Clue one that this is the case lies in the fact that official forecasts for government borrowing have often been wildly wrong – something which would not be true if the government controlled its borrowing. For example, between 1990 and 2019 the average error on official forecasts made in the spring for net borrowing the following financial year was £22bn, or 1.6 per cent of GDP. And there’s a pattern to these errors. Unexpected economic booms such as in the late 90s cause borrowing to be less than expected, whilst unexpected downturns (such as in the early 1990s and 2008-09) cause it to be more than expected.
Our second clue is that this year hasn’t only seen record peacetime government borrowing. It’s also seen a record high for household savings. In the second quarter we saved 28.1 per cent of our disposable income, by far the highest on record. What are the chances that two different records could be broken at the same time?
Well, actually very high because these are not different records at all. Government borrowing is simply the mirror image of private sector net lending: this is true by definition, as every pound borrowed must be a pound lent. Because the government has only limited influence upon private sector net lending decisions so it has only limited control over its own finances.
This explains the forecast record. In booms, the private sector borrows to invest, the counterpart to which is that the government runs a surplus. And in recessions the private sector cuts investment and raises saving and so becomes a net lender – which means the government must be a net borrower.
Hence this year’s events. With pubs, restaurants, cinemas and non-essential shops having been closed, or having limited capacity, or being thought a health risk, those of us who have kept our jobs have become huge savers. By definition, this means the government must have been a big borrower.
Which explains why gilt yields hit record lows this year. The same private sector savings that drive up government borrowing also create increased demand for safe assets: the rise in these savings has of course been a worldwide phenomenon.
Herein lies a reason to suspect borrowing could fall a lot. If the roll-out of a vaccine next year allows us to return to normal, we could see a consumer boom as we celebrate our freedom. As household savings fall, so too will government borrowing. If the reduction in uncertainty also unleashes a pick-up in private sector investment, the fall will be all the greater.
Now, we must distinguish here between conditional and unconditional forecasts. I have my doubts about the unconditional forecast “government borrowing will fall sharply.” This might not happen if fears of job loss keep savings high, if the vaccine isn’t delivered quickly, or if we’ve fallen into more frugal habits.
What I am confident of, though, is the conditional forecast: if private sector net lending falls sharply, so too will government borrowing. This is true pretty much by identity. You simply cannot discuss government borrowing without considering the course of the private sector’s financial balance.
