The World’s central bankers are a close-knit club, given to fads and fashions. In the early 1980’s, they fell under the spell of monetarism, a simplistic economic theory promoted by Milton Friedman. After monetarism was discredited – at great cost to those countries that succumbed to it – the quest began for a new mantra. The answer came in the form of “inflation targeting,” which says that whenever price growth exceeds a target level, interest rates should be raised. This crude recipe is based on little economic theory or empirical evidence; there is no reason to expect that regardless of the source of inflation, the best response is to increase interest rates....That is somewhat unfair - inflation targeting is grounded in economic theory. In particular, it provides a mechanism to improve the credibility of monetary policy in the face of the time-consistency problem (the discovery of which earned a Nobel Prize for Kydland and Prescott). Anyhow... Stiglitz continues:
Today, inflation targeting is being put to the test – and it will almost certainly fail. Developing countries currently face higher rates of inflation not because of poorer macro-management, but because oil and food prices are soaring, and these items represent a much larger share of the average household budget than in rich countries. In China, for example, inflation is approaching 8% or more. In Vietnam, it is even higher and is expected to approach 18.2% this year, and in India it is 5.8%. By contrast, US inflation stands at 3%. Does that mean that these developing countries should raise their interest rates far more than the US?
Inflation in these countries is, for the most part, imported. Raising interest rates won’t have much impact on the international price of grains or fuel...
It should be noted that the countries Stiglitz mentions are not inflation targeters; the practice is more common among high income countries like Australia, Canada, New Zealand and the UK. The issue of how central banks should respond to commodity and energy price shocks applies in rich countries, too (though to a lesser degree because food and energy have less weight in the price indexes of high income countries). Back to Stiglitz:
Raising interest rates can reduce aggregate demand, which can slow the economy and tame increases in prices of some goods and services, especially non-traded goods and services. But, unless taken to an intolerable level, these measures by themselves cannot bring inflation down to the targeted levels. For example, even if global energy and food prices increase at a more moderate rate than now – for example, 20% per year – and get reflected in domestic prices, bringing the overall inflation rate to, say, 3% would require markedly falling prices elsewhere. That would almost surely entail a marked economic slowdown and high unemployment. The cure would be worse than the disease.
So, what should be done? First, politicians, or central bankers, should not be blamed for imported inflation, just as we should not give them credit for low inflation when the global environment is benign. Former US Federal Reserve Chairman Alan Greenspan, it is now recognized, deserves much blame for America’s current economic mess. He is also sometimes given credit for America’s low inflation during his tenure. But the truth is that America in the Greenspan years benefited from a period of declining commodity prices, and from deflation in China, which helped keep prices of manufactured goods in check....
At VoxEU, Axel Leijonhufvud makes a similar argument:
Critical to the central banking doctrine was the proposition that monetary policy is fundamentally only about controlling the price level. Using the bank’s power over nominal values to try to manipulate real variables such as output and employment would have only transitory and on balance undesirable effects. The goal of monetary policy, therefore, could only be to stabilise the price level (or its rate of change). This would be most efficaciously accomplished by inflation targeting, an adaptive strategy that requires the bank to respond to any deviation of the price level from target by moving the interest rate in the opposite direction.
This strategy failed in the United States. The Federal Reserve lowered the federal funds rate drastically in an effort to counter the effects of the dot.com crash. In this, the Fed was successful. But it then maintained the rate at an extremely low level because inflation, measured by various variants of the CPI, stayed low and constant. In an inflation targeting regime this is taken to be feedback confirming that the interest rate is “right”. In the present instance, however, US consumer goods prices were being stabilised by competition from imports and the exchange rate policies of the countries of origin of those imports. American monetary policy was far too easy and led to the build-up of a serious asset price bubble, mainly in real estate, and an associated general deterioration in the quality of credit. The problems we now face are in large part due to this policy failure.
To be clear, the Fed is not an explicit inflation targeter, though it does seem to have an informal "comfort zone" of 1-2% for core inflation (i.e. inflation calculated with food and energy prices removed).
Inflation targeting has generally considered to be a success: it is credited with helping to reduce and stabilize inflation, and anchor inflation expectations in the countries that have adopted it.
The problem now arises of responding to external shocks (although Leijonhuvfud sees a failure in the Fed's past response to positive shocks - of course, there weren't too many complaints at the time...). Accommodating the shock - i.e. allowing the inflation rate to rise above the target level - would damage credibility, but maintaining the target means reducing aggregate demand so that the average price level meets the target - this means lower, or even negative, changes in the prices of other goods.
Supply shocks, a.k.a. "cost push inflation," are not a new dilemma for monetary policy, but this is the most significant instance of this problem since inflation targeting became widespread in the 1990's.
Willem Buiter takes a much different view from Stiglitz and Leijonhufvud. In a Maverecon blog post, he asks "who or what causes inflation?" and answers:
This one is easy. In a fiat money world, central banks cause inflation, or, more precisely, only central banks are resposible for inflation. Other shocks, real and nominal, can influence the general price level if the central bank does not respond swiftly and determinedly, but these non-central bank-induced changes in the general price level can always can be offset by the central bank, given enough time, freedom to act and courage.
So, in the medium and long term (at horizons of two years and over, say) central banks choose the average rate of inflation. Not globalisation; not indirect taxes; not bad harvests; not OPEC and the price of oil; not the Chinese and their exchange rate management. There is no oil inflation, food inflation or cost-push inflation. There is just inflation. Inflation may be accompanied by changes in key relative prices - in the real prices of oil, of food, of oil and of labour for instance - if other relative demand and supply shocks accompany the inflationary impulses created by the central bank. Large increases in the real price of food will be bad news to food importers (including most urban households) and good news to rural food producers and exporters. But don’t confuse it with inflation....
I am willing to grant the old-Keynesians and new-Keynesians among us, the empirical regularity that at very high frequencies, the fact that most nominal commodity prices (and prices of non-core goods in general) are flexible (both ways), while most nominal core goods and services are sticky in the short run. So relative demand or supply shocks that cause the relative price of non-core goods to go up will tend to do so in the first instance through an increase in the nominal price of non-core goods rather than through a reduction in the nominal price of core goods and services; likewise relative demand or supply shocks that cause the relative price of non-core goods to do down will tend to do so in the first instance through a decline in the nominal price of non-core goods rather than through an increase in the nominal price of core goods and services.
So for a given stance of past, current and future monetary policy (as measured by the sequence of past, present and contingent future policy rates), relative demand and supply shocks that cause an increase in the relative price of non-core goods (the kind of shocks we are seeing globally today), will temporarily raise inflation above the level at which it would have been without these relative demand and supply shocks but with aggregate demand and supply at the same level. Such general price level blips work their way through the system quite swiftly; much of it is gone within a year, virtually all of it within two years. They do not ’cause inflation’.
Or, as Milton Friedman famously put it, inflation is always and everywhere a monetary phenomenon.
Update (5/15): The FT's Martin Wolf urges Britain to stay on target. He says: "The lesson of the 1970s was simple: letting inflation rip, to avoid pain in the short run, greatly increased pain in the long run. The UK must not repeat that error."