George Osborne says yes. I say no.
For the past two decades, since its introduction by the Reserve Bank of New Zealand, the dominant monetary policy framework has been inflation targeting. This replaced the concepts of the previous generation, such as exchange rate targeting (e.g. in the ERM) or money supply targeting (as in the early Thatcher period).
Under inflation targeting in its best-known form, the government sets the central bank an annual target for inflation which the central bank then aims to deliver. For example, in the UK the Chancellor each year writes a letter to the Bank of England setting that year's target rate of inflation. Since 2003 the target index has been the "consumer prices index" and the target rate of inflation has been 2%. Until 2007 it was understood that the Bank had a 1% margin of error either side of its target - so, in the short-term it was allowed for inflation to fall to 1% or rise to 3% provided that over the medium term inflation would return to the 2% target. After this target was missed in March 2007, with CPI rising to 3.1%, the MPC argued that this was no more a violation of its target than inflation at 2.9% - a view that I condemned vigorously at the time (I set out my position in some detail here). In 2008 inflation rose considerably above the 3% threshold for many months (and urgings from various quarters, including me, for the government to raise the inflation target in acknowledgement of this reality were ignored), and in 2009 CPI is expected to fall well below 1%. The UK's inflation target has collapsed.
In addition, inflation targeting has come to be seen to be a failed regime. It didn't stop a credit boom and bust (with M4 growth above 14% at one stage and now expected to go negative); it didn't stop a housing boom and bust (with prices rising at 25% per year and now falling at 15% per year); it didn't stop inflation going to the highest level for decades, and it isn't expected to stop deflation; and it didn't stop a general growth boom being followed by the worst recession in at least three decades and perhaps in nearly nine decades. Since it achieves none of these things, what is it good for?
A number of financial journalists, particularly in the US (where inflation targeting was never formalised), and represented in the UK particularly by The Economist magazine and by Roger Bootle of Capital Economics, had always expressed doubts about the focus on inflation, arguing that policymakers should not care only about retail or consumer price inflation, but also about inflation in asset prices - particularly in the price of houses or of shares in the stock market. They advocated a regime in which monetary policy decisions (e.g. whether interest rates go up or down) should take asset price movements into account. The idea was that policymakers might spot "bubbles" developing in asset markets and raise rates so as to "prick" these "bubbles" so they would deflate in a controlled manner rather than bursting suddenly, causing disruption. (A few academics also supported this position, but mainstream opinion was against them for reasons of the sort I shall set out shortly.)
Probably the best-known opponent of the asset price targeting concept was Alan Greenspan, Chairman of the Federal Reserve. His view was that, although central banks might be able to note that asset markets were behaving "irrationally" ("irrational exuberance" was his famous description of the stock market in the late 1990s), the proper role for central banks was to place themselves well to "mop up afterwards" when the bust occurred.
At this point in the discussion we should make a a couple of vital distinctions, underestimated by almost everyone except those academics and central bankers that specialise in the area. First we need to distinguish between (i) the monetary policy regime called "inflation targeting" (the regime we have in the UK); (ii) focusing exclusively upon inflation when operating under the monetary policy regime called "discretion" (what Greenspan is sometimes alleged to have done in the US). This may seem arcane, but you really have no hope of understanding anything about the relative merits of different monetary policy regimes without grasping this first point.
Under inflation targeting, as we mentioned above, the monetary authority (e.g. the central bank) is set (or states itself) a target for the annual inflation rate. Since it has a target, something counts as "missing" its target. The central bank's performance can therefore be assessed by comparing the actual outturn for inflation versus the target. In contrast, a central bank like the Federal Reserve, operating under "discretion", has no formal target for anything. It just sets policy as it sees best. If, for a time, it happens (without specifying any public target for it) to think the most interesting feature of the economy to consider when formulating its interest rate decision is consumer price inflation, that is just an exercise of its discretion. It does not make it an inflation targeter.
Next we need to distinguish between three things: (a) the central bank being set a formal target for asset prices; (b) the central bank taking close account of asset prices in attempting to meet an inflation target; (c) the central bank taking close account of asset prices when exercising discretion.
Once we see that (a), (b) and (c) are separate, we should see that arguments for doing (c) - for example, arguments that the Federal Reserve, as a central bank with discretion, ought to take account of asset price movements - are not necessarily good arguments for (a) - i.e. arguments that the central bank ought to be set a formal target for asset prices.
Thus we should understand the Greenspan doctrine as concerning what central banks operating under discretion should do - he says they should, despite having the discretion to act early, ignore asset price movements and focus on inflation. As it stands, this argument is simply not applicable to the Bank of England's own policy setting, for the Bank of England does not have discretion. The closest we could come would be a claim relating to (b) - the question of whether the Bank of England ought to take close account of asset prices in attempting to meet an inflation target. To make the claim that the Bank of England ought to have taken more account of asset price movements, we would need to argue either that significant asset price movements will lead inflation to stray away from its target, or that pricking an asset price bubble will not affect inflation materially but would promote growth (thereby meeting the second part of the Bank of England's mandate - that, subject to meeting its inflation target, the Bank should act so as to support the government's objectives for growth and employment).
I have argued previously that we should not expect asset price movements necessarily to affect inflation rates within the timescales relevant to an inflation target (specifically, on a year-on-year basis out for about two years). As to the latter case, it is pretty clear that bursting asset price bubbles would be likely to have a short-term impact on inflation. For example, the most recent case of asset price targeting in the UK was John Major's raising of interest rates when Chancellor to "take the steam out of the housing market". House prices crashed, presumably by more than was hoped but perhaps by less than if they had been permitted to find their own maximum. RPI inflation rose as rates were tightened, peaking at 10.9% in September/October 1990 (RPIX was 9.5%) before falling back to 1.2% in June 1993 (RPIX was 2.8% then and eventually pitted at 2%). If an inflation targeting central bank were to burst a bubble that was not causing inflation, and that resulted in inflation going under target, that would seem a pretty straightforward violation of its mandate.
So, perhaps the problem is the use of inflation targeting itself? Well, I've long been an advocate of switching away from an inflation target to a price-level target, but the introduction of a formal asset price target would be another matter altogether. Could the central bank (or whoever sets its target) really know better than the Market what was the "correct" as opposed to "bubble" price for the asset? A standard line of thought is that the central bank could not possibly know enough to set any useful target. Even if it thought that prices were probably too high, it could have no accurate concept of how much they should fall.
Worse than this, by definition these "bubbles" are the result of markets behaving in some sense "irrationally" - not making proper use of the information before them, or trapped in momentum beynod the point at which smart participants understood matters could not continue indefinitely. Since behaviour in such circumstances is irrational, the response to interest rate rises would be likewise irrational. For example, those in the asset market might react to asset-pricking interest rate rises by concluding that there is clearly a bubble going here with many people that might prove "bigger fools" than me and buy from me at a higher price, so now is the time to enter - that way, asset-pricking actions might serve to exacerbate bubbles (yes, such behaviour would be irrational and unsustainable, but that's the alleged nature of these situations). Alternatively, even very slight bubble-pricking actions might induce panicky selling with the market deflating in a far more disorderly fashion than if it had been left to itself. (After all, the Federal Reserve tightened monetary policy a bit in the late 1920s to try to deflate the "bubble" in the stock market, and not many people regard subsequent events as an orderly smooth deflation.)
A further problem is better-known and perhaps easier to see than some of the above (actually it's a version of the last argument, but let's think of it separately). If the central bank is targeting the price of an asset, then the price of the asset today will embody the market's expectations about how the central bank will react to today's price. In other words, prices of assets cease to reflect some underlying value of the asset, and start instead to be distorted by expectations of policy response. In this way we replace occasional distortions of asset prices (during bubbles) - which might create economic inefficiency - with permanent distortions of their prices (by the policy) - which definitely creates economic inefficiency.
(In addition to these arguments, those academics keen on inflation targeting produced all kinds of theorems to "prove" that asset price targeting added nothing to an efficient inflation target whilst fans of asset price targeting produced their own counter-theorems - my sympathies in this case are fairly squarely with the inflation targeters, though their models usually proved much less than they thought (as did the models of the asset price targeters) because their simplifying assumptions eliminated the most relevant real-world cases (specifically, for technical nerds, in my view the strategy space considered in standard models (Markov and quasi-Markov strategies) was too restricted - don't worry what this means if you don't already know).)
OK. So if you've bought my arguments so far you might think that it is infeasible to have some kind of formal asset price target. But perhaps, instead, we could have something vaguer - insisting that in addition to taking account of inflation, the Bank of England also took into account asset prices? After all, the European Central Bank isn't an inflation targeter - it takes account of inflation, to be sure, but it also takes account of money supply growth.
Well, of course. We could do that. But over the past twenty years monetary economists have thought that there were considerable advantages to not simply having vague targets but, rather, to having explicit targets in which there was a concrete sense of when they had been met and when missed. But using a concrete target for something visible (e.g. inflation, but the argument would also apply to an exchange rate target, say) the monetary authority is being transparent and accountable. This was thought to increase credibility, following the experience of the previous two decades in which many monetary authorities had lost credibility by allowing more inflation than they said they intended.
I have always thought of inflation targeting as a framework of constrained discretion (this is Ben Bernanke's term). What it aims to do (and actually does quite successfully apart from its flaw in terms of asset price bubbles and a bit of a problem dealing with a liquidity trap) is to allow the central bank to focus on stabilising growth and employment in the economy, subject to meeting the target for inflation. It is thus essentially a growth- and stability-promoting framework. The target acts as a focus for inflation and interest rate expectations, reducing policy uncertainty considerably and granting the central bank sufficient discretion to react to events so as to smooth out growth. It is a beautiful and elegant solution, precisely because it is so exact and simple. Politicians can debate what is the appropriate inflation rate to target each year (that they have not done so properly in the UK is a failure of democracy and I believe reflects total confusion about the merits and nature of central bank independence - it is not a failure of inflation targeting itself). (Indeed, this ability to debate regularly what annual inflation rate is best is probably the single most important advantage of inflation targeting over price-level targeting in a developed economy.) Consumers and businesses can understand well what is being targeted (Who understands what a Sterling M3 target means? Who does not, at some level, understand what an inflation target means?). Because voters and consumers can understand the target, expectations formation is eased and democratic accountability is possible in a way it would not be for vague targets.
The elegance, transparency, simplicity, accountability, and credibility of an inflation target are not things to give up lightly. I agree absolutely that inflation targeting has failed. But we should not forget or underestimate its strengths. In my view, we should seek a new monetary policy framework that has as many as possible of the strengths of inflation targeting, but without its key weaknesses. The most important weakness was that I have discussed previously - its tendency to create asset price bubbles and then act against them too late. But there is another framework that is also simple, elegant, transparent, credible and accountable, and which represents only an incremental development of inflation targeting - absorbing its strengths but moving on - but which does not share its key weakness. That framework is price-level targeting.
(Of course, under price-level targeting as I envisage it the measure of inflation would include housing costs (unlike the CPI targeted since 2003, a change I objected to - my single most important achievement in politics was convincing the Conservative Treasury team in 2003 to oppose this switch). But housing costs, properly calculated, are not quite the same thing as house prices, and insofar as they did reflect house prices that would be a version of (b) above - taking asset price movements into account within a price-level target rather than targeting asset price movements themselves.)
Don't throw the elegant baby out with the burst bubble bathwater. Price-level targeting allows us to react to asset price movements within a framework very similar to that which we thought was serving us well these past seventeen years. Asset price targeting would be vague, distortionary, as likely to create unnecessary busts as to prevent them, and would imply some mechanism by which government decides the "right" price contrary to the Market. Let's not go there.
