This pulls together tweets that I sent on the history of inflation targeting, having read this article in the New York Times.  Warning:  this is a highly subjective, UK-centric and stream-of-consciousness, and still very brief ‘history’.

A key thing to recall about inflation targeting was that it was about targeting the only thing that central banks had not yet tried targeting and failed.  It came after broken promises to target the relative price of money and gold;  the exchange rate;  and the growth rate of money aggregates;  even after periods when, confused about the difference between an instrument and a target, central banks ‘targeted’ the interest rate.

That statement leaves out nominal GDP targeting.  Despite elegant works by Meade and others, this never seemed to be a runner at the time.

Another feature of inflation targeting, remembering the haste with which it was embarked on, particularly in the UK, was ‘we have to target something, inflation is something, so we have to target inflation’.

At the outset, what seemed scary about doing it was the notion that you can just promise to target the thing that policy really cares about, rather than specifying a value for an intermediate target like money or the exchange rate.

One way of interpreting this new ‘promise’ was:  ‘we won’t tie our hands, because we’ve tried that before, and found that we always have to untie them again;  so instead we will just do it.’  That borrows from Bennet McCallum’s use of the Nike advertising campaign back in the day:  instead of making a commitment, just do good policy.

Naturally, there were sceptics.  Why would markets believe a promise just to get inflation down, when promises to keep intermediate targets had been broken?  The answer was that the lack of resolve that had led to those past failures had their ultimate cause in the variable link between intermediate targets and the ultimate goal, meaning that there were times when, with respect to the ultimate goal, the intermediate target would be better broken.  There was no better demonstration of that than the UK’s exit from the ERM in 1992 which precipitated inflation targeting.

These flaws with intermediate targeting were known at the time, but the dominant view was that the credibility benefits of sticking to a verifiable intermediate target trumped the credibility costs of promising to do something that was occasionally harmful.

I wonder too if part of the reason for pre-inflation targeting strategies was the lack of understanding about what caused inflation.  Money growth and exchanges rates were a monetary policy phenomenon, but inflation was to do with costs, trade unions, oil prices, and lots of stuff that wasn’t monetary policy.  So how could a promise be framed in terms of a variable so little under central bank control?

Subsequently, inflation targeting central bankers enjoying the good times of macro stability would crow about the optimality of their regime, but the benefits of final goal targeting were not at all universally subscribed to at the outset.

Another aspect of the history not mentioned in that NYT piece is the shift from lexical mandates that stressed the primacy of the inflation goal, using ‘subject to’ texts to mention other goals associated with the real economy, towards mandates that emphasised the trade-offs that existed between inflation and other goals.

In my opinion the ‘subject to’ language was always nonsense.  In responding to shocks that did not generate a trade-off, the ‘subject to’ clause was redundant.  Stabilising inflation would stabilise other goals automatically.  In circumstances when there was a trade-off, the ‘subject to’ clause was simply wrong, directing policy, essentially, to ignore the trade-off to the detriment of the economy.

But ‘subject to’ probably seemed necessary at the outset given the worries about our monetary policy misbehaviour in the past, the sense that simply promising to hit your final goal sounded a bit like magic, and the need to sound like you had engaged conservative central bankers, and not lily-livered ones worried about unemployment.  Who, in the aftermath of our ejection from the ERM, could imagine a Chancellor declaring that ‘we will henceforth target a weighted sum of the variance of inflation and resource utilisation’?  Though that is precisely where the logic of giving up on intermediate targets leads, ultimately.

Central bankers did slowly become bolder and less confusing about the trade-off language.  Two reasons:  1) They were emboldened by the apparent taming of inflation, and perhaps felt that there was a credibility dividend that could be spent.  2) There was a torrent of applied monetary policy work on evaluating regimes articulating why even in the simplest models dual goals were warranted [culminating in Woodford’s Interest and Prices].

But initially – and still occasionally – multiple goal pursuit was hidden in tricksy language about there being only one target, just variations in the horizon at which it was optimal to meet it.

The most recent chapter in inflation targeting history has been the great financial crisis, which has had many consequences.

The first is that it retroactively boosted a line of thought that held that inflation targeting had led to a neglect of asset prices, and this had brought about financial instability.  The second is that it thrust interest rates at the zero bound, essentially handing back to fiscal authorities the job of hitting the inflation target.  The third was a bursting of the bubble of thought that had clearly greatly exaggerated the contribution of inflation targeting and central bank independence to macroeconomic stability.  A fourth is the unprecedented level of political controversy generated by persistently low interest rates, and quantitative easing, both of which are resented in some quarters as conspiracies to aid the rich, borrowers, or both.

The criticism that inflation targeting ignored asset prices – the BIS were the most persistent advocates of this – was always wrong.  It was sometimes based on a misunderstanding of the capability and inclination of inflation targeters to respond to multiple goals, to things not defined in their headline quantified index.  It also exaggerated the power of monetary policy to do anything about what, in essence, was a ‘real’ and not a ‘monetary policy’ phenomenon, whose root cause lay in inadequate regulation, not loose monetary policy.  This critique therefore misconstrued symptoms as causes.  The focus on fine tuning the details of inflation targeting practice, and lack of focus on regulation were both caused by a failure to see the risks building up in the system.

The accidental return of the job of hitting the inflation target to the fiscal authorities [who had delegated it in the first place] was unfortunate, and came at a time when those authorities faced their own credibility issues attempting it and when fiscal policy was so politicised that fiscal branding took precedence over confronting the technical problem of assisting monetary policy.

A highlight of the bubble bursting on the contribution of monetary policy frameworks to stability was of course the UK Bank of England conference on the ‘Great Stability’, which took place in September 2007, and was punctuated by the senior attendees leaving to follow up the breaking news they had seen on their Blackberries.

This broad brush history misses out some of the details that transfixed central bank economists.  The developments in communication towards increased, but still incomplete transparency.  The use and abuse of measures of ‘core inflation’.  Discussion of point versus range targets;  of price level versus inflation targets.  Developments in the method of estimating price increases and biases that remained.  The spread of DSGE models in central banks.  The curious phenomenon of the ECB and Fed inflation targeting quietly while saying that they weren’t.

It also begs questions about the future:  how inflation targeting should be reformed to better weather future macroeconomic storms.  Much ink has been spilled on that, so I won’t repeat here.