OBR wrong to continue absolving Osborne’s polices for a deficit now 3x larger than planned
TUC has argued that the economy and public finances were derailed over the coalition years because government spending cuts had a larger-than-expected impact on the wider economy (see e.g. blog, paper).
Earlier this week the Office for Budget Responsibility (OBR) issued their annual Forecast Evaluation Report which gives the official take on this derailing. And some derailing it is. Putting figures on a consistent basis, the OBR find for 2014-15 borrowing was expected to be £30 billion; it turned out to be £90 billion – three times larger.
Happily for the government (on the day before the ‘Charter for Budgetary Responsibility’ debate), the OBR once more find blame lies elsewhere (i.e. in the eurozone) for the weakness of the economy that is behind the derailing of the public finances:
We continue to believe that the most important factors underlying our real GDP forecast errors – in both directions – have been the occurrence or absence of shocks to credit and confidence, in particular those emanating from developments in the euro area. The magnitude, composition and profile of the errors we have made mean it is unlikely – though of course not impossible – that underestimating the effect of the Coalition’s fiscal consolidation on UK growth was a material explanatory factor, either in the unexpected weakness of GDP growth through to 2012 or its unexpectedly rapid pick-up thereafter.
I’m afraid that contesting this gets a bit technical.
The OBR conclusions are drawn on the basis of comparing outturns since the coalition took office with their original forecast from June 2010. We have argued instead that the performance of the economy should be assessed not only against the OBR’s forecast but against the performance of the economy ahead of the crisis. This allows also an assessment of the plausibility of the original OBR forecast.
So the analysis here is based on examining the annual growth of nominal/current price GDP and the contributions of demand by sector (C: households, G: government consumption and investment, I: private investment, X-M: exports – imports, i.e. net trade, ‘other’ corresponds mainly to stocks). This is done for the pre-crisis (2004-2008) and post-crisis (2010-14) episodes, looking both at forecast and outturn for the latter. Nominal figures are used because economic transactions in reality are in cash, as are public finance results; moreover ‘price’ is also an important outcome of economic processes that should not be ignored, as it is implicitly when real terms figures are used (there’s a fuller discussion in the paper).
Results are presented graphically, below; the first column shows growth ahead of the crisis; the second, how the OBR expected the coalition years to turn out; the third, the difference; the fourth, how growth turned out in reality; the fifth, the difference with pre-crisis outturns; the sixth, the OBR ‘error’ (NB as in the OBR document, this is defined as ‘outturn – forecast’ so that the OBR overstatement is indicated as negative).
Economic growth and contributions to economic growth, percentage points
It is simply seen that the OBR expected the economy not to be deflected by austerity, with GDP growth virtually unchanged before and after the crisis. Private demand was therefore expected to compensate for reduced government demand (red), with stronger investment (green) and net trade (yellow).
This was not how it turned out. GDP growth declined, as private sector spending did not rebound sufficiently to offset the reduction from government spending.
(This analysis is based on the new ONS ‘Blue Book’ figures issued at the end of September, and so is a little inconsistent with previous work. The overall differences are small: average post-crisis GDP growth is now 0.9 percentage points below pre-crisis growth, compared with 1.1 per cent on the previous iteration. )
On a demand interpretation, the results indicate that the OBR have underestimated the ‘multipliers’ which measure the strength of the relation between changes to public spending and changes to the economy as a whole. When public sector jobs, salaries and purchases direct from the private sector (e.g. staples, vehicles) are cut, there is an obvious impact on private sector activity, that then has a series of repercussions (because employees in the public sector or vehicle factories earn less or are sacked, and then spend less on food and the cinema, so cinema staff are under pressure, and in turn spend less themselves etc).
The anticipated strength of the private sector in the original OBR forecast shows they implicitly assumed a very low multiplier. (In their June 2010 forecast they stated the figures used, e.g. 0.6 for departmental spending and 1.0 for capital.) For many economists, these figures are way too low. My sense is that a number of commentators (including myself – see here from 2009) would regard 1.5 as more realistic, around double the present OBR view. A few months ago (here), Simon Wren-Lewis, Professor of economics at Oxford University, observed an absence of public debate of these critical factors in the UK. This remains the case.
Instead the OBR continue to maintain that growth was derailed by the (2011-12) euro zone crisis. But it is striking that even in spite of this crisis, net trade added more (or rather subtracted less) to UK GDP over 2010-14 than it did over 2004-08. So while net trade may have contributed less than the OBR expected, this expectation was already pretty punchy on the basis of past performance. Exactly the same goes for investment spending. Though plainly these expectations had to be punchy to compensate for the low multipliers. The results the OBR cite are circular – assuming a low multiplier necessitates high forecasts for investment and trade. So when trade falls short of their forecast it might just indicate a high forecast, not that trade derailed the economy.
3. Annual figures
Of course trade was affected by the eurozone crisis, and so some argue that taking averages across several years abstracts from more important action year by year. The chart below shows the difference in contributions to growth between the OBR forecast and ONS outturn on an annual basis.
GDP(E) growth: outturn minus OBR June 2010 forecast.
The most substantial shortfalls come over 2012 and 2013, including a sharp shortfall in net trade, and so fitting with the timing of the eurozone crisis from Q2 2011 (There’s a nice timeline of these crises here.)
But equally it fits with the impact of austerity building from the start of parliament (and the previous stimulus wearing off), and note as well that investment and consumption were already in decline relative to expectations as soon as 2011 – when at this point net trade was outperforming (overblown) expectations. Trade is only one part of a fuller story.
Moreover the eurozone crisis was/is rooted also in spending cuts, for many countries far more severe spending cuts than in the UK. It was questionable economic policy to choose to rely solely on external demand when these dangers were surely so obvious.
But, ultimately, looking simply at annual results abstracts from the basic point in the previous sections that the original forecasts for private sector activity were inherently optimistic.
That said, it is particularly notable that the reduced aggregate shortfall in 2014 coincides with government spending coming in much stronger than expected (accounting in a numerical sense for the whole of the gain between 2013 and 2014). Just as the impact of cutting spending is underestimated, the scale of the impact of this reversal in reviving the economy is also ignored. Over only a short period of years we are confronted with evidence of the macroeconomic importance of government spending in both directions.
4. International evidence
The most telling evidence, however, is from international data. These show a very clear and very strong relation between the extent of austerity and the extent of the deterioration in economic growth across all OECD countries. The summary regression from our original paper is shown below.
The correlation is 0.8, which is very high for any economic relation.
The implied average multiplier is as high as 3, well above than the value of 1.5 that was suggested above. (Though even this is far from ludicrous as work by Qazizada and Stockhammer from Kingston University has shown – here.)
Simply looking at the concentration of individual data points in the bottom left quadrant indicates that for virtually every country where government spending growth was reduced, GDP growth declined. Conversely the only countries that saw a rise in GDP growth were those where there has been an increase in the contribution of government spending (top right). This seems categorical and undeniable.
As far as I am aware, the OBR have not brought any such evidence to bear on their judgements.
With renewed economic growth, there may be a tendency to think that the impact of spending cuts is less important, but this is very far from the case.
Austerity policies have harmed economic growth and the public finances more than expected. As above, the deficit in 2014-15 was three times larger than planned; I have commented elsewhere on the even more plainly apparent failure to reduce public debt.
Yet the government plans to plough on with spending cuts in defiance of Einstein’s aphorism about doing the same thing over and over again and expecting a different result.
But a way it is the OBR that permits this defiance.
Meantime, in the real world, the post-crisis period has seen real hardship that has been entirely unnecessary, or at worst inflicted for political ends. A higher multiplier holds out the prospect that expanding government spending is the means not only to higher growth but also to material and rapid deficit reduction, let alone doing useful things and creating decent and well-paid work.
I hope the OBR will consider seriously the possibility that they are wrongly rationalising the government’s actions and so standing in the way of better policy.