Lies, damned lies and percentages of GDP

A bit of a row blew up over Christmas and New Year about the government’s claim to have halved the deficit. Frazer Nelson called them out over it, saying that they hadn’t halved it at all. A government spokesman replied to the Spectator that they meant “as a percentage of GDP”.

Meanwhile, the Conservatives have been getting annoyed by claims that their proposed cuts would take public spending back to the 1930s. That figure comes from data in the OBR’s report and is also based on a percentage of GDP.

So if the Conservatives have halved the deficit….

Screen Shot 2015-01-09 at 18.39.58

…then their spending plans must also be taking us back to the 1930s.

Screen Shot 2015-01-09 at 18.36.44

In fact, both these claims are exaggerations. The government hasn’t halved the deficit and, even if George Osborne’s spending cuts were implemented, they still wouldn’t take us back to 1930s spending levels.

The problem with expressing things as a percentage of GDP is that the economy grows and shrinks. As Colin Talbot pointed out in a comment on a post by Jonathan Portes, a percentage is a fraction, so a movement in either value can change the result. It will change if the number on the bottom moves, even if the number on the top stays the same. When the economy is growing and GDP rises, even if the deficit stayed the same, it would reduce as percentage of GDP because the number on the bottom of the fraction is getting bigger. Likewise, even if it stayed the same, it would increase as a percentage of GDP if GDP fell.

The thing you are trying to measure can even move upwards yet appear to move downwards if GDP is rising at a faster rate. It is possible to reduce a country’s debt-to GDP ratio while still running a deficit, provided that GDP rises at a faster rate than the accumulated debt.

In recessions, governments get a triple whammy. The tax take falls and social security costs rise, so the deficit goes up. But GDP falls too, which amplifies the size of the deficit and the debt.

Two charts from the recent ONS report on long-term trends in public finance illustrate the point.

The first shows current (as in day-to-day) central government expenditure compared with current receipts, in cash terms.

Screen Shot 2015-01-09 at 18.48.34Since the Second World War, public spending has increased with inflation, spending commitments and the rise in population but, on the whole, revenues have kept pace with it. After recessions (the grey bits on this chart) expenditure exceeded revenue for a time but the gap closed again once the economy recovered.

The second chart shows the same data as a percentage of GDP.

Screen Shot 2015-01-10 at 14.33.50


On this chart, the increases in expenditure after recessions appear to be much greater but the steep fall in tax revenue after 2008 is not as clear. It looks no worse than that of previous recessions. This is because the collapse of tax revenues was matched by (and caused by) a collapse in GDP. Both parts of the fraction are therefore going in the same direction so the fall doesn’t look as sharp.

Comparing public finance data as a percentage of GDP, then, only gives part of the picture. It is important to look at other measures too, like real-terms spending and, especially for public services, per capita spending.

While the Conservatives’ spending plans would take public service spending back to 1938 as a percentage of GDP, it would only take it back to 2000 in real per capita terms. Even that doesn’t give a full picture, though, because the over 65s go from 15 percent of the population in 2001 to 20 percent in 2020. The same money, even in real terms, won’t go as far.

Measuring public spending and debt as a percentage of GDP is, as Jonathan says, the economists’ preferred measure. There is a good reason for this. Size of GDP is an indicator of a country’s ability to sustain its spending and cover its debt repayments. When economies recover, the triple whammy goes into reverse. GDP rises, tax revenues increase and welfare costs fall, which makes the numbers look better all round.

At least, that’s what used to happen. Where this recovery is looking very different from previous ones is that, despite healthy-looking growth, tax revenues have fallen well short of expectations and the social security bill is still stubbornly high. The deficit is as much a labour market question as a public spending one. Rising GDP might flatter the government by making it look as though its fiscal worries are easing, but unless they can translate that rising GDP into revenue, the deficit will still be there in 2019.


This entry was posted in Uncategorized. Bookmark the permalink.

3 Responses to Lies, damned lies and percentages of GDP

  1. roGER says:

    According to BBC Radio 4’s More or Less show, the “1930s level of spending” is actually taken from the figures for a single year – 1938. That year was exceptional as frantic efforts at rearmament were taking place, so government expenditure that year was high in comparison to the earlier years of the decade.

  2. Bill Wells says:

    You conclude that ‘the deficit is as much a labour market question as a public spending one.’ Yet, given the enormous tax cuts in direct taxation since 2007 (under both governments) – in both income tax and corporation tax – I suspect that taxes rather than wages is the more important factor in explaining the deficit. The lack of recovery of tax revenues is that these tax cuts have led to a structural reduction in the revenue gathering aspects of these taxes.

    In 2007 the main rate of corporation tax was 30% and of income tax was 22%. By April 2015 both rates will be 20%. In addition, the income tax allowance will have more than doubled from £5,225 in 2007 to £10,600 in 2015.

    Recent publications by ONS and BIS provide some support for this view as well as confirming the IFS evidence that wage inequality is narrowing at the bottom end of the distribution.

    The January ONS Economic Review has considered ‘Tax receipts, the personal allowance and the earnings distribution.’ and concluded that ‘the number of people earning less than the personal allowance – and therefore paying no income tax – rose to more than five million in 2014, almost double the number who would have been below the threshold had it moved in line with the CPI.’

    The BIS final evidence to the Low Pay Commission on the National Minmum Wage published on 9th January – particularly Charts 2.5, 2.6 and 2.10 – show that real wages/take home pay for the low paid (particularly for NMW workers) have risen since the pre-recessionary period.

    Click to access bis-15-25-national-minimum-wage-final-government-evidence-for-the-low-pay-commissions-2015-report.pdf

    Bill Wells

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s