Philip Hammond MP. Photo: Oli Scarff
Note to Philip Hammond: In ALL 32 OECD countries that cut spending, economic growth was seriously damaged
Since 2010, under the direction of international organisations and most economists, governments across the world have cut spending in order to restore public finances to health. OECD figures show this strategy has failed. Cuts have greatly damaged economic growth – to a far greater extent than anticipated. As a result, little order has been restored to the public finances. I leave aside social consequences.
This blog reviews the OECD evidence (updating previous work to include 2015), implicitly reasserting the case for an alternative approach – in the light of encouraging remarks by the new Chancellor. (It’s long, inevitably.)
First the idea of cuts needs clarification. In general fiscal policies in the post-financial-crisis period have not amounted to cuts in the level of spending; instead the growth of public spending has been greatly reduced. The following chart (hard to read I know, fuller figures are in the annex) shows the growth of spending in the pre- and post-crisis periods (defined as 2003-2008 and 2010-2015) and the change, for each OECD country.
Government spending, annual average growth
This is important because keeping spending growth positive means a boost to aggregate demand, even if it is still very painful on the ground. In today’s FT Rupert Harrison, George Osborne’s key economic adviser in the coalition parliament, admitted “the rhetoric of the cuts was always worse than the reality in order to gain public support”.
The exceptions to the general rule are Greece, where the level of spending after the crisis was reduced severely year after year, Portugal and a handful of others. Conversely, at the top of the chart, in Germany, Israel and Japan the growth in government spending has increased – albeit relatively modestly but still materially (less so in Japan perhaps).
[In an ideal world we would include government investment as well, but in the national accounts government investment is part of ‘investment’ and, in the OECD data, not separately identifiable. As in the UK, the cuts in spending growth would likely be higher if investment was included. Note also that the figures used here and throughout the document are spending in current/cash prices. There are big problems with translating government spending into real terms so cash figures are better here, but, more generally, with the ultimate interest in public finances, cash figures are the relevant measure.]
The impact of spending cuts on the economy
Given policy has acted on the growth of government spending the relevant outcome is the growth of the economy. The next chart therefore shows the change in growth between the pre- and post- crisis periods (as above, defined as 2003-2008 and 20010-2015). The total change in GDP (Y) is subdivided into the various sources of demand: household consumption (C), government (G), investment (I), exports (X) and imports (M).
Contributions to the change in growth since the crisis, percentage points
In aggregate, in virtually all countries, GDP growth has been reduced relative to the pre-crisis period. On the left of the chart the scale of the reduction in growth has been extreme – and this is a rate per year. In the detail, all sources of demand have contributed to these reductions, with the exception of imports (a reduction in imports has a positive effect, but this is symptomatic of the decline in growth and not a good thing in its own right).
Given the severity of the declines in certain countries, the next chart focuses in on countries where the annual decline in GDP growth is less than 3% a year.
Contributions to the change in growth since the crisis – zoomed in, percentage points
The declines may be smaller relatively speaking, but even say for the EU as a whole an average annual decline of 1 per cent a year still amounts to 6% in total. The UK decline is hardly the most severe, but it is worse than the EU as a whole (and the US), and, as has been discussed before, has still amounted to the slowest recovery on UK records that extend back to the 1850s. At this detailed level the same is still true, pretty much all sources of demand have been reduced – the stand out exception is Ireland with huge export growth – but this is related to the peculiarities of their FDI-based model, and associated measurement issues which are under some scrutiny at the moment.
Most strikingly of all, the only three countries where GDP growth actually increased were the three countries at the top of the government spending chart.
Spending cuts and GDP – further examined
According to the most optimistic doctrine of ‘expansionary fiscal contraction’, cuts should have given way to higher growth elsewhere that more than offset the decline in government spending. THIS HAPPENED NOWHERE. Everywhere that G growth was cut, all other sources of demand also weakened.
The next chart shows the relation between changes in G against changes in Y (GDP) as a scatter plot (UK in red), so that we can assess the strength of the relationship between the two factors. It is fairly obvious that in general the higher the cut in spending, the higher the cut in GDP. The standard measure of such associations is correlation (from 0 to 1 for positive associations), and this scores 0.8 – indicating a strong relationship. Though the relationship is far from a straight line (correlation = 1), which would only be the case if spending cuts were the sole determinant of outcomes – which obviously they are not. For example UK monetary policies have fostered disproportionately higher household consumption – seen on the previous chart as a smaller-than-might-be-expected reduction in the red column.
Change in GDP versus change in the government contribution
Nonetheless, as in the headline, all 32 countries that cut spending, saw GDP growth reduced. The only countries that saw growth rise were the three that increased spending. There were in fact two other countries that increased spending very marginally (Switzerland and Chile), and growth still slowed. To reiterate – there are no countries in the top left quadrant, where they should all be if expansionary fiscal contraction had validity. The full table of results by country is in an annex.
From GDP to the public debt
In the specific instance of the UK, in 2010 the OBR forecast GDP growth over 2010-15 little changed by spending cuts. Not going as far as expansionary fiscal contraction, they still broadly expected most other sources of demand to compensate for the reduction in government demand. The private sector was expected to be ‘crowded in’ by the reduced public sector. This was wrong. The OBR and others maintained that the Eurozone crisis blew their forecast off track. But this is hard to sustain, given the same decline is true of all OECD countries.
The relation with the public sector finances is not straightforward. Given spending cuts operated on the rate of growth, most countries continued to ‘enjoy’ positive GDP growth and so public sector revenues also continued to grow, and therefore public sector borrowing was reduced (and notably the cuts were originally implemented under conditions of growth, thanks to the global stimulus of 2008-09). But the key point is that these improvements were at a far lower rate than expected. UK public sector net borrowing was expected to be around 20bn in 2015-16, instead it came in around four times higher at close to £80bn. As a result the public debt ratio in the UK has continued to rise and has repeatedly defied expectations of improvement. The same is true across the OECD, as the below chart from the IMF indicates.
Repeatedly public debt ratio forecasts are revised up, and the point at which the figures improve moves into the future (I suggest ignoring forecasts). Here is an extract from a more detailed IMF table:
Notably the only advanced economy to actually have public debt materially improving is Germany – and this is the factor behind any improvements in the aggregate euro measures. In Germany higher spending has meant stronger growth and therefore more rapidly improved public finances. (There is also a slight gain in Japan – another spender, but hardly worth shouting about, and a miniscule one in Spain. Should also remember that privatisation permits gains in the public debt that do not follow from the underlying strength of the public sector finances – UK again.)
More generally most measures have crashed through the 90% threshold (owed to Rogoff and Reinhart, and somewhat controversial in its own right) that the whole of policy was aimed at avoiding.
Given the political constraints under which they operate, the IMF and OECD have now come as close as they might to saying that cuts policies have failed. The OECD now recognise that increasing spending will quite possibly improve the public finances. They confine their policy recommendations to investment spending, but the analysis here suggests that changes to current spending will also have this effect.
George Osborne might have the excuse that in 2010 he was simply following the advice of the same international organisations as well as his economists. His economic instincts were sharp enough to recognise that his early cuts were threatening recession, and so towards the end of the last parliament shifted to a more expansionary course. He was also seemingly well served by political instincts that kept the change of course under wraps. But the spending plans that are his legacy still amount to a further period of severely restricted government spending, around the order of the last parliament – as the OBR has pointed out (see chart 3.33 of their latest forecast). The same is doubtless true of all other OECD countries.
The TUC has argued that clear fragilities were evident ahead of the referendum vote. All are agreed that decisive action is now needed to protect the economy, and the Bank of England have emphasised that there are limits to what they can do. The TUC has set out its own proposals for an alternative way forward: here.
The evidence in this post is offered to a new government that has the opportunity to break decisively with the failed policies of the past.
Annex: contributions to government spending (ppts) and GDP growth
(each table is ranked by the change in government contributions)