Christopher Joye rebukes John Quiggin for this post where he violates the territory of these guys. Quiggin criticises Central Bank Independence (in its strong from from the 1990s) and raises the possibility of higher inflation target to get more desired outcomes. Although from context I’m fairly sure Quiggin is speaking in context of the ECB and BoE 1 (where outcomes are far less optimal than here) Joye takes this as a call to interfere with RBA independence by changing the Australian target.
Incongruously he then says “we” (I presume elected government) should consider intervene by lowering the target instead. “Independence” seems to be a synonym for hawkishness rather than autonomy.
I come to this because I’m interested in the way inflation targets by reserve banks are discussed because it seems to me to have suffered from ossification of a make do policy into dogma.
A little history as I understand it. The technocratic monetarist dream of targeting the aggregate money supply and achieving 0% inflation proved impractical. Aggregate supply proved too endogenous and velocity too volatile. But monetarist understanding of inflation dynamics, the augmented Phillips curve and wage-price spirals etc. still seemed valid enough. Fortunately those spirals had ceased with the assistance of nasty recessions (and a lower oil price) but it was understood that a spiral could happen again were inflation to reach too high again.
This is the climate from which inflation targeting comes from. We haven’t the technocratic power to control the price level directly, so we’ll settle for making sure inflation doesn’t get so high it has its own inertia. We’ll target a rate, but there’s nothing special or technocratic about the rate chosen. It’s merely lower than the hypothetical and unknown level which would cause a breakout 2. The RBA has an upper target of 3%, which is higher than the 2% of the BoE, but Joye is also concerned that the actual growth in CPI has tended towards the upper bound and over in the past ten years.
Something I notice in the preceeding ten years is the lack of an inflation spiral. Despite regular adventures north of 3% there’s been no new inertia or self propulsion. Certainly nothing like what we saw in decades past.
So strict adherence to the 2-3% interval hasn’t been achieved, but the broader point of not returning to the 70s and 80s has been achieved so far. Whatever the hypothetical threshold that would see another spiral is, it seems to be higher than 3, or maybe 4%. There’s no real need for us to explore where that threshold is in the Australian experience given there’s no masses of clearly cyclically unemployed here, and from context I’m fairly certain Quiggin isn’t talking about Australia.
But the UK form instance is a different matter. There’s ample room to explore out from the arbitrary target rate (which is lower than the RBAs) before coming close to the hypothetical threshold, and it’s more pressing given the human misery on which adherence to the target would require.
Unfortunately the fact that the choice of target is arbitrary is neglected. On top of this is the notion that adherence to the target is necessary for Central Bank Credibility. This is tied to a fairly interesting implicit notion; an inflationary spiral won’t come as expectations get built into prices in contracts and decisions and bond yields across the entire economy as workers and employers etc. assess their own experience – it will instead come because commentators and pundits and Very Serious People are dismayed that the arbitrary target has been missed and thus their faith, and that of the public which partakes of their wisdom, will be destroyed and the currency debased.
Not surprisingly this notion, which emphasizes the crucial role of commentators and VSPs is very popular…amongst commentators and VSPs. But it seems rather separated from the understanding of inflationary dynamics that led to targeting in the first place. The technocratic veneer is mistaken for necessity.
2 And in the Australian example at least we add a lower bound rather than a cap, seemingly in acknowledgement that disinflation/deflation can also spiral and that mild positive inflation can help in adjusting real prices when nominal prices are sticky.
- edit – see JQ in comments[↩]
- fn1[↩][↩]
Great post, Richard, and I enjoyed your ‘maps’ post, too.
Olivier Blanchard partly laid out the case for a higher inflation target here: http://www.imf.org/external/pubs/ft/spn/2010/spn1003.pdf
His point is that higher average inflation implies higher average nominal interest rates, which gives greater room to cut rates in the event of a shock before the zero nominal bound is reached.
Richard, I often have trouble understanding your pieces on Troppo but not this one. You’ve written word echoes my thoughts perfectly.
It does need to be acknowledged that inflation is not inherently a bad thing (arguably the supply side sort is)
If you pay more, someone gets more. If you pay more, someone gets more. Rising prices = rising incomes.
As you surmised Richard, I wasn’t talking about Australia. Rather I was linking to a discussion about the US, which is also applicable to the UK and EU
Blanchard, being a macro rather than a micro guy, actually misses the biggest reason to have a target somewhere in the 3-4% range rather than the 1-2% range. It’s Bob Solow’s arguemnt that you advert to in the last sentence about relative wage adjustment.
In any economy some wages have to rise faster than others – which means that others sometime have to go down, at least in relative terms. Having significant positive inflation means you can do this without actually cutting anyone’s cash wages (if inflation is 4%, when you freeze someone’s nominal wage then their real wage will drop 4% a year. To do that if inflation was at 1% means you’d have to cut their actual pay packet by 3% – which will be resisted).
The point is that stable inflation of 3-4% means labour markets work better than at stable inflation of 1-2%. This is quite separate from Blanchard’s point about steering room to avoid deflation because inflation rates may not in fact be stable.
Spot on. It never ceases to amaze me how you don’t qualify as a VSP in public policy unless you’re dead keen to inflict pain on those who aren’t VSPs.
response here:
http://christopherjoye.blogspot.com/2011/05/what-happens-when-central-bank.html
Cheers CJ. I meant to aknowledge that there is a plausible case that 2008 was an incipient breakout (which the prevention of which I’m taking inflation targeting as jerry rigged methodology) and that the GFC was fortuitous in that regard. But firstly, it’s a case we can’t determine empirically (at least whilst alternative universes remain unobserved) so arguing about it would be alternate history – entertaining but of little value in determining policy. Secondly, assuming the counterfactual of no GFC, I think tightening fiscal and monetary policy (to get down to ~3% territory) in Australia in 2008 would have been doable without a massive human cost in the form of cyclical unemployment.
I don’t see that in the Northern Hemisphere experience. There’s already a great deal of cyclical unemployment, and greater leniency on the target front would still end up with an inflation rate well south of breakout territory, even if Australia (as you consider) was in breakout territory in 2008. Abandoning the 2% ceiling may bruise central bank egos, but I can’t see it violating the spirit and ultimate goals that led to targeting in the first place.
Senexx – Please call me on it if I have a post that makes no sense, because it means one of three things.
1/ I communicated ideas poorly.
2/ It’s nonsense.
3/ The ideas don’t have a good vocabulary yet.
The first two are probable and I want to avoid them. The third is hard to avoid, but also unlikely given my modest capacity.
Derry D – In the current Australian context I’d also point out the revaluing is useful with property bubbles (at least of modest proportions) since we can correct real prices without nominal prices falling with nominal debt constant.
An older analysis from Brad DeLong.1995.
http://econ161.berkeley.edu/econ_articles/theinflationofthes.html
As a voter this issue has some appeal for my attention if what looks like a sacred piece of government is going to change.
So first how truely independent are the various central banks? Presidents can’t exert sufficient pressure on selection?
It was my understanding , as a layperson the emergence of inflation of the 70s was not understood and the result was targets and banding.I’m not sure it is understood why that bout of inflation resolved.
Just seems like a risky response to the GFC.
Richard, for what little it is worth it was primarily your first blog I think about Institutions and the couple after that if I am remembering correctly (I haven’t actually checked). It probably only made little to no sense to me. Your co-authors probably understood it perfectly.
I did read many of your more recent blog posts yesterday and had little to no trouble.
The intellectual background is more complicated than suggested here. Point one is the acknowledgement that there is no long run tradeoff between inflation and output and unemployment. So, stabilizing inflation at a higher level doesn’t mean easier financial conditions once the transition to the new target is achieved (policy still has to be neutral when there is no output gap). Second, is inflation bias. One of the motivations for independent central banks was that they would have more credibility because they didn’t face the same political incentives to run the economy above capacity. In Australia’s case, central bank independence was seen (and in the UK) as a way to improve credibility as the central bank tried to reduce inflation to more reasonable levels and then keep it there. Third, there is the separate debate about the optimal level of any inflation target. Here things get trickier. On the one hand a higher inflation target can help grease the wheels of the labour market (as DD says). On the other, higher rates of inflation tend to be more distortionary and are associated with more regular price resetting (with obvious menu costs). In the US at least there doesn’t seem to be much evidence that the Fed has had a deflationary bias over the past decade or so. They helped blow two major asset bubbles with policy that was too loose. And although the current environment may be disinflationary in the US, that doesn’t mean that it will be in the medium-term. Monetary policy institutions need to be appropriate for a range of macroeconomic environments.
Could I get the term disinflation explained please? I’ve seen it several times and the best I can come up with is decelerating inflation but maintaining an inflationary rate.
LO is mostly right but Bernie Fraser brought in an inflation target AFTER inflation had been tamed.
It could be argued Independence helped in keeping the credibility although the RBA was already independent because of Bernie Fraser before it ‘became’ independent.
Senexx yep you have basically got it right. We had disinflation in the early 90s.
You need to avoid deflation like the plague however.
Disinflation is simply a reduction in the rate of inflation short of becoming outright deflation.
VT – yes, the formal inflation target was introduced after inflation had been reduced, but I would argue that the inflation target and independence helped to ensure that inflation expectations remained well anchored over the course of the next decade. Stable inflation is not just about changes in policy rates, it is about the credibility of the overall regime.
LO – Which is what brings me to my main question (the title originally was “The point of central bank “Credibility””. Is credibility determined by how well outcomes (i.e inflation rates) match to a fairly arbitrarily drawn bracket? Inflation expectations are a product of the whole population that deals with prices, but knowledge of the official CPI rate and it’s relationship with the bracket is the domain of a small population which may be inflated their own importance.
I’m not even sure to the degree that when forward prices are negotiated expectations are based on something other than rough linear extrapolation of the preceeding few years price changes, for instance anticipating central bank action, let alone their action based on their ability to stick rigidly within the bracket.
I occasionally see a strong form (which I don’t think you’re advocating)of the argument that goes “The unions must know they will be punished if they push too hard with wage claims”, which is very strange for a number of reasons, of which the idea of collective punishment is just a start.
Basically, I’d need convincing that inflation dynamics and expectations are substantially based on more that the experience of price changes, based both on the knowledge available to relevant actors (i.e everyone) and the vagueness of “credibility” as a notion. It is possible that spectators and commentators are unduly influential on inflation dynamics based on their notions of credibility, but the burden of proof lies with those who suggest an outsize importance for themselves based on an unquantifiable variable.
Richard, my points aren’t arbitrary. There is a pretty solid literature looking at different measures of central bank transparency and credibility and other institutional features and relating that back to inflation variability and other things central banks care about. For example, applied work will look at whether the variation in the magnitude of second round effects following negative supply shocks is related to countries’ monetary policy regime. The issue of whether inflation expectations (and wage negotiations) are adaptive or forward looking (or a combination of the two) is also the subject of a pretty substantial literature. The basic point is that in an environment where a central bank is unsuccessful in convincing households, firms and market participants that they are committed to price stability (which can be defined in a variety of ways) the effects will show up in a variety of ways – wage determination, asset prices, and price setting more generally. Anyway, the best thing is for you to take a look at the empirical literature that is already out there and decide whether you find it convincing or not.
If you want an example of what a central bank without credibility looks like and what the consequences can be, take a look at Turkey at the moment. Inflation expectations can become unanchored much more easily and quickly than you think.
I never meant to imply your points were arbitrary. Consistent with the post and the topic in it, the target bracket is arbitrary. i.e The choice of 2-3 as opposed to 1.5-2.5 or 2-3.5 or whatever isn’t grounded in a deep and rigourous analysis. It’s a rule of thumb to get to a policy goal that is based on rigourous analysis (i.e inflation below spiral territory). The credibility of the authorities is based on the latter goal, not adherence to the arbitrary target boundaries.
Were the BoE fearful that a looser adherence to their target, or even the choice of a higher target, would turn them into Turkey, I’d probably reply “If you want to see a central bank that acts [fn1] as if the target is a guideline rather than something that requires rigorous adherence but can still avoid breakouts and retain credibility, look at the RBA”.
I fear you may be placing me in a role defined by sides in debates made when I was a child, rather than anything I’m saying.
[fn1] regardless of what they might think or say, they certainly act this way
The Bank of England haven’t really behaved any differently than the RBA as far as I can tell. Inflation has been running well above their target for more than a year and they haven’t tightened policy at all. All inflation targeters use discretion and flexibility in their pursuit of price stability.
A higher target really wouldn’t help the UK very much either. Let’s say they increased the target to 3%. Policy might be looser for longer, until underlying inflation increased to around the new target, and then it would have to react to the state of the economy and changes in inflation expectations to ensure that the new target also wasn’t breached. It would be no more than a stop gap measure. As I said in an earlier comment – there is no long run tradoff between output (and unemployment) and inflation. Central banks can put different weights on output and inflation variability of course, but the experience of the last decade or so isn’t consistent with them putting excess weight on inflation.
The RBA btw is very concerned about its own target. You are right that that the intitial target is somewhat arbitrary. But once determined, credibility is bound up with achieving that target on average over the medium term. Thus, the target is no longer arbitrary but a critical part of maintaining credibility. Whenever the RBA’s own forecasts are for inflation to rise above (or fall below) the target range by the end of the forecast horizon, you will almost always see them changing policy.
Maybe I can put the point a little differently, with respect to the UK. Again, assume that they raise their inflation target to 3%. They go through another business cycle and then in 10 years there is another major recession, with unemployment increasing sigificantly. Although the higher starting point for inflation would raise real adjustment a little, the higher inflation target would not be enough on its own to prevent the recession and again the Quiggin’s of the world would be talking about how arbitrary the 3% target is, and how allowing inflation to increase wouldn’t hurt. Afterall, 3% inflation isn’t particularly harmful. So, the target is then raised to 4%. Wash, rinse, repeat. 5%, 6%, where does it stop? However arbitrary at the beginning, central banks have to establish an inflation rate (or path for the price level) that they deem consistent with price stability and stick to it. The UK’s problems weren’t caused by their inflation target and the weakness of the current recovery is not being caused by tight policy now. The problems lie in the financial sector and they way it was regulated before the crisis and how it is being managed now. Deleveraging is painful.