Winner of the New Statesman SPERI Prize in Political Economy 2016

Monday 22 July 2013

Getting unhinged by ‘unhinged’: macroeconomic trade-offs and taboos

Central bankers, and even some of the best economists, sometimes talk about inflation expectations becoming ‘unhinged’. I do not like this term, and have been known to react quite badly to it. Some might say overreact. Let me say why.

But before doing so, I want to make three things clear, lest I be misunderstood. First, I think inflation expectations are really important. Second, I largely believe the great moderation story. As a result of setting inflation targets (explicitly or implicitly), and acting to achieve them, central banks did succeed in stabilising inflation expectations at low levels, and this has made the job of stabilising the economy as a whole rather easier. I may be wrong about this, but that is what I currently think. (I chose this as an example of the achievements of the microfoundations revolution in macro in my mild disagreement with Paul Krugman on this issue.) Third, I think it is more than likely that if inflation stays above/below target for some time, inflation expectations will adjust. 

But I would not call this expectations becoming unhinged. It is all about language. I would have no problem if another term was used. I used the term ‘adjust’ above, but we could also say ‘increase’, or ‘become less predictable’, or even ‘shift’. In fact any of the other words we normally use for macroeconomic variables. When discussing consumption, we do not talk about expectations of future income becoming unhinged. When discussing exchange rates and UIP, we do not obsess about expectations of future rates being unhinged. Whether intentional or not, the use of the term unhinged is designed to create an impression. The impression is of disastrous uncontrollability. If we talked about a person become unhinged, we indicate madness.

It is as if inflation expectations can be in one of two states: either low variance with mean reversion to the inflation target (or something close to it), or as highly volatile and could go anywhere. In this second imagined state, as expectations of inflation drive actual inflation, we could have ‘inflation bubbles’, which would become very costly for the central bank to prick. As we really do not want to go to that second state, we have to do everything we can to stay in the first state.

It is this view of the world that I find very difficult to believe - in fact I find it absurd. Why would inflation expectations become so unanchored from a central bank’s inflation target? They would do so if people thought the central bank had no intention of trying to achieve that target. So the only circumstances in which inflation expectations might become unhinged are when the central bank itself became unhinged. That could happen if the central bank was ordered to permanently monetise growing budget deficits, but that is not the world we are currently in.

When central bankers talk about unhinged expectations, they nearly always mention the 1970s and early 1980s. Do we really want to go through that again, they ask? Yet that was a period, in the US and UK, when it was very unclear what the central bank’s inflation target was, or indeed whether it had one. (This was not the case for Germany, as I note here.) So the lesson of that time is that inflation targets are important, but not that they should never be changed or missed.

I would draw a very different lesson from that period. It is important not to have taboos in macro. If there was a taboo at that time, it was that rising unemployment would mean a return to the 1930s. This prevented many seeing variations in unemployment as a means of stabilising inflation.

Could it be that we have a similar problem today, except roles have become reversed? With nominal rates at the zero lower bound, and doubts over unconventional monetary policy, we could use higher inflation (raised in a premeditated and controlled way) as a means of getting unemployment down. Of course that entails costs, and so we need to do the cost benefit analysis, and look at alternatives (like fiscal policy) that may be less costly. But if raising inflation is taboo we will not have that discussion. In this context, talk of expectations becoming unhinged reinforces that taboo.

In memory of Mel, who showed even as a teenager an appreciation for the absurd by helping me found the LUWS.


  1. The major trend the last decades as far as inflation goes is lowering it and stabilize it.
    Both have had a considerable influence on the yield investors demand for things like money/debt kind of things and subsequently for stuff that are mostly bought with a lot of leverage (like houses).
    Yield demand has dropped during that period because of this.
    As a consequence of lower yields asset prices have gone up and have gone up considerably.

    This makes structurally reversing this trend extremely difficult. Basically all pensionrights would (considerably) drop in value and so would most houses. A similar thing plays in the businesssector.

    It is unlikely that CBs and government can over a longer period fully control these higher 'natural' yield demands if inflation would go up and would become less stable. Such an unnatural situation longer time will lead to higher demand for loans and lower supply (say looking for yield what we see now).

    So it is imho a trade off the pros of higher inflation in a high debt enviroment especially and the cons (dropping asset prices including pensions). Priority setting and depending on a specific situation.

    To optimize the thing one could think of shorter term increase the expectations so the positive Macro effect will occur while keeping longer term expectations low (and not cause huge asset price drops).

    This also more or less indicates that if you have to get Japan going this is very hard to optimize in that way, the problem is way too large. You either need very high inflation or during a very long time and likely both. Difficult to see that this will not start to play via longer term expectations into higher yield demands.
    So imho unless there is an emergency scenario playing (like alternative being a default) this is something more suitable of smaller or medium size dips or as one of several measures.
    It worked imho fine, on a macrolevel at least, with UK RE. Higher inflation took so much hot air out of the market that the bubble may be not is gone, but not like France or Belgium still constitutes a huge danger that could tank the economy basically at any time. Giving some push to consumption without people worrying too much about things as their pensions. Well managed debubbling I would say.

    NB A lot of inflation now seems caused by rising VAT and other taxes. Difficult to see that this will give a big boost (except a one time effect ico VAT rises).

    NB I am not sure how the internal devaluation in the EZ will work out in this respect. Hard to see how increasing inflation at Macro/EZ level will not jeopardize internal revaluation. Making the inflation tool effectively ineffective.

  2. There is stability and then there is stability.

    I believe that the Japanese did not have the stagflation which the UK underwent due to OPEC I and II, yet land prices in major Japanese cities 1991-2006 fell by 68% (Shiller and Akerlof, 2009).

    The late Alec Cairncross wrote of the returning inflation in the 1980s as "old habits die hard", but didn't the UK have a new habit, the housing bubble explained through spurious ideas of globalisation?

    I have a copy of Martin Daunton's 'A Property-owning Democracy", with a preface written just before the 1983-8 bubble burst, in which the economic historian proffers the fear that the London home market in particular is in a bubble - not least as he was trying to buy a home there at the time.

    I heard on Radio Four this morning the City analyst of choice (I don't think they have names, do they?) talking of the Japanese fiscal stimulus, and the BBC's Simon Jack making sure he didn't say fiscal stimulus.

    So politically and economically, it's all eyes on Abe's Japan, as their right-wing government seems more frightened of the Chinese than of their creditor class, and what impact that will have on the UK debate.

  3. Prof Simon,

    You propose cutting unemployment while not bothering too much about the inflationary consequences. Assuming NAIRU has not risen since the crunch (and I don’t see why it should have) there is no reason to suppose returning unemployment to pre-crunch levels would raise inflation too much. And that sort of ties up with your conclusion on the 19th July that the size of the output gap has been underestimated.

    But, if NAIRU has in fact risen, then we need to think long and hard about exactly why it has risen. E.g. – just one possibility – could be that the proportion of the so called unemployed who are not seriously looking for work has risen.

  4. I think you're right that 'unhinged' is unhelpful, because it's a vivid word and implies an analogy with hinges when there isn't one. (Like Carney's use of the words 'escape velocity'). However, one explanation for what is going through the typical central banker's mind might be found in, say, a sticky price model with constant gain least squares learning. In such a model, provided there are only small shocks, and the central bank acts promptly, inflation expectations and inflation can be brought back to target quickly, and with relatively small output cost. But when there are large shocks, or a succession of medium sized shocks which the central bank misses, the system can easily explode without the central bank itself changing its targets.


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.