Freethinking Economist A voice of reason against illiberal nonsense

This post is aimed very specifically at a particular characterisation of how government spending has ‘splurged’ since 2007.   It is sufficiently influential that it must have a profound effect upon the fiscal debate going forward – and it is superficially very appealing.  It goes like this:

“Before the recession, in 2005/6, government spending was already 41% of GDP.  Then, the crisis hit.  The Labour government lost all common sense and resurrected Keynesianism.  This meant a spending splurge – as their own figures show (tables B13 and B14), spending leapt from 43% to 48 or 49% over the crisis.   Cash spending went up from £627bn to £706bn in two years.  If that ain’t a splurge, what is? So, the government used the cover of a crisis to relaunch defunct economic policies and takeover the economy, to the detriment of our long term prosperity. “

If you attended as many events with right-leaning speakers as I do, you would have to listen to this sort of narrative a lot.  And it is difficult to refute.  Cash spending has risen.  The government spend is a higher ratio of GDP.  It is true.

But at the same time it uses a methodology that is incredibly misleading, in terms of the stimulative effect presumed, and the extent to which fresh government spending happened.  Bear in mind several facts:

• Businesses and households plan forwards. And they do so in terms of nominal cash. When a large business makes investment decisions, it shoves out many spreadsheets, all of which try to anticipate how much cash spending there will be for its products and savings.  Similarly, you and I, when planning whether to, say, buy a car or house or holiday, make estimations of our future cash income, and what it can buy.   We may compare it to our assets and our debts – so if our house price is falling, and our mortgage debt steady, we may pull in our horns.  So too for businesses.
• But what we do not do is make our plans based on what proportion of GDP our spending is.  If GDP expectations fall massively from £1.5trn to £1.2trn, , so that my £1500 holiday has leapt from being 1 billionth of GDP to 1.25 billionths, I do not think I am spending more.
• More to the point, if a business is expecting to get paid £150 million for a piece of work, and GDP expectations fall as above, the business does not go around thinking “Great! I was going to be paid one ten-thousandth of GDP for this, now I will get 1.25 ten-thousandths of GDP!  Party ON!
• More to the point again: if one set of people – say, 10,000 teachers – were expecting to be paid £300m, and were basing their consumption decisions around that, the fall in GDP would not make them think they were getting more money, just because it represented a bigger chunk of GDP.
• People’s current expectations of what their future incomes will be – the sum total being the economy’s expectations of forward Nominal GDP – play a dominating role in determining CURRENT spending. About a million hat-tips: Scott Sumner.  It is funny how easily the Right recognises this fact when trying to use Ricardian equivalence to disprove the ability of government to achieve anything with changes in its stance.
• In 2007, the government was forecasting spending in 2010-11 of £678bn. Table B11.  So the latest forecast spend is £30bn higher.  About £12bn is in higher debt interest; about £15bn in tax credits and social security.   There has been no remarkable increase in actual public works. The ‘stimulus’ for what it is worth was on the revenue side: failing to tax spending as much as before, for 13 months.

So putting these all together, what do we get?

• The boost in spending/GDP almost entirely reflects future GDP falling.  This does not stimulate ANYONE!   In particular, none of the sudden increase in that ratio would have fed into some businesses and people revising upwards their previous expectations of income derived from the government – or indirectly from it.
• The cash increase that DID happen was hardly stimulatory.   Out of work people got more benefits than expected, and our creditors got paid interest.
• Attempting NOT to do these spending increases would have immediately lowered expectations of future GDP, which as they had stood would have anticipated the increase from 2007-8 of £589bn to £678bn in 2010-11 in their CURRENT plans.  (This ~4.8% p.a increase was originally expected to match the rise in NGDP, incidentally)
• Since the financial sector was ****ed, the lower expectations of government-derived income would not have been substituted with expectations of other incomes derived from private sources
• As a result, businesses and households that were expecting incomes of a certain size (whether directly or indirectly from the government) would have had to drastically lowered their expectations of future NGDP, which would have lowered their current nominal spending plans, which would have lowered current GDP.

This is blindingly obvious: if the government had (madly) targeted the level of spending as a proportion of GDP during a nominal GDP slump, then all it would have done is made GDP slump even more.  Even if it had done this at the same time as cutting taxes, because tax cuts would have gone towards groups with a lower marginal propensity to consume, particularly given the collapsing asset markets, than the recipients of government benefits and incomes.

Don’t be fooled by ratios.  Yes, we need to get spending down as a ratio of GDP.  yes, Brown spent too much, relying on mirage revenues.  But what happened over 2007-10 is not the splurge, and using spending as a ratio of GDP is, for such situations, a lousy metric for understanding macroeconomic relationships.

UPDATE: if you want robust confirmation of this view point, read Martin Wolf today. Read him anyway.

What would happen if governments also slashed their spending? In an economy without monetary or exchange-rate offsets to austerity, any reduction in spending is likely to lead to at least an equivalent short-run reduction in output (a “multiplier” of one). An attempt to cut a fiscal deficit by 10 per cent of GDP, via cuts in spending, would require an actual reduction of 15 per cent of GDP, once one allows for falling fiscal revenue. GDP would also shrink 15 per cent. As Desmond Lachman of the American Enterprise Institute pointed out in FT.com’s Economists’ Forum, the decline could be even larger.

Some of you wiseguys may retort that we DO have monetary and exchange rate offsets.  But where is the evidence that (a) they work for the real economy at this point and (b) that even if the ER did fall, exporters would increase volumes?  As Conway points out, and I did in Slash and Grow, they may just increase margins.