In this essay he warns in general against the danger of following slogans (for which he here uses the term "shibboleths",) expressing simplistic ideas which "often become badges of identity for groups of like-minded people, who repeat certain phrases to each other, and eventually mistake repetition for truth." The particular slogans he argues against – with his customary gusto and rigour – in the two main sections of this essay are:
- that central banks should aim for fast growth, as America's Republican Party and much of its business press demand, and
- that central bank policies should pursue the goal of price stability, i.e. zero inflation, which is what the central banks of most countries claim to be doing.
Suppose you reject both the miracle cures of the growth sect and the old-time religion of the stable-price sect. What policies would you advocate?Central banks may have to avoid intelligibility and to spout slogans, such as "zero inflation", lest the sensible policies most of them follow are pounced upon. But among the dangers of spouting slogans is that one may start to believe in them oneself – as Krugman warns is happening in some countries. He concludes:A shibboleth-free policy might look like this. First, adopt as a long-run target fairly low but not zero inflation, say 3-4%. This is high enough to accommodate most of the real wage cuts that markets impose, while the costs of the inflation itself will still be very small. Monetary policy, however, affects inflation only with a long lag, so you also need some intermediate target. A reasonable strategy is to try to stabilise unemployment around your best estimate of the level consistent with stable inflation at the desired rate, even while recognising that such estimates are imperfect. So you should be prepared to adjust the unemployment target if inflation is worse or better than expected. And of course if past misjudgments have caused inflation to move above – or below – the target, policy must seek to bring it back into line.
This proposal will presumably bring angry objections from both sides. The growth sect will denounce it as an acceptance of defeat, insisting that the West needs higher growth to raise living standards and solve its budget problems. Unfortunately, economics is not only about what you want, it is also about what you can get. Growth may be good, but achieving it requires more than simply declaring inflation to be dead.
Meanwhile, the stable-price sect will denounce this strategy as irresponsible, a return to the bad old inflationary ways of the 1970s. But the strategy is not outlandish – on the contrary, it is intended to be a description of the actual policies followed by several of the world's big central banks. In particular, what I have described is close to the behaviour predicted by the "Taylor rule". This rule, which suggests setting interest rates automatically based on a comparison of the actual and potential output of an economy, successfully tracks the policies of the Federal Reserve. (It is ironic that the Fed, whose policies are in fact more growth- and employment-oriented than any other western central bank, is the target of most of the growth sect's attacks.) But the strategy described is also arguably a pretty good description of the behaviour of other central banks, including the Bank of England and even – dare one say it? – the Bundesbank, which talks a monetarist game but rarely meets its own announced targets.
Shibboleths make you feel good. They are an alternative to the pain of hard thinking and, because so many people repeat them, they offer a reassuring sense of community. But you must go beyond shibboleths, however comfortable they make you feel: monetary policy is too serious to be conducted on the basis of slogans.
Under the heading "Better off, but not much," The Economist reported on October 4th 1997 as follows:
According to the Census Bureau's annual report on poverty and income, the typical American household's income rose by 1.2% in real terms in 1996 ... By many measures, the country's wealthiest families did better than the poorest ones. Average income for the poorest fifth of American families fell by 1.8%, while average income for the top fifth rose by 2.2%. This explains why America's poverty rate did not fall despite an increase in median household income.Figures from the Bureau of the Census also show that the median household income has stayed almost constant during the 20 years to 1996, but that the share of aggregate household income of the top fifth of households has risen from about 43% to about 49% during that same period, while that of households in the other quintiles has fallen.