How a Greek drama became a global tragedy

Maybe that title is too strong, but there is an arguable case that what happened to Greece in 2010 was crucial in the move to austerity not just in the Eurozone, but in the UK and US too. As most reasonable people now recognise that the global move to austerity was a terrible mistake, understanding what went wrong in Greece is important. By this I do not mean how Greece came to behave fiscally in a total irresponsible way, interesting and important though that is. These things sometimes happen, but they do not usually have global consequences. What is more important is how the Eurozone and IMF subsequently handled events, which helped turn a Greek crisis into a Eurozone crisis and more.

These events have recently been analysed by the IMF. (Subsequent references are to paragraph numbers.) All credit to them for publicly and critically analysing their role in this affair. In terms of the substance, the facts of the case are not really in dispute. There were two feasible responses to the true fiscal numbers as they began to be revealed by the Greek government. [1] The first was forgiveness, in the form a large fiscal transfer from other Eurozone governments. These governments might have noted that the Greek people did not intend its previous governments to act in such a profligate and deceptive way, and that the Eurozone had failed to put in place effective institutions to stop it happening, and as a result they could have (one way or another) paid off a large part of Greek debt as a gift. That was never likely to happen, and it did not happen. [2]

So the other feasible option was default. As the report also notes (para 55), a number of Fund programs since 2000 had started with some ‘private sector involvement’. Yet initially the Eurozone ruled this out, as the report makes clear and which coverage of the report has widely noted. Why was it ruled out by the Eurozone? The charitable explanation was a concern that once the default possibility became reality, markets would turn on other vulnerable governments. Predictably, they did this anyway. A rather more base explanation was that with default the banks in other Eurozone countries might lose a lot of money, and become even more fragile. There may also have been a bit of pride.

So what we got was a transfer of ownership of a large part of the Greek debt from the private sector to other Eurozone governments, and crippling austerity for Greece. This was not feasible: default was just postponed, but only partial default on the remaining privately held debt. Additional austerity was imposed, and Eurozone governments swore there would be no default on the debt they now owned. And so it goes on.

So why did the IMF not point out that the original plan was not feasible, and refuse to participate? That would have been a hard political call to make, but the IMF is practiced in making such calls when the economics dictates. The report talks about failures in Greek implementation, but it also recognises that the subsequent turnaround in Greece’s underlying fiscal position has been dramatic, so its difficult to believe that plans to do any more should have been taken seriously. The real problem, as I note here, was that projections for the Greek economy were hopelessly optimistic.

Why were they too optimistic? As the IMF acknowledges (para 41) and has acknowledged before, it underestimated the impact on the economy of austerity. It got the multipliers wrong. [3] Yet while the report draws a number of lessons from the episode as a whole, I cannot see any analysis of how this fundamental, and arguably critical, mistake was made.[4] Talk to most people who know anything about the basic theory involved, and they will tell you that when nominal interest rates are fixed, the starting point for the government spending multiplier should be one. [5] In a financially crippled economy there are likely to be a lot of credit constrained individuals around, so the tax or transfer multiplier is also likely to be a lot larger than normal. So numbers based on looking at the past evidence which includes times when monetary policy was able to counteract the impact of the fiscal change were always going to be seriously wrong. You didn’t need to be a nobel prize winner to work this out, you just need to ask people who have worked through the theory of fiscal multipliers. [6] Knowing a little bit about the IMF, I would love to know how this mistake came to be made. [7] [8]

The IMF document is not just about deriving lessons looking back. It also speaks to a real issue that should be being debated in the Eurozone, and that is what are the best institutional arrangements for the lender of last resort to Eurozone sovereigns (hereafter SLOLR)? As Paul DeGrauwe has pointed out, the last few years tell you that you need a SLOLR to prevent a bad equiilibrium where debt crises are self fulfilling. But should the SLOLR be the central bank, other Eurozone governments or the IMF?

What the IMF report clearly shows is that it should not be other Eurozone governments. A good SLOLR needs to be effective (in having the fire power to prevent a bad equilibrium), but it also needs to be able to know when not to intervene, but instead allow default to happen. Eurozone governments failed on both tests: they were never willing to devote enough resources to ensure they were effective, but with Greece they failed to see that default was inevitable. They have the wrong incentives to make the right decision.

You might think that the Greek episode also tars the IMF with the same brush. But as Karl Whelan points out, the ECB does not come out of this episode very well either. Which is a little worrying, because with OMT the ECB is now the SLOLR. It does have back-up, but from those very same governments that got it wrong in the case of Greece. So in a future crisis, does the ECB have the right mechanisms in place to know when it should pledge to buy government debt and when it should do nothing to prevent default? While the IMF deserves credit for being publically self-critical, I cannot help but ask how long will we have to wait for a similar self-analysis of the ECB’s role in this affair?  

[1] As the IMF notes, in October 2009 a new government revised up its estimate of the 2009 deficit from 4% to 12.5% of GDP, and even this was 3% below the final estimate.

[2] In technical terms, this would have amounted to there really being just one set of Eurozone fiscal accounts. If one government gets its people into trouble, the other governments will raise taxes or cut spending to prevent default. When I was writing this paper with Campbell Leith, we ruled out this possibility as unrealistic, but I never imagined that judgement would be tested so quickly.

[3] Assuming a multiplier of 0.5 rather than 1 does not account for all the forecast error. But what remains does sound a bit like ‘we expected the confidence fairy but she did not turn up’. For example, errors “reflected the absence of a pick-up in private sector growth due to the boost to productivity and improvements in the investment climate that the program hoped would result from structural reforms.”

[4] Am I overemphasising this point? Consider this hypothetical. Using the correct multipliers, the IMF just cannot show that the original no-default programme adds up. (As the report makes clear, the actual projections only just did so.) The IMF tells the rest of the Troika that it cannot participate without some initial default. Fearing the impact that such news would have, the Europeans recognise that some default is required. Rather than spending the next year or so putting off the inevitable, the Troika instead focuses on establishing that Greece is a special case, and ensuring there is adequate support for other periphery countries on beneficial terms without crippling austerity (using the right multipliers). The rest of the world increasingly sees Greece as an isolated incident and not the beginning of a worldwide debt crisis. The other possibility, explored by Barry Eichengreen (HT Brad DeLong), is that the Greek government could have insisted on default.  

[5] In a closed economy at the ZLB it is almost certainly above one. In a monetary union, even if the government spending is entirely on domestic goods, in a model with unconstrained consumers it will be below one, but with credit constrained consumers it can be above one (see this paper by Fahri and Werning, and a forthcoming post). Ideally, from a macro point of view, you would choose to cut government spending on goods produced overseas, where multipliers could be negative. The amount that Greece spends on arms, the extent to which it has been part of the fiscal consolidation programme, and the fact that many of these arms come from other European governments is highly controversial.   

[6] Doing the analysis is what is crucial here. In this particular case it just so happens you could have asked a well known nobel prize winner who had also done the analysis. However just asking an academic who has won a nobel prize for macroeconomics but who had not done the analysis could get you into trouble.

[7] Even though the report notes that the assumed multipliers were too low, it also comments (para 46) that “The adjustment mix seems revenue heavy given that the fiscal crisis was expenditure driven”. Yet if you want to protect demand, you should (in the short run) focus on tax increases rather than government consumption or investment. So even within this document, the basic macroeconomic theory behind fiscal consolidation has still not been fully appreciated.

[8] Of course you can say that the numbers were chosen to fit the politics, and the real lesson is that the head of the IMF should not be a past and/or prospective French politician. But those within an organisation like the Fund should try and make it as difficult as possible for politics to overrule economics in this way.