Winner of the New Statesman SPERI Prize in Political Economy 2016

Friday 14 June 2013

Why Bernanke was right to speak out on fiscal policy

This is a comment on Cardiff Garcia’s post on fiscalists and market monetarists, and also some related criticism of Bernanke’s recent remarks on fiscal policy, criticism which I think is totally wrong. I want to argue that a ‘monetarist’ position which is indifferent to what fiscal policy is doing in current circumstances is untenable. As a result, central bankers have to speak out on the dangers of austerity. [1]

There are two lines that monetarists might take. The first is that unconventional monetary policy, Quantitative Easing (QE), is a perfect substitute for conventional monetary policy. The second is that an appropriate monetary policy regime can, through expectations, undo the restriction imposed by the zero lower bound (ZLB). Let me take each in turn.

The first argument is wrong mainly because of uncertainty. Macroeconomists know little enough, but we do know something about how conventional monetary and fiscal policy works, and we have a lot of data that can help us. We know so much less about unconventional monetary policy. What kind of model we should use is unclear, and we have very little data.

The second argument would be right if we could fix inflation expectations in exactly the same way as we could, absent the ZLB, fix nominal interest rates. Would a nominal GDP target do that? Of course not. I think it would help, particularly compared to an inflation target regime, because the latter actually inhibits inflation expectations rising above that target. That is why I have recently argued that a path for nominal GDP should be adopted by central banks as an intermediate target. Would adopting such a target raise inflation expectations and speed a recovery? - I think it would. Would it raise inflation expectations by enough to negate the need for any fiscal stimulus (or, more realistically, to counteract the impact of fiscal tightening)? There is no logical reason why it should. But let us just suppose it did. Does that mean we can ignore fiscal policy?

Absolutely not. What we are getting in this case is a recovery achieved by raising expected inflation above (in the UK, US and Eurozone) 2%. That is costly, because it means actual inflation must be allowed to go above 2%. The more we deflate demand through fiscal austerity, the higher inflation has to go (or the longer it has to be above 2%). So monetarists who believe in the expectations channel cannot be indifferent to fiscal policy, unless they also believe it has no effect, or that inflation above 2% is costless. (I make a similar point a little more carefully here.) If, as Paul Krugman says, fiscal policy makers are doing the wrong thing, that is a cost worth paying, but it is a cost nonetheless.

This is why it is really important that central banks, like the Fed, make it publicly clear the difficulties that fiscal tightening is causing them in meeting their mandate. Either this is because they are, quite rightly, uncertain about the impact of QE, or they are aware that the more fiscal tightening there is, the more inflation will have to go above 2% to counteract its impact.

The idea that to speak this truth is wrong because it might frighten the horses is silly. I have used the following analogy before. No one wants to hear a pilot tell passengers that they are no longer in control of the plane. However a better analogy in this case would be the pilot not telling the co-pilot, which would be highly dangerous. The horses that matter here are those in charge of fiscal policy, and they need frightening.

[1] Sorry Nick. I have a lot of sympathy for the point that we should not routinely exaggerate with language. The (I think just British) phrase I hate is ‘black hole’ when used to describe a worsening in the government’s accounts. The use of austerity to describe what is happening in parts of Europe and the UK right now is less obviously loaded or misleading, but I’m open to persuasion.


  1. Does that mean we can ignore fiscal policy?

    I completely agree with absolutely not. The change of one thing will always bring a change in another. These changes are made to keep all areas balanced so there will be no major change in one area.

  2. Macro folks make a lot of mainly methodological mistakes in this discussion.
    Just to mention the probably most important (no ambition to write a book).

    1. No proper definitions. Austerity in particular is simply not properly defined. Looking at most possible definitions in most the present situation it isnot austerity. Governments are spending more in percentages of GDP than pre-crisis (and often in real terms as well).

    2. A non-methodological one. Is reversing structurally unsustainable spending austerity? I think you can put a lot of ??? there. Anyway it is doubtful if society behaves in the same way on that as it would on a stimulus in more or less normal circumstances. Simply eg will people go consuming (in the same way) if they know there might be growth but anyway the financial muscle built on steroids will be gone anyway?

    3. Another one missed by many (not by Simon btw). In a lot of countries there is simply no borrowing capacity left to finance any stimulus. From a practical pov there it is a totally useless discussion. If we had room what would it do? Well if my sister had a d... she would be my brother.
    A related question: will it be effective if half or more of the world doesnot do it (but has to make cuts) and is it worth the trade off. Basically what is your multiplier (will it not be (much) lower than usual?

    4. Applying stimulus stuff basically designed for smaller dips WITHOUT asking any further questions like most do. Size (of the crisis) matters or better might matter. The stuff there is, is probably the best there is at first sight. However a continuous check if it does the job and how acurately is simply appropriate.

    5. Greece and co donot show that stimulus works. It Basically only shows that austerity in the form used there doesnot work. And doesnot work as far as generating growth is concerned.
    It might be wise to check if that is the main target of the decisionmakers btw (probably not). And do not be so naive to only look at their words, look at the whole picture. Eg a politician will never say we goi for a healthy bankingsector and take your financial suffering simply as collateral damage. They will always give a spin to that.

    6. Not looking at example where in a roughly similar situation stimulus is used. Best cases are probably US, China and Japan.
    US it looks to be sort of working. Although you can be pretty sure the FED is a bit disappointed by the results (setting up against the potential risks caused). But the US is always special because of its reserve currency status. Hard to compare with any other thing. They simply keep money get coming in from abroad because of that. Not absolute but the cliff is much further than that of say the UK or Japan. People (the Johnny Foreigner kind) donot need those currencies and will simply drop them when things get too bad.
    China now has a borrowing problem. So stimulus is clearly not without dangers.
    Japan has as well and policies to tackle the negative consequences of fall out seem not to be working. The appropriate authorities simply didnot oversee the complete picture and missed a lot (probably assumed away in the model). And as a side remark: a future BoE CEO more or less suggesting the same thing for the EZ hardly gives confidence in the technological capability of the people dealing with the matter (compared to the complexity of the medicine).
    Organisational mistake: don't let an organisation do things it is technologically not able to do or in this caes intellectually able to oversee unless you are completely desperate.

  3. Part 2 (I am aleways boringly long).

    7. Increasing interest rates danger. You should check that over the period that debt by stimulus will possibly be higher than acceptable.
    Basing that on looking at long term rates is simply complete rubbish. Long term pricing developments are crap as long term prediction instruments.
    Probably the only thing that might work (and only for a part) is working with probabilities.

    8. The issue is not full stimulus or cutting everything away. Margins are very small. Basically if you look at full austerity (in the definition used now in practice, well there is no proper definition but how it is now applied) and not austerity (in the present situation) are very small. The extra stimulus or less austerity that it will generate will never do the job not even half (probably much less). The theoretical sufficient stimulus amounts are simply not on the table.
    Basically a non discussion for most of the countries hit the hardest.

    9. Japan is most interesting imho. As it is the most aggressive. And alot probably can be seen there. However Japan was imho F'ed anyway. Either its debt went up till it hit a wall or it does something wild.
    Will be good to see if it fails on other stuff than the interest rate or not.
    Imho it will always go down either now as markets will demand higher interest (and with a 230% debt you can not grow against that) or longer term as realistic growth and deficit estimates simply indicate it will not do the job and lower debt level (and that wall looks close as well). It is either or and it will hit a wall in both situations, only questions being which one and when.
    In this respect the present set up isnot that bad FOR JAPAN it gives the opportunity to go further with inflating away the debt (likely the bust 'solution'). In other words a failure doesnot have to implicate that it doesnot work (it will only have optics against it from thereon (not unimportant).

    So basically:
    You folks got a lot of your basic facts not properly checked.
    Assumptions made on basis thereof and choice of formulae suck eg as a consequence thereof.
    Things are not properly defined.
    The formal logic of sucks as well with many.
    Higher degree of uncertainty not taken into account.
    Not put it in a proper decisionmodel per country (as the situation is different).
    Practicality of the discussion is only marginal. Be glad it is Macro and not management (as priority setting is way beyond sucking).

  4. Basically on this issue my personal view is that if possible first of all the bottom should not fall out.
    Which is/was a possibility in say the UK/US/Germay and Co. But not in the porcine brotherhood for instance.
    Japan being a weird case they are basically certain ti hit the wall but had the possibility to prevent it. Imho the right decision to prevent ther bottom falin out.

    From there it simply doesnot look like growth will return to normal percentages in the West and Co. So for the not bankrupt it probably has to be a mix from structural things and stimulus. Situation being different from country to country.

    US needs to get its healthcare costs under control (Obiecare simply sucks in that perspective, it is even worse than the pre-crisis situation).
    Europe needs to get answers on aging, healthcare, immigrationdisaster iso solution and a few others. Effectively they are much too late. And would take a huge hit anyway. problem being that it now comes out in a very unconvenient time and within a shorter period of time.

    In general donot create extra problems. Like the Dutch did with their RRE sector. They kicked the bottom out of it themselves and completely unnecessary the way it happened. As Simon's post also indicates one of the worst crisis-managed not bust countries, if not the worst.

    Point that economist are missing is:
    a) structural stuff is simply necessary (the West is losing competitiveness to the rest fast, something has to happen). But probably largely other stuff than is happening.
    b) politically a crisis is probably necessary to get that through to the population. And even now almost 6 years after for alot of simply necessary stuff there is no political platform. Mention healthcare cuts anywhere and everybody gets red spots and runs to the GP.

    Anyway these reforms like higher pensionage and paying more for that stuff if you want to keep the same level of services or get cuts somewhere else, better education, more real economy (but in the advanced sector), more R&D are likely having a higher priority on the political aganda than stimulus. So would cutting costs of the bankingsector for the total economy (not see that one happening).

    Would be great if the economy could be back to a self sustaining growth in that process. Still far from that with a multiplier of say 0.5, and ofsetting against a sustainable growth and inflation level combined, all Western countries have negative growth. Which also shows some structural stuff needs attention.
    fact we are having a normal cycle dip at the moment is also hardly helpful.

  5. Amen.

    Monetary Policy no matter how good, cannot sufficiently compensate for bad or inadequate fiscal and regulatory policy.

    -jonny bakho

  6. "What we are getting in this case is a recovery achieved by raising expected inflation above (in the UK, US and Eurozone) 2%. That is costly, because it means actual inflation must be allowed to go above 2%. The more we deflate demand through fiscal austerity, the higher inflation has to go (or the longer it has to be above 2%). So monetarists who believe in the expectations channel cannot be indifferent to fiscal policy, unless they also believe it has no effect, or that inflation above 2% is costless. (I make a similar point a little more carefully here.)"

    I don't think you'll convince any market monetarists here. You've once again assumed a negative welfare cost from short-run inflation variability even if the level path of NGDP level is kept stable, which they'll take issue with.

    Mark Sadowski pointed this out previously:

  7. Simon: take a real example: the Canadian (Conservative) government increased (or rather preponed) investment spending during the recession. Was that "austerity"? Of course not. Was it "profligacy"? No, it wasn't that either. It certainly increased the current period deficit, but was probably good for government finances in the long run, because they took advantage of a period of abnormally low real interest rates and cheap resources to build stuff they would probably have built later anyway. It seemed like a very sensible decision to me (despite the inevitable problems with lags), even if judged on purely micro grounds. You could maybe say they were practicing "being cheap", which sounds a bit like a household practicing austerity. But they borrowed money to do that spending, which sounds a bit like profligacy!

    Words like "austerity" or "profligacy" are just totally useless when we want to discuss the pros and cons of fiscal policies like that. It's not that they exaggerate. They are just orthoganol to the real issues. They garner both support and opposition for all the wrong reasons. Calling cutting deficits for the wrong reasons "austerity" makes it sound good.

    (BTW, I like to use the words "black hole" to refer to the theoretical possibility, in a neo-Wicksellian model, of a deflationary death-spiral. Because I find it weird that we don't actually seem to observe black holes, even though the model predicts their possibility. Though maybe the Eurozone will be the first??)

    Fiscal policy almost always matters. ("Almost", because I once observed a case of Ricardian Equivalence, when my local municipality asked me whether I wanted to pay for paving the road in one lump sum or in 10 annual installments with the same PV, and it made no difference to anything either way.) That's true whether or not monetary policy can hit an AD target regardless of fiscal policy. It is precisely because fiscal policy almost always matters that we want monetary policy to be assigned to controlling AD, so that we don't need a constrained optimal fiscal policy because monetary policy is sitting around idle. If fiscal policy didn't matter (except for controlling AD) then we could toss a coin to decide whether to use fiscal or monetary policy to control AD. If you have two instruments and only one job it doesn't matter which you use. (I need to find a clearer way of expressing what I'm saying here, but it goes back to the old targets and instruments debate.)

    1. Nick: I think we are largely in agreement here. For example, I looked at my recent paper on Labour’s fiscal policy 1997-2010, which was all about periods of fiscal tightening and loosening, and I do not use the word austerity once. However I think it is useful to have a word that distinguishes the context in which fiscal tightening or loosening occurs, so how about the following definition: fiscal austerity=severe fiscal tightening undertaken in a recession. If you wanted to be really specific you could add: at the ZLB.

      I also think you and I largely agree on fiscal policy. My remarks there were mainly a response to some comments I keep getting along the lines of ‘why do you keep writing about fiscal policy? The right monetary policy would solve everything’. I cannot justify that view with standard macro. Now of course this aversion to using fiscal policy may have other roots. As David Beckworth wrote recently: “Okay, but not all fiscal policy is equal. Fiscal policy geared toward large government spending programs is likely to be rife with corruption, inefficient government planning, future distortionary taxes, and a ratcheting up of government intervention in the economy. So I will pass on this type of fiscal policy.” If that is why some people are not concerned about current austerity, then they should follow David and say so upfront, so we move the discussion on to the microeconomics of what current austerity involves, rather than wasting time with macro.

    2. Simon: yep. I think we are on the same page.

      I'm trying to formalise this.

      Suppose social welfare U depends on monetary, fiscal, and other stuff: U=U(M,F,X)

      Define the optimal M and F policies, conditional on X, that maximise U, so that U*(X)=U(M*(X),F*(X),X)

      Question: Under what functional forms of U(.) would it be possible to decentralise this problem so that there exists some optimal target for monetary policy N*, such that the monetary authority could be told to target N*, and the fiscal authority could be told to max U subject to N=N*?

      I think the answer is: if we can re-write U(.) in the form:

      U(N(M,F,X),F,X) where U*=U(N*,F*(X),X)

      In words: both M and F and X affect N (AD or NGDP or inflation or whatever), and we care about N and F and X, but we only care about M via its effect on N.


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. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.