What should central banks target? Since the early 1990s the answer has increasingly been “consumer price inflation”. But this has never been unchallenged. Today, there are four alternative positions. One thinks central banks should target asset prices. Another thinks they should target a “just” interest rate. Yet another thinks they should target real activity. The last thinks they should target some other nominal goal, such as the price level or nominal gross domestic product. These are important debates. But the reality remains: central banking is art, not science. The art must be guided by sensible goals coupled to deep awareness of uncertainty.
Since the early 1990s, the dominant view among central banks and economists is that the best target is inflation. The approach was pioneered in New Zealand in 1990 and quickly followed by Canada and the UK. The US Federal Reserve followed in 2012. The European Central Bank is also effectively an inflation targeter, though its target is a ceiling of 2 per cent and so not symmetrical. According to Paul Fisher of the Warwick Business School, 67 central banks had inflation targets in 2018.
The rationale for inflation targeting has three components. The first is that one instrument — in this case, monetary policy — can only be aimed at one goal. The second is that a central bank can only target a nominal goal of some kind. The third is that inflation is a comprehensible and politically acceptable aim.
A subset of those who think central banks should target asset prices agrees there must only be one goal. But they want them to target the price of a commodity, usually gold. Maybe, at some point, they will want them to target bitcoin as the ultimate reserve asset instead. The objection to a gold standard is it eliminates discretion, which has proved intolerable. Today, the “gold bug” view has become a curiosity.
A more widely-shared view nowadays is that central banks should target the growth of debt or a range of asset prices, such as those for houses or equities. The New Zealand government has, for example, just told its central bank to target house prices. People who think this often assert that explosive rises in asset prices feed inequality and create macroeconomic instability.
There are at least four strong objections to targeting asset prices. First, since asset prices are volatile, targeting them would make monetary policy highly volatile and so generate great macroeconomic instability. Second, nobody knows what the “right” equity or house price or level of private indebtedness should be. Third, central banks must in any case take asset prices and debt growth into account in assessing the state of the economy. Last, if central banks want to influence asset prices or leverage, they should use a wide range of regulatory tools instead.
Perhaps the most grotesque argument is that over wealth inequality. It is absurd to argue for tighter monetary policy and so higher unemployment, which is painfully real for its victims, in order to lower wealth inequality, which is merely a ratio. In 2020, wealth at the 20th percentile of the US income distribution was $6,400 and at the 40th $67,500. The top 0.1 per cent started at close to $43m. What difference would it make to those with next to nothing if the latter were to double? If people want less wealth inequality, they should argue for wealth and inheritance taxes.
A view that is not unrelated to the above is that the central bank should target a positive real interest rate. The implicit belief is that we know what the real interest rate ought to be and that the central bank has been responsible for setting it so low for so long. Neither is justified. The appropriate interest rate depends on economy-wide conditions, especially savings and investment. Moreover, the central bank has to set a rate that is consistent with a macroeconomic equilibrium. Thus, underlying realities ultimately shape what it can do.
Using monetary policy to target asset prices, rather than merely take them into account, is folly. Yet it is not absurd to consider real activity. Indeed, the Fed is already mandated to consider unemployment. Happily, inflation-targeting is consistent with making real activity the ultimate goal. The weaker the relationship between inflation and unemployment, the more the bank’s aim must be maximum employment, subject to the inflation constraint. Recently, the Fed has even shifted towards targeting average inflation. This will allow it to target real activity more aggressively, by compensating for a long period of below-target inflation with one of above-target inflation. Price-level or nominal GDP targeting could reinforce such an objective.
In sum, today’s broad approach to central banking is clearly the least bad. That does not mean it is easy to operate.
It is quite likely that the monetary policy needed to generate maximum employment and stable inflation is consistent with all sorts of crazy activity in the financial system. We are surely seeing plenty of that now. But no sane central bank would deliberately create recessions in order to save finance from itself. Rather, it must save the economy from finance by tough regulation, especially of leverage.
Similarly, it is possible that an exceptional crisis, such as the pandemic, will generate policy mistakes, including unexpectedly high inflation. Andy Haldane, the Bank of England’s chief economist, has stressed just this risk in a thought-provoking recent speech.
Central banks must always be alert. But nobody should imagine there exists an alternative regime that will solve all the difficulties. Central banks will make mistakes. But they must keep it simple.
Follow Martin Wolf with myFT and on Twitter