Yet again, government borrowing has exceeded expectations. So far this financial year, the deficit on current borrowing* has been ''42.1bn compared to ''30.9bn in the same period last year. This puts in doubt the OBR's forecast that this deficit would fall this year, from ''98.9bn to ''95.3bn.

Duncan says this shows how austerity is self-defeating; a squeeze on spending weakens growth and thus reduces tax revenue.

I'd add that this will remain the case for as long as the private sector deleverages. This is because of a simple, basic, unavoidable identity - that for every borrower there must be a lender. Government borrowing - by definition - means that other sectors are net savers.

The problem is that there are only three other sectors, and all three have reasons to want to save or pay down debt:

- Households. Low consumer confidence means these are loath to borrow. And it might be the case that - with the debt-income ratio still high - they will continue to deleverage.

- Foreigners. The desire of Asian economies to save heavily is unlikely to cease soon. And the credit squeeze in the euro area is creating forced savers.

- Companies. Spare capacity, a dearth of investment opportunities and depressed confidence (partly thanks to the euro crisis) are all holding back investment, and encouraging firms to build up cash piles and pay off debt.Today's CBI survey found a fall in order books, consistent with a continued reluctance to invest.

Now, as long as these three sectors want to save, the government will have to borrow, simply as an accounting identity. The mechanism through which this happens is, of course, that higher private savings mean weak activity, which means weak tax revenues and higher welfare payments.

There are three implications here.

1. Government borrowing will fall when and only when the private sector saves less. The government is not in control of the public finances. Austerity works as a deficit reduction policy only insofar as it encourages the private sector to save less. Whilst this might happen sometimes - if tighter fiscal policy encourages borrowing by lowering interest rates or increasing business confidence - these mechanisms are weak here and now.

2. The only deficit reduction strategies that make any sense at all are those which aim at reducing private sector net saving. The Funding for Lending Scheme, aimed at encouraging borrowing, is the right kind of idea, though I doubt how practically effective it'll be.

3. An overshoot in government borrowing will not greatly raise gilt yields; although gilt prices fell today they are extraordinarily high by historic standards. This is because the same private sector savings that cause the government to borrow also create a favourable climate for gilts - namely a weak economy that causes investors to favour safer assets, and a pool of cash from which to buy assets.