(Originally published in February 2010 on the IMF's PFM blog)
The financial pages have over recent weeks been full of stories about the accounting “innovations” used by Greece and other European countries to artificially lower their reported deficits. However, the accounting strategies in question – in particular swaps and securitization – were banned by the EU years ago, and can no longer be used to push deficit numbers down. The real story is the continuing scope for creative accounting around Maastricht debt figures – that is, the debt measure used by the Europeans to measure national debt against the (now greatly exceeded) limit of sixty percent of GDP under the Stability and Growth Pact (SGP).
There a four key features of Maastricht definition of debt which leave it open to many of the manipulations which have been in the spotlight recently:
It is a cash-based rather than accrual-based measure of government liabilities;
It measures debt at its face value – i.e. redemption value – rather than its market value;
It is a gross debt measure – that is, a measure which does not net off against debt any financial assets (even money in the bank);
It fails to count a wide range of financial liabilities.
Let’s take each of these in turn.
First, by looking at the cash rather than accrual value of government liabilities, the Maastricht definition of debt is inconsistent with the debt measure used in Europe’s own accrual-based statistical reporting standards (the European System of Accounts 1995 – ESA95). It is, as a consequence, out of line with the current Maastricht deficit definition, which is based on ESA95. This means that there is a limited relationship between deficits and changes in debt with the Maastricht framework. The inevitable result is substantial “stock-flow inconsistencies” (to use the fiscal policy jargon) which would, under a more coherent accounting framework, provide a clear signal that “creating accounting” was going on.
Second, measuring debt at its face rather than redemption value opens the door for transactions which reduce reported debt without in any way changing the real debt burden. Consider the Greek swap transaction involving Goldman-Sachs which has attracted so much interest in the past few days. Essentially, what this swap involved was reducing the face value of debt at the cost of a higher interest rate on the debt. That is, those who provide the replacement loan are compensated for its lower redemption value by receiving a higher interest rate. Such an artificial transaction does not reduce the market value of debt, because the country concerned is no better off. It is precisely for this reason that best practice – as followed by many OECD countries and required under ESA 95 reporting rules – is to measure debt in market value rather than face value.
But because Maastricht, once again, diverges from the EU’s own fiscal reporting standards, it remains possible to use these kinds of transactions to understate the value of their liabilities.
Third, the use of a gross rather than net debt measure has the consequence of encouraging governments to minimize their cash balances and other financial assets below what may be required for good liquidity and financial management purposes simply in order to dress up their reported debt levels. But because cash can be used at any time to repay debt, such “reductions” in debt are purely artificial. It can also create incentives to dispose of financial assets (like shares) during period where the government is least likely to get a return (like in the middle of a recession).
Fourth, by far the biggest shortcoming in both the Maastricht and the ESA95 debt measures is the failure to recognize unfunded government pension liabilities as a form of debt. These liabilities are, for many European countries, enormous – as I illustrated in a recent blog piece in the case of France [include reference]. By contrast to the European measures, both the OECD’s net debt measure and its IMF equivalent (“net financial assets” in the framework of the Government Financial Statistics Manual 2001) clearly recognize accrued but unfunded pension rights of civil servants as a government liability.
Not only Greece, but the whole EU, needs to take a look at its fiscal data. Starting with the way it measures debt.
We should not be afraid to be transparent about the debt position for fear of playing into the hands of those who are urging an unduly rapid fiscal consolidation which would put in jeopardy the fragile recovery presently underway. The policy challenge in Europe at present is that of providing the necessary short/medium term fiscal support while simultaneously addressing the significant medium/long-term fiscal sustainability problem – e.g. through reforms to pension systems. It is crucial to focus on both elements of the problem. Understating debt merely gives succor to those in Europe – of whom there are many – who claim that there is no underlying sustainability problem.