forthright in presenting this unfunded pensions a single massive liability, it took a bit of digging to find out per-annum expenditure on pensions. Under Key Figures in paragraph 3.2 it is listed as the largest component of social benefits, weighing in at £70bn (paragraph 3.20 states it as £69.5bn). However, under paragraph 7.34 there is a separate component called Finance Cost of Pension Liabilities, which is a further £49.9bn. This £119.4bn represents 16% of government expenditure, or 20.5% of government revenue.

Private Finance Initiatives

The section on Private Finance Initiatives (PFIs) is highly evasive, in particular the way it semi-separates service charges from capital charges/interest, and the not-exactly-clear division of annual and total costs. Being a non-accountant, there is no logic in the separation, and getting a full picture is less than straightforward. Extracting the raw numbers gives:

£27.2bn payable under PFI contracts
£30.9 net book value of PFI assets
£33.4bn interest year-to-date on capital commitments
£28.1bn liability for capital repayments
£131.5bn present value of future obligations
£6.4bn estimated annual service charge

Best guess for an annual expenditure figure is summing the first and third figure, which gives £60.6bn, which is 10.4% of government revenue. This is of concern as it is substantially higher than the £30bn it cost the government to service the official government debt.

The £131.5bn figure is clearly not an all-in figure, and as far as I can tell it is basically just service charges and some replacement expenses. Assuming the replacement expense component is negligible (a dubious assumption), and that £6.4bn multiplied by the average length of remaining time of PFI contacts was used to calculate it (paragraph 3.69), this implies that the average remaining time of PFI contracts is 20.5 years (realistic as most PFI deals are 25-30 years). If the interest on capital expenditure is 8%, then annual interest payments of £33.4bn imply a total underlying capital liability of circa £415bn. That implies an outstanding total liability of £546.5bn. If 20.5 years worth of interest is also included, that adds another £685bn,which brings the total bill to £1,231bn.

We're screwed

Pension & PFI payouts and servicing the national debt, between them, represent £210bn of contractual costs that cannot simply be axed. To make matters worse, they take up 36% of revenue, and hence 28% of public spending. That is a major straight-jacket when a good quarter of total spending is funded by borrowing. If the report is to be believed then the country has to pay out £2.4trillion, but since it is doubtful that PFI liabilities have been properly acknowledged, even this 170% of GDP looks like a conservative estimate. With PFI deals fully accounted for, I suspect the real total is nearer £3.5trillion (i.e. 170-248% of GDP). And unlike 1946 there is no Marshal Aid.