How the lessons from austerity have not been learned

The UK and the Eurozone are both vulnerable to the next recession, but both politicians and central bankers think each other should deal with it.

I don’t want to talk about the likelihood of a recession in the UK, US or Eurozone. Forecasting is a (necessary) mug’s game, where there are just too many variables to make anything like an accurate prediction. It is worth outlining the risk factors, and Grace Blakeley does an excellent job here. Instead my concern is the vulnerability in both the UK and the Eurozone to the impact of a recession if it happens. This vulnerability was clearly illustrated by the mistakes made after the Global Financial Crisis, yet in many ways the lessons of that failure have not been learnt.

Most people know the story of austerity after the Global Financial Crisis (GFC). In the UK the negative impact of the GFC was so severe that even cutting interest rates from around 5% to 0.5% was not sufficient to counteract its impact. As a result the Labour government in 2009 undertook various fiscal stimulus measures. They, together with lower interest rates, succeeded in stopping the fall in output, but by 2010 the signs of a recovery were still fragile. The new Coalition (Conservative and Liberal Democrat) government decided to focus on the rising budget deficit rather than the recovery, and undertook a large fiscal contraction in what became known as austerity.

The consequence of UK austerity was the slowest recovery from a recession in centuries. Here is a nice chart from a recent report by James Smith of the Resolution Foundation, which clearly illustrates the extent of the weak recovery.

Employment eventually recovered, but at the cost of an unprecedented fall in real wages. James Smith gives some evidence to suggest that the reason employment got off comparatively lightly but wages suffered by much more than we might expect was the 2008 sharp depreciation in sterling. This allowed firms to respond to the recession by keeping wages low, whereas in recessions in which sterling did not fall firms resisted nominal wage cuts and so had to resort to cutting jobs.

The idea that austerity was essential to reduce the deficit is simply wrong. It undoubtedly was a large factor in the weakness of the UK recovery. The tightening of fiscal policy has continued until today, with the consequence that interest rates have had to stay low to offset this fiscal tightening. The net result is that instead of rates being near 5% as they were before the GFC, they are below 1%. As James Smith points out, interest rate cuts in previous recessions have ranged from three to ten percent. This means that conventional monetary policy has almost no room to counteract a new economic downturn if one came.

The Eurozone is in an even worse position. Their history is of two recessions since the GFC, the second of which was largely caused by fiscal tightening as a result of the Eurozone crisis of 2010-12. The last time core inflation in the Eurozone touched 2% was in 2008, and it is currently around 1%. (More details from Frances Coppola here.) Interest rates set by the European Central Bank (ECB) remain at their lower bound. If a new recession happened, caused for example by a disruption in trade due to Donald Trump, conventional monetary policy would be unable to do anything about it.

Of course central banks in the UK and Eurozone still have various unconventional monetary policy tools. But the clue to their reliability at ending a recession is in their name. They are unconventional because they have only been used since the GFC, so we have limited evidence on their impact. It is like having an accelerator on a car where how far you have to push your foot down varies from second to second. You will end up driving slowly, which in economic terms means a prolonged recession.

All this is now largely understood by central bankers. All have said in one place or another that they will be relying on fiscal stimulus to help counteract the next recession. The ECB needs fiscal stimulus right now to get out of the last one. Yet fiscal stimulus is in the hands of politicians and not central bankers, and many of the politicians and political parties that were crucial in implementing the austerity that hit the post-GFC recovery are still in power.

There is therefore a danger that the policy of fighting the next recession will fall between two stools. Central bankers will say, in their own quiet and politically sensitive way, that they are not equipped for the task, but politicians may be deaf to these messages and will once again start worrying about deficits that inevitably rise in an economic downturn. To say more, we need to differentiate between the UK and the Eurozone.

In the UK some may think that with a new Prime Minister the problem of austerity has disappeared. In order to get elected both candidates have promised all kinds of tax cuts or spending increases. But as I have argued recently, what we are seeing here is what economists call deficit bias: the tendency to borrow just for political gain. Worse still, if the borrowing is mainly for tax cuts (including tax cuts for the rich), there is a danger that it is part of a strategy called ‘starve the beast’, which involves increasing the deficit with tax cuts and then demanding spending cuts to bring the deficit under control.

The upshot is that a Tory leader wanting to spend and cut taxes to please party members provides no guarantee that they will undertake effective fiscal expansion in any future recession. Neither of the Coalition partners has apologised for the mistake of austerity, and we have no reason to believe that they wouldn’t do it again in any future recession. The only major party that has a fiscal framework that would automatically move to fiscal expansion when interest rates hit their lower bound is Labour.

In the Eurozone there is also too little recognition among senior politicians that fiscal stimulus is required when ECB interest rates are at the lower bound. This is why Eurozone inflation is still below target. The OECD point to a crying need for more fiscal stimulus. Germany in particular has a great need for additional public investment, but is restrained by a fiscal rule that is worthy of an economic stone age. Efforts to create a Eurozone budget that could act in a countercyclical way have also been blocked by politicians, despite support from the ECB.

We can hope for a change of political attitudes in both the UK and Eurozone, but central bankers should not be content with hope. They have been delegated the task of stabilising the economy, and if they fail to complete this task in economic downturn after downturn many will come to believe that delegating monetary policy to central banks was a huge mistake. In addition, it is not the case that all central banks can do when interest rates hit their floor is unreliable types of unconventional monetary policy.

A fail safe way for a central bank to bring a recession to an end when interest rates are at the lower bound is to create money and give it directly to citizens. It could also create money and give it to borrowers by subsidising borrowing rates. The former is called helicopter money, a term due to Milton Friedman, and would require cooperation from government. The latter has been undertaken by the ECB in the past (see Eric Lonergan here), and so could be done to a greater extent without involving government. In essence it involves cutting interest rates on borrowing to well below the lower bound, but keeping rates for savers at the lower bound, and funding the difference by creating money.

Central banks in most of the major economies have been happy to create money during a recession, but nearly always this money has been used by central banks to buy assets. The impact on the economy is then difficult to predict, because no one's income has increased and the interest rate for borrowing has not fallen significantly. Giving the money directy to people rather than buying assets would have a direct and more predictable impact in stimulating the economy, as Frances Coppola argues in her new book.

So why do central banks not do this? There are two major reasons. First, they worry that increasing people's income is the job of elected governments, although I would argue that it is central bank's job to stabilise the economy if the government does not. (See my article with Mark Blyth and Eric Lonergan for more on helicopter money.) Second, if they create money to buy assets, when the economy recovers they can if necessary take money out of the economy by selling those assets. If they give that money away they wil not have those assets to sell. However this problem could be dealt with by governments guaranteeing the supply of assets a central bank needs.

Besides these arguments, I think there is a third argument why most central banks have not proposed doing these types of measures in a major way, and that is conservatism with a small c. The problem is that, if politicians unprepared to undertake fiscal expansion in a recession remain in power, that conservatism may be very costly both to us and to central banks themselves.

Another lesson of the GFC unlearnt: the Consensus Assignment is dead

Martin Sandbu recently responded to critics of an earlier piece of his arguing that central bankers could and should have done more to tackle the aftermath of the Global Financial Crisis. I stress aftermath here, because I have no doubt that central banks (including the Bank of England) were culpable in both ignoring warning signs before the crisis, and in the UK and Eurozone not reacting quickly enough to the consequent recession. (It is extraordinary that on the MPC only Danny Blanchflower understood what was going on, and I would argue that part of the reason for this was the primacy of the inflation target.)

It is also obvious that the ECB were wrong to raise rates in 2011 and not to introduce QE much earlier. That the UK almost raised rates in 2011 is not reassuring, and suggests they did take their foot off the peddle over that period. I would also argue that the ECB were very wrong to wait until September 2012 to introduce OMT.

In the UK and US I do not buy the argument that no further stimulus was needed.

UK unemployment rose from around 5% to around 8% in 2008/9, and stayed at that level until it began coming down in 2013. US unemployment was also above 8% until 2013. Both economies needed more stimulus in 2009, and in its absence in 2010 and so on. To think otherwise means you are placing too high a weight on temporary changes in inflation and too low a weight on the costs of the recession. 

Where I think I might disagree with Martin is that this stimulus could have reliably come from monetary policy. A good policy instrument is one that has a reliable impact on demand, and the only reliable monetary policy instrument that fulfills that criteria is short interest rates. Central banks could have done more QE, or they could have reduced rates below the Effective Lower Bound (ELB), but they wouldn’t have known how much to do. They might have got there in the end, but extra years of unemployment and probably a permanent hit to output through hysteresis were an avoidable cost.

The biggest mistake central banks made was not to recognise this and be honest with the public. They should have said, clearly and repeatedly, that once rates hit the ELB monetary policy was no longer the most reliable instrument to stabilise the economy, and fiscal policy should be used. This does not break any implicit concordat about not commenting on fiscal policy (which most central banks break anyway), because the statement is about a delegated authority being honest with the public about whether it can reliably do its job..

The fact that central banks in the UK and Eurozone didn’t do that may reflect dishonesty or it may reflect negligent ignorance. The fact that options like QE existed may have allowed central banks to convince themselves that they could still do the job assigned to them, and it discouraged them from being honest with the public. I say negligent ignorance, because instrument reliability is pretty basic stuff.

Martin writes

“Besides, there was broadly shared understanding among macroeconomists and central bankers of the best division of labour. Fiscal and budgetary policy should be set to achieve microeconomic and distributive goals, and the desired share of the state in the economy; while monetary policy should take care of stabilising aggregate demand.”

This is what I call the Consensus Assignment, and as the name implies it was certainly the consensus among mainstream macroeconomists before the 1990s. But the experience of Japan’s lost decade where they also had interest rates stuck at the ELB began a process of rethinking. By the time the GFC came around many macroeconomists had realised that there was an Achilles Heel in the Consensus Assignment. Fiscal stabilisation was still required when interest rates hit their ELB. That is why we had fiscal stimulus in 2009.

The importance of this cannot be overstated. The policy consensus in 2009 was that fiscal stimulus was required, because monetary policy was not enough. This consensus didn’t evaporate in 2010. What overrode it was mainly politics - what I call deficit deceit. There was also a bit of panic in some quarters caused by the Eurozone crisis. However a majority of academic macroeconomists continued to believe that further fiscal stimulus was required, and that majority got steadily larger as time went on.

Which means that the Consensus Assignment that Martin talks about is dead, or at least dead until monetary policy makers can agree some form of fiscal delegation (e.g. helicopter money) with governments. As this paper from the Boston Fed points out, downturns where interest rates hit their ELB are likely to become the new normal, but policymakers have not adjusted to this. (The exception is Labour’s fiscal credibility rule.) Quite simply, most policymakers have not learnt a major lesson of the Global Financial Crisis.

The IMF as a transmission mechanism for academic knowledge

In my recent post on the ‘biggest policy mistake of the last decade’, I emphasised the irrelevance of the academic consensus on austerity if politicians did not want to listen. It was, inevitably, a picture painted with a broad brush.

I did not discuss, for example, an element that should form part of the transmission mechanism for academic knowledge but didn’t, and that is European central banks. As I have discussed here, these central banks are full of economists applying state of the art macroeconomic knowledge, so they should be a source for the current academic consensus. But these central banks are also very hierarchical, and if the senior staff want to give out a different message they can. In Europe that message was that austerity was necessary, and worse still that the lower bound for interest rates was no impediment to their ability to control the economy.

This was a serious mistake for two reasons. First, central bank leaders were going against the knowledge that their own economic models and analysis gave them. Second, their implication that the lower bound for interest rates didn't matter was not only very wrong but also encouraged politicians to continue with austerity.

But there was a perhaps surprising route by which the academic consensus did get through, and that was the International Monetary Fund. The IMF itself wavered on austerity. At first (before 2010) it encouraged coordinated fiscal stimulus. As the Eurozone crisis began to unfold it changed its mind, and advocated austerity. But this did not last that long. I remember visiting the IMF in September 2012, and being told of empirical work by their Chief Economist Olivier Blanchard and Daniel Leigh that suggested multipliers might be much larger than the received Fund wisdom at the time. It was nice for me, because one of the talks I gave was why from a theoretical point of view multipliers might be large when interest rates were stuck at their lower bound.

This was not the only piece of Fund work that undermined the case for austerity. This analysis questioned the empirical case for expansionary austerity, as I discussed here. Economists at the IMF also showed clearly how unusual the behaviour of government spending after the Global Financial Crisis was compared to previous recoveries: austerity, far from being the norm, was an untried experiment. Indeed I think it is fair to say that if you wanted a source of empirical analysis on the impact of austerity, the IMF was your first port of call.

As Ben Clift discusses here, the IMF have also pioneered analysis of how inequality, and perhaps even large financial sectors, may be bad for growth, and much more that you would not have expected from the IMF of the last century. But he also points out something I emphasised in a post I wrote after my visit. The IMF is extremely heterogeneous. Alongside more modern views of the role of fiscal policy you will also find traditional fiscal hawks. The IMF also has its hierarchy with more political masters, but the difference is that at the IMF today there is no rigid control of what gets published by its economists.

For example, the IMF have an Independent Evaluations Office, which appears to be lead by economics rather than politics and which is often critical of IMF practice. I noted here, for example, a 2014 analysis of austerity, which criticised the support the IMF gave to austerity from 2010. The report essentially suggested that parts of the IMF had been panicked by the Eurozone crisis, which also presumably gave the fiscal hawks in the institution the upper hand. The report also explains why this panic was unwarranted given what we now understand about the Eurozone specific causes of that crisis, and this together with the Blanchard and Leigh analysis helped turn the tide against a belief in the virtues of austerity in the IMF.

All this IMF work was clearly very helpful to those economists like myself who were arguing against austerity at the time. It didn’t change policies in the UK and among Republicans in the US because those policies were ideologically based. I doubt it had much impact in Germany either. However it might be possible to argue it had some influence in softening the line taken by the EU Commission. If you look at the OECD’s estimate of underlying primary balances, 2013 was the last year of fiscal contraction in the EU as a whole.

Should Eurozone central bankers keep quiet about fiscal policy?

The Governor of the Irish Central Bank, Philip Lane, has called for higher taxes on savings and investment (property taxes). Frances Coppola, in a very clearly argued and well informed article, says this “attempt to dictate to the Government is a serious threat to Irish democracy.” Is it?

The conventional view is that there is an implicit quid pro quo between independent central banks (ICBs) and governments. Governments do not interfere with monetary policy and in exchange central banks do not suggest to governments what fiscal policy, or any other non-monetary policy, should be. It is important to note that this implicit quid pro quo has never seemed to extend to central bankers in the Eurozone, who have frequently pontificated on matters outside monetary policy, so in that context Lane’s comments are not out of the ordinary. But that does not make it right or wrong.

I have argued in the past that the situation where interest rates are at their lower bound requires ICBs to tell governments what to do, or at least say they can no longer do an effective job unless the government undertakes fiscal expansion. In that context I rather like the suggestion in a paper by Ed Balls and colleagues [1] that at the ‘Zero Lower Bound’ (ZLB) central banks should be mandated to write every three months to the government suggesting how much stimulus they think is required for the economy to get of the ZLB.

The situation for central banks in Eurozone countries is similar in the sense that cannot change interest rates (which are set by the ECB). They do, however, as Frances points out,  have responsibility for the health of the financial system in their own economies which requires macroeconomic expertise. So it seems reasonable to apply the quid pro quo to financial supervision by an ICB in a Eurozone country in much the same way it is applied to a ICB that is not in a currency union or fixed exchange rate regime.

But is the quid pro quo sensible in the first place? There is a real concern that an independent central bank might be intimidated by government politicians telling them what to do. That is understandable, as part of the whole rationale for ICBs in the first place is that their independence from politicians gives them additional credibility that they are not acting for political reasons.

The argument for a quid pro quo is that independent central banks should not abuse their authority to interfere with political decisions that have nothing to do with monetary or financial (macroprudential) policy. Let me put a counterargument that I sketched out in a different (UK) context here. While it is obvious no central banker should give advice on who should win the next General Election, that is not what we are talking about here. We are talking about issues were the central bank has some expertise.

Given that, it would be strange indeed if the central bank was prohibited from telling the government what its expertise suggested. You could see the outcry if the ICB guessed a recession was on its way, but kept that knowledge to itself and it subsequently turned out it was right. So in this case Lane would undoubtedly tell the government his views. What we are therefore talking about is secrecy. Is it best that this expertise is kept from the public so as not to embarass politicians when they ignore it? That does not sound so clever. More generally, the lesson of the last ten years is not one where governments would have made good macropolicy if only they hadn't been intimidated by central banks, but rather than in Europe central banks gave bad advice.  

I wanted to talk about this particular case because it perfectly illustrates this dilemma. Back in the early 2000s, while he was still a lowly academic, Philip Lane was one of the few public voices suggesting that the Irish Republic needed to use fiscal policy to cool down its economic boom. He was ignored, and the result was a financial and economic crash. He therefore not only has expertise but a reputation for being right on the very issue he is giving his advice about. Is it really better for Irish democracy that this advice is kept secret from the public?

[1] Balls, E, Howat, J and A Stansbury (2016) 'Central Bank Independence Revisited: After the financial crisis, what should a model central bank look like?' M-RCBG Associate Working Paper No. 67

The advantage of a central bank not being ‘ahead of the curve’

Imagine the following economy. Growth has been strong for a number of years: 2.7% 2014, 3.8% 2015, 3.1% 2016 and is expected to be above 3% again in 2017. The OECD also think the output gap is positive i.e. output is above the sustainable rate. Inflation was bobbing around zero for a few years, but since 2016 has gradually crept up to the target of 2%. It was just over 2% in the summer, but dipped just below target in the last two months. Fiscal policy is broadly neutral, and is expected to remain so. The unemployment rate is still slightly above 6%, but the average rate since the crisis in the early 1990s is over 7%.

We are talking about the very healthy Swedish economy. An economy where inflation is at target and some experts think the economy is running hot. What level do you think the Riksbank, Sweden’s independent central bank, has set its interest rate at? The answer is -0.5%. What is more the general expectation (the Riksbank publishes its interest rate forecast) is that rates will not start rising above -0.5% before min-2018. In addition, the Riksbank is undertaking its form of QE.

What can explain this dovish behaviour? Central banks are supposed to be inflation averse, and elsewhere they talk about the need to ‘normalise’ rates the moment the economy starts recovering, so that they are ‘ahead of the curve’. Part of the answer lies in the past. I have told the story in real time (here, then here, then here), so just a short synopsis this time. The Riksbank started raising interest rates from its then lower bound of 0.25% towards the end of 2010, because they were worried about a potential housing bubble. Rates continued to rise to 2%, but inflation began to fall, and did not stop until it hit zero at the end of 2012. There was no growth in GDP in 2012. The eminent macroeconomist Lars Svensson resigned from the Riksbank in protest at this departure from inflation targeting.

Sweden: Consumer Price Inflation and Short Interest Rates, plus forecast (from OECD Economic Outlook)

In 2012 the Riksbank admitted their mistake, and started lowering rates to the new lower bound of -0.5%, where they have been since the beginning of 2016. Having made the error of prematurely raising rates once, they are in no hurry to risk doing so again.

The Swedish experience I think illustrates a number of points that could also be applied to other central banks.

1. Macroprudential policy is the way to deal with financial instability like housing bubbles. Using the interest rate instead can be very costly. Controlling inflation should be the only thing short term interest rates are used for.
   
   
2. Low, below target, inflation can be very sticky. Swedish interest rates went below zero at the beginning of 2015, but inflation only went above 1% just under 2 years later, and this despite growth in GDP of 3.8% in 2015.
   
   
3. Inflation went above target in July, August and September of this year. But the central bank, looking at the basic determinants of inflation like wage growth relative to productivity growth, held their nerve and kept rates at -0.5%. Inflation fell back to 1.7% in October, and has stayed below 2% in November.
   
   
4. The argument that interest rates must be raised above their floor so that central banks are ‘ahead of the curve’ has not been as influential in Sweden as it has been elsewhere.

It could still go wrong for Sweden, but if it does not then the Riksbank's 2010/11 mistake may have a silver lining. By making the central bank extremely dovish, they have allowed the Swedish economy to grow strongly and unemployment to fall substantially. Perhaps the mantra adopted by other central banks of needing to be ahead of the curve in terms of early ‘normalisation’ is not such a clever idea.

Could independent central banks be advisory?

With fiscal councils (or Independent Fiscal Institutions) now commonplace in advanced economies, a natural question arises. Why are all these councils advisory, while independent central banks have control over monetary policy? For fiscal policy we seem to have delegated advice [1], while for monetary policy we have delegated control. In this post I want to focus on control over how policy instruments are changed, and not control of the goals of policy. For clarity assume that governments still control the ultimate goals of monetary policy (e.g. an inflation target) and fiscal policy (e.g. a target for the deficit in 5 years time).

As fiscal councils are the less familiar, it is natural to try and answer this question by asking why fiscal councils are not given control over fiscal policy. I am, of course, not talking about controlling the detail of government spending or taxes, but instead setting a target for the projected deficit which governments should aim to achieve in a budget. There are lots of potential answers to that question, which I have written about elsewhere.

However we could ask the question the other way around, and I cannot remember anyone asking it this way. Why are there no independent advisory central banks? In the UK, for example, imagine having the MPC meeting, and then immediately advising (in secret for a short time) the Chancellor of their recommendation for interest rates. The Chancellor would very quickly (within an hour or day?) decide whether to accept that recommendation or do something different. After that, the decision and the MPC’s recommendation would be announced.

Two straightforward points. First, a system of that kind could only work in the US if Congress gave the President the power to accept the Fed’s recommendation or impose the President’s own decision: perhaps not something we would want to contemplate right now. In the Eurozone the ECB would have to give recommendations to Ecofin, which might make it both impractical and perhaps undesirable. Second, this form of delegation is obviously weaker than giving complete control to the central bank, and that in itself may be a reason why it is not adopted.

Nevertheless, for a country like the UK, it would be a mistake to underestimate the political pressure the Chancellor would be under to accept the central bank’s public advice. The Chancellor or Treasury minister would be entirely responsible from deviating from the recommendation given to them, and if it went wrong they would incur a considerable political cost. In these circumstances, it would be understandable for governments to reason that there was little to be gained from having the power to overrule central bank advice. They would get it in the neck if they overruled this advice and turned out to be wrong, but equally if the MPC make mistakes they would also have ultimate responsibility for accepting this advice. If in practice nearly all of the time they are going to accept the central bank’s recommendations, why not give them complete control so that at least you are not implicated when things go wrong.

If this reasoning is correct, it raises a difficult question for those who argue against central bank independence but still accept monetary policy’s primary role in stabilising the economy outwith the ZLB. Of course many governments used to be happy to control monetary policy, as long as the advice they were getting was secret. But if that advice is public, as surely we all agree it should be, would even formally advisory central banks start to in effect control monetary policy because governments would never incur the risk of going against their advice? In which case, why so much fuss about independent central banks that do control monetary policy being undemocratic? I stress again that I’m talking about control of month to month interest rate changes, and not the goals of monetary policy (inflation targets or NGDP targets). I think those should be democratically decided (as in the UK, but not the US or EZ), and that central banks should be accountable in a meaningful way if they do not achieve these goals. But for the day to day business of setting rates, I cannot see that much would be gained by putting those under democratic control. 

[1] In the absence of delegating advice to an independent institution, advice would come from the the internal civil service.