One of the best things about conference season is that it gives you a good insight into which of last year’s policies each party thought were popular, and which need to be quietly binned. The result? Both main parties seem to agree that austerity is a vote winner. Consider: Labour pivoted to the ‘cost of living crisis’, whereas the Conservatives think voters austerity so much that they have decided to propose semi-permanent austerity, spending and tax cuts be damned. George Osborne sees vindication in recent output growth. However as Simon Wren-Lewis notes, a single quarter’s good performance is weak success at best, pointing out that:
‘We could close down half the economy for a year. The next year economic growth would be fantastic. Only a fool would argue that this showed that closing down half the economy for a year was a great idea.’
This is a more general problem. In politics austerity seems like a debate between partial truths, with recent events seeming to have austerians carrying the day. But empirical economics has spent the last five years testing the claims of austerity, and fairly emphatically found the reverse.
Looking back on the 2010 general election campaign and after, the argument rested on two points:
1) Multipliers on government spending were small, so fiscal contraction would barely harm overall demand
2) Austerity could be expansionary for growth due to business confidence effects,
This was taken to mean that the overall costs to growth were small.
Government spending multipliers
These are the fall in total demand associated with a given pound of tax rises or spending cuts. Put simply, small government spending multipliers mean austerity will be relatively painless. When the OBR and the IMF forecast the economic effects of coalition spending plans, they used a multiplier of 0.4-0.5. However the balance of evidence suggests under current circumstances, multipliers are about three times as high, around 1.5. We know this because of a very public mea culpa from the IMF – their chief economist, Olivier Blanchard, released a study earlier this year working out multiplier values based on how overoptimistic IMF forecasts had been. These estimates are also consistent with the best historical evidence we have. So when we see that UK growth performing very badly relative to its OBR forecast in 2009, this is completely consistent with much more severe than anticipated effects from austerity (on which more below):
So why were forecasters so over-optimistic in 2009? Normally monetary policy can offset the effects of a fiscal shock by cutting interest rates. Given that they have been at their zero lower bound since 2008, monetary policy has essentially been exhausted as a tool. The jury is still out on quantitative easing, but most studies over the last few years give it a much smaller and less certain impact than conventional rate cuts.
The complementary argument put out (mostly by George Osborne) in 2010 was about austerity boosting growth, through so-called ‘confidence effects’. The basic idea was that households and businesses were cutting back spending because of uncertainty over high public debt, potentially higher tax rates, or a sovereign debt crisis. Regardless of whether or not this is a sound analytical argument (and it is difficult to more than assert a theory of the mass psychology of households and businesses) it had some empirical backing from a paper by Alberto Alesina and Silvia Ardagna, which the Treasury explicitly relied on to make their case. Alesina-Ardagna claimed that fiscal consolidation led to higher growth. However they did so on the basis of some quite flimsy statistical work. In a nutshell, if their result holds at all it is when output growth is strong (i.e. the opposite of the UK recently), and interest rates can be cut by central banks. The IMF debunked it in 2011 on that very basis, after controlling for whether fiscal contraction takes place at the peak or trough of the business cycle. This explains the massive underperformance of the Euro Area and the UK versus less contractionary countries:
Costs to growth
Given the above, then, it’s no surprise that the best estimates of the costs of austerity have been huge. Alan Taylor, one of the world’s pre-eminent macroeconomic historians, attempted to estimate the total cost to output of budget contraction since 2009 along with his co-author Oscar Jorda. The result? Austerity probably cost the UK 3% in terms of lost output. By the very nature of counterfactuals, we can never know if this figure is correct. But serious criticisms of the paper, including by the authors themselves, have tended to focus on the fact that they fail to control for interest rates being at the lower bound, when monetary policy is exhausted. In this light, 3% is probably a conservative estimate of the total cost of austerity. Given that there has been no comparably rigorous piece of contradictory analysis, this should be, at the very least, our base case. Jorda and Taylor underline this in a graph:
As important as the work above is the total absence of the reverse – serious and unrefuted empirical economic research coming out in favour of the pro-austerity arguments has been slim to none. Given the debates that sometimes mark out macroeconomics as a discipline, this speaks volumes.
So does writing against austerity equal staking out a political position? I would argue exactly the opposite. At least in terms of the cost to growth, the line that austerity is disastrous is apolitical. Rather, from a macroeconomic perspective it is a firmly refuted position, like the gold standard, or money supply targeting. And yet it doesn’t seem as if key political actors have updated their beliefs to fit this – it is for this reason that Simon Wren-Lewis likens the debate to ‘climate change denial’.
Economics and politics seem to have reached opposite conclusions to the same debate. What does this say about the impact of economics on policymaking? And if the evidence isn’t enough to sway this particular political discussion, are we going to be permanently worse off?
Update: Turns out the OBR broadly agrees with Jorda and Taylor