Wednesday, October 21, 2009

Darling’s borrowing forecast may turn out right

The UK's public sector net borrowing reached £14.8bn last month - a high for September… Net borrowing for the six months of the financial year so far now stands at £77.3bn - the worst figure for the April-to-September period on record.

The figures raise speculation that the government may have to revise its forecasts on borrowing.

Possibly. But I’m not sold.

I think the Budget forecast – for public sector net borrowing of 12.4% of GDP in 2009/10 – looks like turning out right, or possibly even a bit pessimistic.

This graph shows rolling 12-month figures for public borrowing as a % of GDP, based on official ONS borrowing data and monthly GDP estimates from the NIESR (rough explanation at the bottom).


The solid line is what’s actually happened, and the dotted line is how things will go if the average rate of increase over the last six months continues. By next March (virtually the end of the 09/10 fiscal year), it comes to a bit over 12.3% – just below the Budget forecast.

The usual caveats apply: past performance is not necessarily a guide, etc. But one thing that could make a difference is the end of the recession. GDP growth over the next six months is (touch wood) going to be considerably better than the last six. This will reduce any given amount of borrowing relative to GDP.

Even if I’m right, though, let’s not go dancing in the streets. 12ish% is still bloody high. And then there’s next year to worry about.

Rough explanation: I used the monthly cash borrowing figures from the ONS (series PSF2, ANNX) and the monthly GDP estimates from the NIESR, and did what I could to control for the differences in the ways the two institutions present their data. This involved using the ONS’s rolling four-quarter figures for borrowing as a % of GDP (series PSF9, J4DD) as a series of anchor points (for March, June, September and December – i.e. the final month of each quarter), weeping at my lack of social life, and also comparing the NIESR’s quarterly GDP figures with the equivalent from the ONS to generate a set of monthly GDP figures that cohere with the ONS quarterly ones. If you want more details, please shoot me and/or yourself.

3 comments:

chris said...

The problem here - if there is one - is solely of the Treasury's own making.
It has always - and this pre-dates Darling by decades - presented forecasts as if they were single points. But in fact they are (wide) ranges.
The average error in Budget forecasts for net borrowing is around 1% of GDP - £14bn in today's money.
We should therefore read Darling's £175bn forecast as meaning: there's a roughly two-thirds chance of borrowing being between £161bn and £189bn.
The figures so far are consistent with this.

Tom Freeman said...

Yes. This deranged fetish for exactness is of no use to anybody. Better to be roughly right than precisely wrong.

Tom said...

The trouble, of course, is that the media would report that as both

"PUBLIC BORROWING COULD INCREASE BY AS MUCH AS £189BN"

and

"CHANCELLOR DOESN'T KNOW HOW MUCH HE'S BORROWING TO NEAREST £20BN"