Winner of the New Statesman SPERI Prize in Political Economy 2016

Thursday, 22 May 2014

How to avoid the austerity mistake next time

In recent posts (e.g. here) I have been rather pessimistic about what might happen the next time we have a large negative demand shock that puts interest rates to their zero lower bound (ZLB). In theory fiscal policy can come to the rescue, and we can avoid a severe recession. But many of the reasons that did not happen this time persist. The political right will see rising debt as a chance to shrink the state. The financial sector will argue a funding crisis (the ‘bond vigilantes’) are just around the corner. Central banks will do what central banks have nearly always done: advise either privately or publicly that we need fiscal restraint.

We can hope that, as recent lessons are learnt, economists speak with more of one voice on what should happen. I like to think this will occur to some extent, as the influence of the anti-Keynesians fades, but the importance of ideology in the discipline is such that economists will never be united on this issue. And in any case, will the majority of economists have more influence than the financial sector and the central bank? I think not, particularly if there is another Greece.

What I think macroeconomists can do is start talking about fiscal rules, and the importance of fiscal councils. This is what Jonathan Portes and I try to do in a new paper (here or here). The paper discusses a number of issues involved in formulating fiscal rules (which I will write about in subsequent posts), and it also stresses the importance that fiscal councils can have in supporting (or modifying) these rules. It also has something very clear to say about what should happen the next time we experience a large negative demand shock. (With unchanged inflation targets and perhaps a lower natural real interest rate, this may happen rather more often than we would like.)

Suppose, for the sake of argument, that the fiscal rule involves some target for the deficit in five years time. A negative demand shock hits, and the central bank judges there is more than a 50% chance that interest rates could end up at the ZLB. At that point, the deficit target would no longer apply. Instead the central bank and fiscal council would be obliged to cooperate in formulating a fiscal stimulus package that would enable interest rates to rise just above this lower bound. Both institutions would then recommend this package to the government. The central bank and fiscal council would continue to cooperate in this way (suggesting modifications to the package as new data became available) until the central bank expected interest rates to rise. Call this the ZLB procedure.

Why do we propose cooperation between the central bank and fiscal council? The central bank should be involved for four reasons. First, they probably have more resources working on short term forecasting. Second, they could help design a package that was effective at raising demand, rather than one that pandered to political preferences. (Fiscal councils on their own would be wise to avoid making judgements about how deficit targets should be achieved.) Third, a government would find it difficult to go against these two institutions acting together. Forth, the central bank would also probably want to implement some form of unconventional monetary policy, and the size of any fiscal stimulus would need to allow for that.

Now you could say that a central bank would be reluctant to make this call, for the reasons I gave at the beginning of this post. However, the central bank bears responsibility here. If interest rates did go to zero, they will clearly have made a mistake, and can be held accountable for that. And once interest rates had hit the lower bound, the ZLB procedure would operate anyway.

Why should the fiscal council be involved? Because part of the package would be an assessment by the fiscal council of what should happen to deficits and debt once interest rates did rise above the lower bound. The previous deficit target would have to be revised. A government that accepted the stimulus package would be asked to sign up to meeting any new deficit target after the recession was over.

This ZLB procedure should be part of any ‘fiscal rule’. This may seem odd, because it is really an account of how the normal rule should be suspended and what should replace it in ‘exceptional’ circumstances. Why not instead simply say that the normal rule should no longer apply when interest rates hit the ZLB and leave it at that? The reasons are those given at the start of this post. Making up policy in a crisis is fraught with dangers, and macroeconomic rationality can easily give way to vested interests and biases, as we discovered in 2010. The whole point of fiscal rules and fiscal councils is to overcome those interests and biases, and they apply at least as much in a crisis as in normal times.  


  1. There is a certain irony on the day in which Europe goes to the polls with disenchantment with the political process at its highest because citizens feel disenfranchised (real decisions being taking by the ECB or the Troika rather than by, say, those elected to represent them in, say, Greece or Spain) in proposing that fiscal policy should be handed over to be monitored by 'fiscal councils' (equivalent to the MPC).

    I agree with the economic prescription, but in democratic terms it is a complete non-starter.

    I know Plato argued that we should be governed by 'philosopher kings'. experts trained in governing the rest of us. Philosopher Kings themselves rather favour this approach.

    But. Representative democracy has more to be said in its favour than you allow for here. Yes elected politiians may make choices on fiscal policy that Wren-Lewis (or I) disapprove of. Let us have experts to advise them, but ultimately the choices should be theirs.

    That they sometimes get it wrong is a price well worth paying.

    1. I do not understand this comment. The proposal involves recommendations to governments - ultimately the choice is theirs about whether they follow these recommendations.

    2. The entire point, I would have thought, is that once a government has opted into the rules, with oversight from the fiscal council (which of course they are free to do or not do,) they would no longer have complete freedom of action from that point on. A rule that you are free to ignore is not a rule.

      Now it may be that these rules of your devising will be entirely benign, and any fiscal council will be made up solely of Philosopher Kings of great wisdom.

      But I would have thought that in the EU we have just had a rather sharp demonstration of the fact that fiscal rules are not always entirely benign, that those in charge of their implementation are not always all that wise, and that the democratic deficit created when we are governed by Philosopher Kings rather than elected politicians is a bit of a problem.

      A common mistake made about politicians is that they like responsibility. They don't. If they can palm off difficult decisions on to experts, and then blame them when things go belly up, they will.

      Better to leave the responsibility with the elected politicians, acting under advisement from the Philosopher Kings of their preference.

    3. I’ve heard the “Spinning Hugo” argument dozens of times. The SpinningHugos never see the contradiction in their argument which is thus: about 95% of them are perfectly happy to have technicians or “philosopher kings” determine stimulus when those “kings” come in the form of central bank monetary policy committees. I.e. CB MPCs determine interest rates and the amount of QE, for example. But as soon as you suggest that technicians should determine fiscal stimulus, they kick up a fuss. It’s bizarre.

    4. Democracy is often misunderstood as requiring voting as basis for decisions.
      Rather than around 'counting votes' (with everyone being allowed to vote and with all votes counting equally) democracy can also (and complementarily) be designed around 'weighing arguments' (with everyone being allowed to enter arguments into decision-making, with the weighing process being transparant so people can check whether their arguments are being taken into account and with -as now- decision makers being held to account if they fail in doing so).
      Democracies place certain types of decisions (e.g. court decisions, 'implementation' of policies) at a distance from representative (voting based) decisions and different democracies place the same type of decisions at various distances from 'the people' in that sense.
      There is no reason why we couldn't experiment with placing certain types of economic policy decisions farther from 'most votes count' processes and compensating that with more openness for input and more transparency of decision making.

    5. " The SpinningHugos never see the contradiction in their argument which is thus: about 95% of them are perfectly happy to have technicians or “philosopher kings” determine stimulus when those “kings” come in the form of central bank monetary policy committees"

      Errr, Ralph, if you re-read what I said you'll see that I specifically said that having decisions taken by the ECB rather than by elected politicians created a democratic deficit.

      So this SpinningHugo, unlike presumably the many others you have met, agrees that having monetary policy decisions taken by Philosopher Kings is also far from ideal.

      No doubt the MPC has proven something of a success in economic terms, but in democracy terms it is horrible.

    6. I agree with SpinningHugo. I would much rather live in a democratic society where mistakes and poor decisions are made than in a society ruled by unelected technocrats. Philospher Kings, Benevolent Despots, Power corrupts ... all terms developed out of historic experiences.

      I would have thought that the US Constitution already has the necessary "rules" with the Congress the "council"? Indeed, the Ayatollah's religious council in Iran and the Politburo in USSR & China are also "councils" developing the "rules" by which govt have to operate. Do SWL understand that such a economic "fiscal council" could become like these religious/political councils, who has the power and authority to decide what is good or not for society? What is stopping the elite to ideological "capture" this council?

    7. Oupoot,

      How far do you take that process? E.g. do we have politicians design nuclear power stations and tell heart surgeons how to do heart surgery? We could vote relevant politicians out of office when the power stations blow up (ten years later) or when too many heart patients die.

      Strikes me that suitably qualified technicians should always take technical decisions with politicians having the ultimate power to over-rule them, as Simon suggests above.

    8. The US does indeed have a fiscal council - it is the CBO. So what we are suggesting is that when interest rates hit the ZLB, the CBO cooperates with the Fed to put together a stimulus package which it then recommends to Congress and the White House. Is that really such a threat to democracy?

    9. "E.g. do we have politicians design nuclear power stations and tell heart surgeons how to do heart surgery?"

      And there we have the mindset of a (certain kind of) Philosopher King revealed.

      Economics as 'science' with 'right' and 'wrong' answers that are capable of proof (save that, as yet, we haven't reached any kind of agreement as to what those answers are.) Decisions too difficult and dangerous to be left in the hands of the electorate.

      No I'd much rather have these decisions taken by Gordon Brown (who I thought completely useless) or Alistair Darling (who did as well as any man could) openly through a democratic process.

      That doesn't mean that they should ignore the advice of Wren-Lewis. Portes et al. Indeed, despite the wailing and gnashing of teeth about how the UK's is the worst government ever, it hasn't been even by a government with different instincts from theirs.However it does mean that we should not institutionalise their power.

      Indeed, the obvious folly of what W-L and Portes are suggesting does not stop there.

      What do you think is more likely to be the form of any enacted fiscal rules (in either the US or UK)?

      A set of rules for a stimulus where interest rates hit the ZLB?

      Or a set of rules to ensure 'fiscal discipline' and a 'balanced budget'? (cf the eurozone).

      Who do you think a Tory (or Republican) government would have sat on its fiscal council?


      No, for good or ill, it is better that responsibility rests with Osborne.

  2. Alternatively, one could look at the governmental budget constraint and realise that there is no mechanism to force future primary surpluses, regardless of the level of the government debt. As a result, economists could then point out that there is no effective constraint on fiscal policy other than inflation. We could then progress back to traching Functional Finance.

    1. *teaching functional finance.

  3. I think the basic problem with this post is the assumption that an independent institution would make wiser economic decisions than elected governments. Here in Sweden the (Right-wing) government stayed clear of austerity during the first years of the crisis and has more recently begun a weak stimulus effort (because economic growth is slowing). Conversely, the central bank raised the interest rate despite a falling core inflation rate and unemployment above trend; we are now on the verge of deflation.

    As Antonio Fatas repeatedly points out, the ECB has implemented no QE even though the case for it is stronger in the eurozone today than it ever was in the US and it has often acted much too slowly. To take an example, the Open Market Transactions were not announced until 2012.

    The problem here is psychological. A politician who refuses to change an economic policy that conforms to his personal bias even though it causes significant economic harm will get voted out of office. A central banker with a bias (against keeping interest at 0% for an extended period, for example) will keep his job and his beliefs. In other words democracy is good because it forces policymakers to be flexible and to adjust their views in response to the evidence.

    1. Just to repeat, the proposal is for the fiscal council and central bank to jointly advise on a stimulus package. The decisions remain with governments.

    2. The Right-wing government (the Alliance) did indeed practice austerity and did it inoptimally by cutting infrastructure investments in 2009 and 2010 by half. Short-term it looked good and the Minister of Finance Anders Borg bragged about their low budget deficit. The Prime minister, Fredrik Reinfelt, practiced all the rhetorics from Merkel and fell immediately into the Hosehold Fallacy trap and The Fallacy of Compostion. He frequently uses the moralizing arguments of Merkel and Schauble of the Economy as a Morality Play.

      The Right-wing will now be ousted from power in September since the electorate dislikes mass-unemployment, low economic growth, low levels of investments, high output gaps etc. (Voting Out the Bums-phenomenon).

      But you are right about the central bank, the Riksbank, and its strange behavior during the crisis. In spite of having its own currency, very low government debt to GDP-ratio, etc. the central bank chose to constantly keep interest rates higher than the ECB and hiking despite low core inflation and high unemployment.

      The Riksbank has of course to a high degree contributed to the bad unemployment figures and the low investment level and is one of the main culprits for the coming change in poltical leadership.

      But the Swedish Right-wing government has dug its own grave by adopting austerity in a crisis environment that called for fiscal stimulus.

    3. First things first, the woman who will probably become Swedens next Finance Minister (Magdalena Andersson) has criticized the governments recent economic stimulus for creating "unsound public finances". When it comes to the deficit, the Swedish Social Democrats are using much the same arguments as the Republicans in the United States. Yes the right-wing will almost certainly get voted out but that will not lead to better economic policymaking, rather the opposite.

      Second, I have trouble finding the austerity you speak of in the statistics. The budget has been in a significant deficit for most of the years since the crisis began. Perhaps more importantly, there has been much less need for stimulus in Sweden than in many other countries. Sweden did not suffer a financial crisis or a housing bust and it is not bound by the constraints of the Euro. Indeed, the way the economy strongly rebounded after the crisis strongly indicates this.

      The recent slowdown in the economy (heavily abetted by the strangeness of central bank policy) has led to some government stimulus which is precisely the right policy in this kind of situation.

    4. Professor Wren-Lewis is right though, I did go a bit overboard in my first comment.

      There is one point I want to make though: While a fiscal council might not be a bad thing, I very much doubt that it will be sufficient for preventing the kind of destructive policies that we have seen over the last few years.

  4. I don't understand the following: "However, the central bank bears responsibility here. If interest rates did go to zero, they will clearly have made a mistake, and can be held accountable for that." What interest rates are we talking about? Short term? Dont they control short term interest rates? ECB has keept them above zero for a long time. Long term Gov bonds? They never hit 0 (see US, JP). So a central bank can easily avoid being held accountable. Could you clarify your thinking? Thankyou

    1. I'm talking about short rates. Defining what zero means here is not a major problem. Implementing our proposal for the Euro area does introduce important additional coordination issues.

    2. PGG has a point here. The ECB, for example, has never admitted they are at the zero lower bound; they insist that they still have room to ease conventional policy and in fact until recently regarded their execution as 'impeccable'. A CB that wants to define goals and claim credit for success should also define the criteria for failure more clearly. The BoE's letters to the government explaining their inflation misses are a good step.
      Still, no matter how much planning is done, hitting the ZLB is still going to occur with some frequency, just as inflation overshoots and overheating will -- the CB usually is complicit but cannot be held solely accountable, either. Monetary policy and fiscal policy are not all powerful. But a non-partisan fiscal authority, with more power than, say, the CBO, that can react more quickly and at least point the compass in the right direction would be a big step forward. As Prof. WL suggests there are powerful forces that reflexively push the wrong way (rather generously, he omits speculating on their motives).

  5. If the exercise is to propose a single institutional innovation that could be expected to lead to better macroeconomic outcomes over the long run, how about the "simpler" step of eliminating central bank independence?

  6. I shall repeat my email response from the BBC 02 May 2012:

    "In terms of combatting the liquidity trap, we have also reported the Bank of England’s estimate of how quantitative easing had affected economic output or GDP on October 6 2011. The wider point, of course, is that the Bank of England and Fed in the US would not agree we are in such a trap, even if there are some economists who do."

    Each BBC response ends with a nod to demagoguery. as did this one: "Nevertheless please be assured that your comments were registered on our audience log. This is an internal report of audience feedback that is made available to editors and their teams, channel executives and senior management. The audience logs are seen as important documents that can help shape decisions about future programming and content."

    Cracking this VSP nonsense will be very hard, and anything that helps lay the ground should be welcomed. As Krugman said in 'The Simple Analytics of Invisible Bond Vigilantes':

    "As far as I know, none of the people issuing dire warnings have actually tried to write down a model of what an attack would look like. And there is, I suspect, a reason: it’s quite hard to produce a model in which bond vigilantes have major negative effects on a country that retains a floating exchange rate. In a simple Mundell-Fleming model (M-F is basically IS-LM applied to the open economy), an attack by bond vigilantes has very different effects on a country with a fixed exchange rate (or a shared currency) versus a country with a floating exchange rate. In the latter case, in fact, loss of confidence is expansionary."

    The BBC and elsewhere need to stop the feeling that they can whitewash at will.

  7. UK debt as a percentage of GDP has more than doubled since 2008, and the conclusion is the next time the solution to a "demand" crisis would be to - what - double or triple it again? Pray tell, is there a limit to the amount of debt as a percentage of GDP which you think the UK should bear?

    1. Government debt does rise rapidly during a crisis (like a major recession or war), and then gradually declines thereafter. To help ensure the latter happens, any package would include an assessment by the fiscal council of how the subsequent decline could be achieved, as I noted in the post.

      As to UK government debt, it is really important to understand that this increase is just the consequence of a large increase in savings by the private sector caused by the financial crisis. If people want to save more, basic Keynesian analysis tells you this can happen in two ways - the government borrows more, or output and incomes fall. We got too little of the former, so had too much of the latter.

      If this government debt was really being 'forced' on an unwilling public, interest rates on debt would rise. The opposite happened.

    2. Anon,

      Like Kenneth Rogoff and Carmen Reinhart, you seem to be horrified by debt, as evidenced by the last word of your comment, i.e. “bear”. The reality is that as long as the interest paid on debt is equal to or less than inflation, there is no “bearing” or “burden” to be carried. Indeed, where the inflation adjusted rate of interest is negative, debtors profit at the expense of their creditors.

      As to what to do if creditors demand a higher rate, that’s easy: print money, pay them off and then tell them to get lost. As for any inflationary effect of that, that can be dealt with by raising taxes and “unprinting” the money collected.

      Government debt at low interest rates is virtually the same as money (base money to be exact). So the answer to your question as to how much debt/money should there be is: whatever amount induces the private sector to spend at a rate that brings full employment.

      The above, by the way, is all standard Modern Monetary Theory thinking. You might be interested in reading up MMT.

  8. "the importance of ideology in the discipline is such that economists will never be united on this issue"

    Hear, hear!
    '... will hardly ever be united on hardly any politically relevant issue', I would say.

    What about acknowledging the role of ideology in both economic practice and economic theory as economists and consequently endogenising ideology in economic theory?

  9. I like the idea of fiscal rules, and deficit targeting. We should run a permanent deficit at all times, with only its size adjusting to circumstances.

    I would like to suggest that the true consolidated deficit that needs to be targeted is the difference between the treasury deficit and the central bank's net income. If the government, in the US for example, has a treasury deficit of $1 trillion while the central bank has net income of $100 billion, then the consolidated deficit is $900 billion. The central bank is partially offsetting the $1 trillion injection by extracting $100 billion.

    It is possible that the interest rate effects of QE policies have been completely offset by the fact that the way QE lowers rates is by buying financial assets that generate a cash flow for the central bank, and thus drain even greater amounts of money from the private sector economy over the short and medium term than the purchases themselves inject in the immediate term. That the Fed has been generating large "profits" from its SOMA portfolio is a fact to be deplored, not celebrated. A central bank does not need profits in order to operate.

    Yet, the interest rate effects have to be taken into account, so I think what is needed is a comprehensive policy rule which incorporates both interest rates and the consolidated deficit in a single policy formula.

    But the interest rate effects shouldn't be overestimated. What I am worried about is that if there is no zero bound next time around, economists and policy-makers will go back to thinking that central bank policy adjustments can handle the job. I think they can't. The optimistic assessments of monetary policy are based on models of the central bank's relationship with the economy which are simply flawed and posit effective policy channels that just don't exist.

    The problem, I believe, isn't the zero bound. It's that in the end, central bank policy consists almost entirely in interest rate adjustments, and that interest rate adjustment just cannot accomplish what we need in a deep recession. The central bank can swap liquidity for other assets with financial sector agents and lower rates, but then it just has to hope that those actions are followed by an expansion of borrowing, investment and consumption. The money multiplier model on which these expectations are based is dead, zero bound or no zero bound.

    The treasury, unlike the central bank, can invest and not just try to stimulate investment; it can consume and not just try to stimulate consumption; it can strategically innovate in the economy, and not just try to stimulate consumption-oriented innovation with better interest rates. We need a more activist economic role for government in carrying out these activities.

    Finally, we should develop new methods of direct monetary financing of deficits, in accordance with the policy rule, in ways that don't necessarily involve the issuing of more bonds or gilts. A macroeconomically justified expansion of the deficit shouldn't be used as an opportunity for the wealthy to rent-farm the ordinary operations of government. The treasury's general fund should be seen as just an operational extension of the consolidated government balance sheet, and the central bank should be free to expand the fund with a direct, credit-free issue of central bank liabilities, if macroeconomic circumstances warrant.

    1. "The treasury, unlike the central bank, can invest and not just try to stimulate investment; it can consume and not just try to stimulate consumption; it can strategically innovate in the economy, and not just try to stimulate consumption-oriented innovation with better interest rates. We need a more activist economic role for government in carrying out these activities."

      Well said and, to me, the practical crux of the issue. Any 'remedy' that reinforces the misplaced faith in the Central Bank as a source of adequate demand management, will be inherently self-defeating.

      Rather than 'fiscal rules' and Central Bank noodling, let's try to overcome the political reluctance to spend adequately in the teeth of a crisis/downturn by advocating now for an increased use of automatic stabilizers on the downside (and, just as important, a scaling back of their use on the upside, where they almost always bite too early, perversely helping to terminate expansions).

  10. Why 'mistake'? Mistakes are fixed when perceived. The policies have not been a mistake but a political decision. Why should another layer of bureaucracy stop this kind of policy, err, mistake, from happening again given that the mistake was deliberately and repeatedly made?

  11. Or, we could change the way we define/measure the deficit. (The current method is rather silly anyway).

    Measure/define it as something like:

    Deficit = primary deficit + (current nominal interest rate - inflation target)xdebt.

    Better still:

    Deficit = primary deficit + (current nominal interest rate - NGDP growth rate target)xdebt.

    Those revised measures/definitions make more sense from the sustainability point of view, and would also have the desired cyclical property.

  12. I agree that rules could be established but think you are on the wrong track with regard to the main triggers.
    1. ZLB is one condition that the economy needs stimulus, but we need to get away from the fallacy that men and women can predict the future and instead determine what the policy response should be to certain events before hand. For example, the bond vigilantes kept warning of inflation, but that's just another event. The rules should say that if inflation comes in x% over predictions for the past quarter, that interest rates rise by y%. In essence, we need a schedule of 'surprise' to match the remedy. Perpetually predicting high inflation is a neuroses and phobia that won't change because it is a personality trait more than an equation. This requires merging predictions from several models and economist, but the outcome is to weigh the predictions by accuracy relative to the past 6 quarters or something like that. The people always predicting high inflation will see their influence on the suggested remedy reduce. This is an attempt to filter the models toward those best predicting the next set of unemployment, GDP, and interest rates and using the preset adjustments they recommended by the level of surprise from the next actual economic reports. For example, let's say that you predicted inflation,unemployment, and GDP at (x,y,z) and your prediction was spot on. Then the scheduled remedy you may have prescribed could have been (--,-,0,+,++) for interest rates and (+++,++,+,0,0) for deficit spending, and so forth. This way you show IN ADVANCE the range of policy changes you would make for a range of actual measures. In this example, since your last prediction was spot on, interest rates would not change, but you would recommend raising deficit spending.
    2. Deficit spending has two flavors, a) bring forward in time government spending that is going to happen in the future, or b) spend government money on projects that are new. In the first case, one can see the government using the logic "since interest rates are so low, it would be prudent to rebuild the bridge scheduled for next year in the current year". For the second case, you get into great political debate about how spending $700B is a waste of taxpayer's money and will turn into permanent spending; but will never discuss the recurring $1T of lost GDP from depressed demand and the waste of time waiting for demand to return. We have lost more than $5T in GDP in the USA(SO FAR) just from reduced demand and all we hear about is Benghazi.
    3. Until you convince politicians that lost opportunity is a cost, the ZLB problem will remain a political battle

  13. If we see monetary authorities as justified, it is somewhat peculiar to oppose a similar undertaking once we turn to a different means of macroeconomic management. Even going beyond the proposition of Wren-Lewis, it is worth asking in what respect does a fiscal council endowed with the authority to enforce its rules differ so much from central banks. Either way, a government provides non-elected officials with legal powers to manage the economy.

    If anyone has issues with the legitimacy of such a procedure, they will have to make it clear how these two cases are different, unless they are ready to provide elected officials with the task of managing monetary policy as well. Does anyone want to see what the grid lock of Octobre 2013 would look like with a monetary version of Ted Cruz who insists that the interest is too low? Personally, I don't.

    Now, on another note, suppose we do see professor Wren-Lewis' solution as morally acceptable. Why should we believe that such a council will not provide ill-founded tips? Why would such a council have come up with a stimulus plan? Is it not possible that a group of expert would agree on doing exactly the wrong thing?

    I am not too familiar with the literature on the question, but we know fiscal policy to be a controversial topics in political debates -- and it's hard to imagine a situation where you'd bring right-leaning economists to advise enlarging the deficit. So, how would we solve such a problem?


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time.