The Austerity Tortoise and the Keynesian Hare:
Why Osborne is probably right, and why his critics are disingenuous
It is all too often said that a lie travels half way around the world before the truth has got its boots on, but there are some academic fields in which results should matter more than they evidently do. Back in 2010 I became involved in some very bitter exchanges over social media in which I tried to explore the idea that austerity was a growth strategy. Support for my view came in part from Carmen Reinhart and Kenneth Rogoff’s research that appeared to show a dramatic decline in growth rates once public debt rose to over 90% of GDP.[i] In other words, limiting public debt was thought to have a measurable, positive impact on long-term growth rates, hence the need to endure short-term pain for long term gain. The essential inspiration for ‘austerity’, along with the knowledge that the short-run is short, and that the long-run, by comparison, is forever.
Against this was thrown the accusation that not only was this wrong, and that levels of debt did not matter much at all (and in the long run, we’re all dead, haven’t you heard?), but that this notion of austerity being a worthwhile policy was so obviously and transparently wrong, the only explanation for supporting it was an ideological commitment to shrinking the state. Indeed, it was argued by Paul Krugman et al at the time that the austerity ‘lie’ — sometimes referred to as a ‘scam’ — had no economic credibility whatsoever and was merely throwing dust in the eyes of a credulous public.
Countering this view, aside from many famous and credentialed economists at the time and since, Jeffrey Sachs has recently weighed in to suggest that Osborne had been shown to be broadly right, and his policies broadly successful in dealing moderately and carefully with a very real crisis of credibility in the UK’s public finances.[ii] For this, he was subject to the usual criticisms and the repetitive assertion that austerity ‘had no economic credibility’ and ‘had clearly failed’ but his response was simply that Krugman was ‘being ideological’, suggesting in the end that the results of the May general election ought to serve as a corrective to Krugman’s hubris.
So who is right? Osborne and the Conservatives were elected on a platform of reductions in public expenditure in 2010, most of which were enacted as planned, then they were re-elected in 2015 with increased representation while remaining unrepentant about austerity and promising more of the same. Did the public not get the memo? How can a party whose principal economic policy has ‘no economic credibility’ being routinely described as ‘economically illiterate’ and which had ‘clearly failed’ coast to an unexpected victory? Because — simply speaking — austerity neither lacks credibility, nor has it failed.
The Missing 16%
As early as September 2013, the Financial Times — not exactly a Tory rag — concluded that Osborne’s strategy of deficit reduction had been vindicated.[iii] The Economist waited until early 2014 but nevertheless concluded that Osborne ‘had won the economic argument’ although it qualified this with ‘at least as far as the public is concerned.’[iv] Neither of these publications is neutral, of course, but nor do they lack credibility, at least on matters of economics. There was always every reason to suspect that these conclusions would be opposed in some circles, but that both of these august journals returned clear verdicts in favour of Osborne’s policies might at least suggest to the economic ingénue that the argument has not been concluded decisively against Osborne either, no matter what some economists claim. Therefore, the accusation that his strategy has no economic credibility is surely ridiculous, no?
The General Election in May 2015 provided an important point at which to reassess these competing claims, not least because an election is the one occasion when the general public have to give their verdict, but also because power is at stake, and the winner of this argument, would likely win power. So once again, we found the old arguments rehearsed but this time with more urgency, and greater linguistic flourishes than ordinarily grace the dry business of economics.
Consequently Paul Krugman — inveterate meddler in UK politics — authored a Guardian ‘Long Read’ entitled ‘The Austerity Delusion’ that, over the course of thousands of well chosen words, plotted some data on growth rates, showing that the UK’s growth rate was lower than the US and Germany between 2009 and 2013. It also implied a clear correlation between reduced government expenditure and reduced growth rates. This was done to show the abject and obvious failure of ‘austerity’ as a policy. He doesn’t get as far as showing causation, but anyone familiar with economics will acknowledge that a reduction in government expenditure will reduce growth. The debate really surrounds whether the long-run benefits outweigh the short-run costs. So, even if we accept that the chart is accurate (on which more later) it still demonstrates nothing more than an advocate of austerity would have conceded to begin with, while not looking for the related upside that said advocate would claim as justification for the short-run costs.
Then in the UK John Van Reenen of the Centre for Economic Performance at the London School of Economics published one of a series of reports prepared in advance of the election, entitled ‘Austerity: Growth Costs and Post Election Plans.’[v] This report contains many relatively uncontroversial findings, along with quite a few that are more contestable, but before the reader gets to these complex arguments, he is confronted with a chart indicating that the UK economy is 16% smaller in 2015, than it ought to have been if the UK economy had followed pre 2007 trend growth, a performance the author describes as ‘dismal’. This forms the very first bullet point in the executive summary. Nail, meet hammer.
There is a little bit of sleight of hand here in that the author does not directly say this is the fault of Osborne’s austerity, indeed the report is ostensibly written to discover whether any of this poor performance is the fault of Osborne’s austerity. But it’s a funny place to start an argument. You might, for example, test the proposition that the performance of the UK economy of this period was ‘dismal’, but by merely asserting that it was, then asking how much George Osborne can be held responsible for this, you don’t even need to demonstrate Osborne’s direct failings, because — given that he was Chancellor of the Exchequer for the bulk of the period — he can’t have done very well, can he?
If, however, you really wish to describe the performance of the economy as ‘dismal’ then logic dictates you must have a comparator, and in this case, the comparator is the trend growth of the UK economy before the 2007 crash. This, to put it mildly, is simply not good enough. Trends are notoriously difficult to measure, but more to the point, by what logic can we simply erase the fact that the financial and economic crisis did actually happen? Why not calculate the trend from 1990 to 2015? Which might at least have the virtue of treating the period of 2002 to 2007 as the unsustainable bubble most economists have now concluded it was. But the bigger problem is simply that many — obviously superior — comparators are available simply in the multiple sources of available data measuring the economic performance of other economies over the same period, like say, the G7 economies.
If the UK’s economic performance really was ‘dismal’ between 2007 and 2015, then surely the only way to show this is to compare the UK to other comparable economies, not to itself in another economic period. You might just as well compare the recent UK economy’s performance with a similar length of time in the 1920s or 1950s, or perhaps — without invoking Minsky or Kondratiev — compare the upswing and downswing within the same business cycle. The comparison would be just as meaningful, or indeed, meaningless. Because by doing this, and simply assuming a deviation from a pre-2007 trend, there is a danger of leaving the reader with the impression that the financial crisis had either never happened, or that it was unique to the UK, two conclusions I remember those on ‘the Keynesian Left’ strenuously resisting with their insistence that the crisis, ‘started in America’ and that Gordon Brown’s hands were clean.
So it may be, if you look at an appropriate comparator, that the performance of the UK economy between 2007 and 2015 was not ‘dismal’ at all, but actually rather good, or at the very least, respectable. Except that in hoping to demonstrate the effect of Osborne’s austerity, it may not be a sufficiently dramatic relief upon which to project rather partisan conclusions, perhaps? Especially in the run up to an election…
In fact, up until 2012/13, the UK was routinely compared with other economies by those seeking to advance the case that austerity had been a dreadful mistake — Brad DeLong — a committed member of the ‘Krugman tendency’ — did this quite explicitly in 2012, showing that growth was indeed a bit lower in the UK than in economies with less severe deficit problems and/or which engaged in more fiscal stimulus.[vi]
I reproduce the chart he published here,
in which all economies are indexed to 2007 and their performances plotted over the following years. And yes, it does show that UK growth was — relatively speaking — low from 2009 to 2011. This is, I think, not surprising. GDP is — famously — measured by adding together C + G + I, and if C falls due to sharply restricted demand, then the normal Keynesian response is to increase G, and if you don’t increase G, something is going to give. So much, so textbook.
To that extent, DeLong’s conclusion is obvious, and in no way provides a verdict on the success or failure of austerity; it simply reveals one of its inevitable effects. Perhaps Osborne didn’t spell out that austerity implied reduced growth over the short term. But then, who remembers the last time a Chancellor of the Exchequer described some negative policy consequences — in that case ‘unemployment in the North’ — as a ‘a price worth paying’ for an important aggregate effect?
But notice this, this very same comparison if updated, no longer provides the same impact. And given both that data is available for another couple of years and that there are forecasts made for the next decade by all the major economic organisations with no particular axe to grind, I compiled a pretty basic data set from World Bank and OECD data sets using annualised Real GDP growth rates in order to replicate and extend this analysis. And if this analysis is extended to 2015, only Germany, the USA and Canada are ahead of the UK in proportion to the size of their economies in 2007. As shown here (although please forgive my poor spread-sheet chart generating skills):
More to the point, if OECD growth estimates are used (and the UK economy has higher long term growth estimates than any other G7 economy) even mighty German will fall behind the UK by 2017.
To be clear, much has been written on the effects of austerity on the UK economy after 2010, but everyone at the time conceded that reduced government expenditure would reduce growth over the short term. The OBR estimated this at about 1% a year over 2010/11 and 2011/12 [NB. This does not aggregate to 16%]. However, those who argue for austerity claim that this results in less debt drag on growth over the longer term. So someone advocating austerity, would assume both a short term negative effect on growth, and a longer term uplift in growth against the ceteris paribus assumption of less austerity and higher long term debt. Now, of course, the further ahead you look, the less reliable the forecast, so it is difficult to provide an exact quid-pro-quo analysis of the short and long term effects of austerity, but if you look at the UK’s growth performance — both historic and forecast — it seems to be conforming exactly to the growth profile an advocate of austerity would expect.
And here is where we can go back to Krugman’s chart in ‘The Austerity Delusion’, which I reproduce here:[vii]
The chart purports to show a clear relationship between harsher austerity and lower growth between the years 2009–2013, and as we can see, the German growth figure comes out at about 2.1% and the UK figure somewhere around 1.4%. It would be interesting to find out how the ‘change in cyclically adjusted primary balance was calculated’ but for now I set that aside, except to note that we shouldn’t just accept at face value that these figures are themselves beyond criticism. Krugman used data from the IMF, which I imagine is fairly comparable to the data I drew upon from the World Bank and the OECD.
The data is all historical so whatever differences are bound to be small. However, it is striking that Krugman says his data comes from the years 2009 to 2013. In the first place, I estimate German growth rates averaging over this period at 0.6% and UK growth rates at 0.3%, which doesn’t match at all. Only when I re-calculated to exclude 2009 (which is an outlier anyway) could I arrive at comparable numbers. For 2010–2013 I got the numbers 2.1% and 1.5% (roundings I suspect) respectively. We can therefore assume that Krugman is actually showing figures starting from 2010 not 2009. Is this an error in his spread-sheet, or in The Guardian caption?
More to the point, though, why stop at 2013? Data is available up to the beginning of 2015, and fairly good — politically neutral — forecasts are available for the rest of 2015, so what happens when you bring this data up to date? Unfortunately the figures for 2010 (Krugman’s ‘adjusted’ starting year) — 2015 are less persuasive for Krugman’s argument, coming in at 1.9% for Germany and 1.8% for the UK, and this is without altering anything on the x-axis. Which, to be entirely honest shows either no real difference at all, or — if we assume the trend implied by the original is correct — reflects rather well on the UK.
Now, is this just quibbling? Or is it arguable that something else is going on? It remains the case that Osborne has had less success in reducing the deficit than he forecast in 2010. Does this — in itself — prove that austerity has failed? Well, no. Of course not. Osborne’s deficit reduction forecasts were not actually his forecasts, but came out of the Office for Budget Responsibility (OBR) and they came up with their own explanation for the discrepancy, which was an unexpectedly large impact on overall tax revenue of raising the income tax threshold to £10,000 in a time of unexpectedly low wage growth. However, this has no particular implications for the debate on austerity, which is really about overall growth rates over a longer time period.
One man’s spreadsheet errors are another man’s judgement call
Economists, like other social scientists, have to make assumptions. They are dealing with messy reality, and the numbers sometimes go astray. Whenever a conclusion is reached, it is necessary to treat it with caution, not to equate a ‘degree of confidence’ with a ‘certain fact’. Economists therefore, or at least good ones, look at the whole picture, test a variety of propositions, find similar patterns across different data sets. Very rarely does an economist conclude from ambiguous data with the same brio a natural scientist might attempt.
So when the two examples I have looked at are treated in isolation, does it matter much? Does the 16% ‘growth gap’ really matter? It is, after all, not a fraudulent misrepresentation. Historical trend growth did seem to take a knock in 2007, and it is ‘true’ that if growth had carried on in the same vein from 2007, and if the financial crisis hadn’t happened at all, the UK economy might — arguably — be 16% larger in 2015 than actually turned out to be the case. But this is for the birds, or perhaps just the appendices. There is no way such a graph should — under any circumstances — form the foundation of a serious enquiry into economic performance in the run up to the 2015 election.
Then Krugman’s chart: a year here a year there, sooner or later it adds up to real partisanship. There is nothing wrong with showing the relationship between growth and austerity between 2010 and 2013 [on the proviso that you label it correctly in the first place] but this should be accompanied by some pretty heavy qualifications concerning the length of the time series, which in this case would not ordinarily be considered long enough to show anything very much, let alone a demonstrative correlation. But more than this, some comment on why the data was not extended to include the actual available data might have been welcome. Why was the time series cut off in 2013? I believe I have shown why, which is that the correlation essentially disappears if you do. The chart, therefore, reveals little more than a bit of duplicitous periodization, dressed up as an indictment of a policy that is — broadly speaking — working.
It wouldn’t matter much, but putting Jeffrey Sachs’ fairly trenchant criticisms aside for a moment, it is worth noting that Sachs is not the first economist of international standing to directly criticise Krugman for a lack of moderation. Raghuram Rajan — now Head of the Indian Central Bank, once chief economist of the IMF and one of a select few economists credited with predicting the 2007 financial crisis — openly pushed back against Krugman’s criticisms of Kenneth Rogoff and Carmen Rheinhart, in 2013 which he called ‘paranoid’.[viii]
Scratch the surface of these disagreements and you will find a deep level of disagreement, even hostility, between some of the world’s best known economists. What you will also notice is that the arguments about fiscal restraint, austerity and public debt are not — despite Krugman’s claims to the contrary — settled. Indeed, this is the most surprising thing about the whole debate, not that economists disagree, but that that a consensus on this issue could ever be finally reached, or even claimed.
Indeed there is an interesting, almost quantum-like, problem that economists are grappling with as a direct consequence of this wider debate. Namely the claim made by Keynesians that higher debt was fully justified in 2010 on account of the low interest rates that ensued, and continue to characterise debt markets despite the well established fears of advocates of austerity. The problem, that Keynesians don’t acknowledge is simply that these self-same low interest rates are thought largely to be a consequence of governments bearing down on expenditure and gaining at least a forecast control on their overall debt level. Without that, what confidence would lenders have to offer at such low rates of return?
This question remains moot, but when Keynesians breezily suggest low interest rates prove their case, they are telling — at best — a half-truth. And historically, whenever the market has been asked to speak in interest rates to heavily indebted governments with no ability or intention to control its overall debts, the result is always the same; higher interest rates, or complete withdrawal from the market provoking an inevitable default. The UK may be a long way from this point, but who wants to conduct an experiment into exactly where that point is? Keynesians, of course. Unfortunately, if Osborne’s shadow ever got the chance to conduct this experiment, markets are such that by the time interest rates do start to rise, the likelihood is the debt trajectory would have already become unsustainable, implying that the corrective measures would have to be far more drastic than anything Osborne has remotely attempted to date.
Finally, when Keynesians argue that ‘governments own printing presses’ so never need to technically default, the only serious response is just to point out that lenders do know this, and they also know that the impact of printing money will be inflationary, hence they would demand a higher interest premium. So yes, debasing the currency always remains a policy option to avoid technical default, just not a very good one. But honestly, an economist who actually advocates this as part of a prudential policy set, ought to have their credentials revoked.
But what is perhaps ironic, is that for one side in this debate — notably the ‘Krugman tendency’ — flaws in spreadsheets are central to their claims that austerity is ‘discredited’. This refers to the finding that Reinhart and Rogoff’s influential text — ‘This time is different: Eight Centuries of Financial Folly’ — was found to contain basic errors in a spread-sheet of one of the supporting working papers. When this was revealed, it caused quite a storm. People who didn’t like the conclusions of the book immediately denounced the entire volume as a fraud and pronounced the question of austerity both flawed and ideological. Which is interesting, given than when the data was corrected, nothing much changed.
True, the original research showed a very clear drop-off point in growth levels when public debt reached about 90% of GDP, and when the data was reworked, this clear drop-off point was less clear. But none of the basic conclusions were reversed. High debt is still assumed to have a negative correlation with growth. At the most basic level this is obvious, as debt requires servicing, and higher debt requires higher debt interest payments out of current income (even if we assume that interest rates remain constant across the entire debt profile), which — in economic terms — must reduce ‘G’ both in the short and long-run (because debt repayments are akin to saving, another no-no for Keynesians).
This remains true even if you concede the other element of this often made argument, that public investment produces useful returns, which in turn justifies debt, for the simple reason that this is always true, or not, depending on the quality of the public investments. The question of what level of debt is sustainable has not the slightest relationship to the value — long or short-term — of any public investment that might flow as consequence. Debt sustainability is a problem regardless of whether you build a High Speed Railway, a giant solid gold monument to the antichrist, or spend all the money on chocolate (notwithstanding the impact on gold or cocoa prices, patterns of consumer demand and possibly even inflation that these investments might produce).
Of course, Krugman is entirely within his rights to draw attention to spread-sheet errors, even perhaps his own, but it might have been better — instead of throwing around accusations of ideological bias — if he had actually spoken about the effects of correcting the data, which certainly introduces a degree of ambiguity to Reinhart and Rogoff’s conclusions but in no way ‘discredits’ what remains an important work. Not only this, but Reinhart and Rogoff’s research findings were broadly replicated by Stephen Cecchetti, Madhusudan Mohanty and Fabrizio Zampolliin, of the Bank for International Settlements in 2011.[ix] And as yet, no one has found any spread-sheet errors in their work.
About 2013…
In April 2013, The New Yorker published a long article by John Cassidy that summed up the debate at the time.[x] Its starting point was recent disappointing economic data, that Cassidy suggests merely demonstrates the obvious error of reducing public expenditure in a recession. Then he directly criticises Reinhart and Rogoff for attempting to distance themselves from both the actions of ‘their disciples’ and even, to some extent, their own words. The tone of the piece is a familiar sort of journalistic exasperation, as if the argument thread is childishly simple, and doesn’t really require elaboration. But he takes the opportunity to stick the boot in both to Reinhart and Rogoff, and — of course — to George Osborne for having the temerity to quote obviously flawed research in deference to conservative ideology.
Now, journalists can be wrong. Indeed they can be very wrong, but a run down of John Cassidy’s article titles over the last few years provides a bracing introduction to journalistic partisanship. November 2011, ‘Austerity Britain: An Experiment that Failed’. December 2012, both ‘It’s Official: Austerity Economics doesn’t work’ and ‘George Osborne’s Game: From Austerity to Cruelty’. December 2013 ‘By George, Britain’s Austerity Experiment Didn’t Work!’ note the exclamation mark. And finally, December 2014, ‘The UK’s “Mr. Austerity” Doubles Down.’
When it came to that all important election result in May 2015, Cassidy had the good grace to admit that Osborne had — rather against expectations — secured a victory, which suggested that people were generally persuaded by Osborne’s arguments. That he put this down to media bias and the stupidity of the electorate or, more precisely “virulent attacks from the Conservative media, and the difficulty of explaining Keynesian economics to non-economists” is perhaps for another essay. However, for a writer to speak of media bias who begins an article thusly: “George Osborne, Britain’s Chancellor of the Exchequer, has been called many things since 2010, when he turned the United Kingdom into a petri dish for his experiment in austerity economics: arrogant, incompetent, supercilious, right-wing, cruel, antediluvian, deluded. Until now, though, nobody has described him as a socialist. Let me be the first” invites pot-ish and kettle-ish considerations.
Nevertheless, the exclamation mark in the title to the December 2013 article is instructive, simply because that is exactly the point at which the economic consensus moved in Osborne’s favour, exactly the point at which the growth figures ticked upwards, where the critics of austerity had predicted they would not. It happened to be a little after the point at which the much mocked ‘double-dip’ recession of 2012 was revised away by the ONS, to be followed by the initial recession before that. And it was the point at which economists, commentators, and members of the general public started to ask ‘what happened to the huge explosion in unemployment predicted by all the Keynesians?’ which, of course, never materialised. There remained the small matter of slower deficit reduction than Osborne outlined in 2010, but this was a clear case of progress and delay rather than the defeat and dejection heralded by his opponents.
The exclamation mark therefore, identifies the metaphorical and psychological point at which — for the Keynesians anyway — the indicative mood failed, to be superseded by mere assertion.
Repetition is not the same thing as persuasion
Much of the commentary on austerity has subsequently taken on a tone of incredulity. It is as if the argument was won, apart from the actual winning of it. How can people not have understood? What explains our failure to explain? For example, Simon Wren-Lewis “The idea that austerity during the first two years of the coalition government was vindicated by the 2013 recovery is so ludicrous that it is almost embarrassing to have to explain why. The half-truths in this case are so flimsy they do not deserve that label … [although] some critics of austerity might have occasionally overstated their case.”[xi]
So now it is the turn of the Keynesians to correct the record concerning their own pronouncements, but note the ‘embarrassment’ at having to explain. When I think back to those poisoned arguments of 2010/11 I too, feel that self same embarrassment. And who was right? I admit there are still many problems, I concede that nothing is certain, and most of all, I admit that I may be, and may always have been wrong. But advocates of austerity were not as wrong as the Keynesians predicted, nor were they as wrong as the Keynesians turned out to be. And that — when the figures are rolling in and the votes are being counted — matters rather more that what some ideologues concluded way back in 2013, when they pronounced the contest over, and stopped the clock.
For while the austerity tortoise keeps trudging on towards the finish line of balanced books, low unemployment and solid growth, the Keynesian hare has been droning on about winning the argument for so long, that the rest of us have fallen asleep. There is economics going on out there in the real world, and not all of it is bad news for Osborne. In truth, the Keynesians constructed not so much a straw-man as a wicker-man, in which they attempted to incinerate their enemies. But George Osborne — their chosen sacrifice — seems to have lived-on and prospered, making a daily mockery of their pagan, Keynesian faith. His career will — as the saying goes — surely end in failure. But, thankfully for the rest of us, not yet.
[i] Carmen Reinhart and Kenneth Rogoff, This Time Is Different: Eight Centuries of Financial Folly, (Princeton: Princeton University Press, 2009).
[ii] Jeffrey Sachs made a number of interventions across different media but this one is probably as good as any ‘Krugman’s anti-Cameron contradiction’ on Project Syndicate, published on April 9th 2015; http://www.project-syndicate.org/commentary/krugman-us-uk-recovery-contradiction-by-jeffrey-d-sachs-2015-04
[iii] ‘Osborne wins the battle on austerity’, Financial Times, September 10th 2013. http://www.ft.com/intl/cms/s/0/03137634-1a13-11e3-93e8-00144feab7de.html#axzz3jFaaYxQ8
[iv] ‘The Weathermaker: George Osborne is vindicated, emboldened and bursting with trickery’ Bagehot Column, The Economist, January 11th 2014, http://www.economist.com/news/britain/21593417-george-osborne-vindicated-emboldened-and-bursting-trickery-weathermaker
[v] John Van Reenen, ‘Austerity: Growth Costs and Post-Election Plans’ Centre for Economic Performance, LSE.
[vi] http://delong.typepad.com/sdj/2012/04/the-recovery-in-the-g7.html
[vii] ‘The Austerity Delusion’ published in The Guardian on 28th April, 2015. This paper rehashed the conclusions of previous long essay in which Krugman had pronounced the failure and consequent death of austerity in ‘How the case for Austerity has crumbled’ published in the New York Review of Books (like The Guardian, a well-known economics journal which no particular political leaning in its editorial stance) on the 6th June 2013. http://www.nybooks.com/articles/archives/2013/jun/06/how-case-austerity-has-crumbled/
[viii] http://www.project-syndicate.org/commentary/the-declining-quality-of-public-economic-debate-by-raghuram-rajan
[ix] “The real effects of debt” by Stephen G Cecchetti, Madhusudan Mohanty and Fabrizio Zampolli, BIS, Sept 2011. http://www.bis.org/publ/work352.htm
[x] John Cassidy, ‘The Reinhart and Rogoff Controversy: A Summing Up’ in The New Yorker, April 26th 2013. http://www.newyorker.com/news/john-cassidy/the-reinhart-and-rogoff-controversy-a-summing-up
[xi] Simon Wren-Lewis, ‘MediaMacro myth 6 : 2013 recovery vindication’ on blog Mainly Macro 26th April 2015 (2 weeks before the general election): http://mainlymacro.blogspot.co.uk/2015/04/mediamacro-myth-6-2013-recovery.html
