ONCE, a central banker who did not believe in monetarism would have been viewed as equivalent to a priest who admits to being an atheist. A quarter of a century ago, control of money was seen as both necessary and sufficient to curb inflation—so most central banks set monetary targets. Monetarism has since become unfashionable. Financial deregulation and innovation made the money supply harder to interpret, let alone control. As the link between money and prices seemingly broke down, central banks scrapped money targets and instead focused on inflation directly. Or as Gerald Bouey, a former governor of the Bank of Canada, once said, “We didn't abandon the monetary aggregates, they abandoned us.”
Today, America's Federal Reserve barely glances at money. Indeed, from this week it will stop publishing M3, its broadest measure of money. The Fed claims that M3 does not convey any extra information about the economy that is not already embodied in the narrower M2 measure, so it is not worth the cost of collecting it. It is true that the two Ms move in step for much of the time, but there have been big divergences. During the late 1990s equity bubble, for example, M3 grew faster; over the past year, M3 has grown nearly twice as fast as M2. So it looks odd to claim that M3 does not tell us anything different. The Fed is really saying that it doesn't believe money matters.
It is ironic that the Fed is dropping M3 only days after a conference was held to honour Otmar Issing, chief economist of the European Central Bank (ECB) and the architect of its highly money-oriented policy (see article). At that meeting many economists and former central bankers from around the world expressed unease about the recent rapid pace of growth in global money and credit. As Milton Friedman famously said, “Inflation is always and everywhere a monetary phenomenon.” Monetary aggregates are a fickle guide to the economy over the next year, but over longer periods the link between money and prices still holds. Many big mistakes in economic history were made when policymakers ignored monetary signals: the Great Depression in the 1930s, the great inflation of the 1970s, and the financial bubbles in Japan in the late 1980s and East Asia in the late 1990s.
Those experiences surely suggest that central banks should keep a close eye on the growth in money alongside their immediate inflation goals—which is exactly what the ECB has done, with its much (and unfairly) criticised “two-pillar” strategy. The first pillar looks at how economic activity might influence inflation over the next year or two; the second focuses on the growth in money as a cross-check on medium- to long-run developments. The Bank of Japan's new monetary-policy framework adopted earlier this month has borrowed several elements of this approach. The Bank plans to focus on price stability and growth one to two years into the future, but also to carry out a broader assessment of medium- and longer-term risks, such as asset prices and credit growth. So, in scrapping M3, the Fed is looking like the odd man out.
Research by the Bank for International Settlements has confirmed that monetary aggregates do still contain useful information. In particular, rapid growth in money and credit as well as asset prices usually signals the build-up of economic and financial imbalances, which often cause financial stress later on. Central banks cannot use the money-supply numbers as a way to set monetary policy on auto pilot: but they would be foolish to ignore the hazard warning lights.
