The Most Important Number in Economics Was Plucked Out of the Air
On April 1, 1988, New Zealand Finance Minister Roger Douglas appeared on television and casually floated an idea: inflation should eventually fall to roughly zero to one percent. The remark caught almost everyone off guard, including the Reserve Bank of New Zealand, which had neither been consulted nor prepared a strategy for achieving it. Over the weeks that followed, officials worked to turn Douglas's spontaneous suggestion into an actual policy and eventually settled on a target band of zero to two percent. Douglas was particularly taken with the proposed timetable for reaching it — and with the memorable phrase attached to it: "0 to 2 by '92." The slogan endured, the range became official policy, and in 1989 New Zealand made history by becoming the first nation to put an inflation target into law.
Don Brash, who led the Reserve Bank at the time and was responsible for delivering the result, later offered an unusually frank account of where the number came from. According to him, the figure was essentially "plucked out of the air."
The surprising part came afterward. The policy appeared to work: after nearly twenty years in which inflation had routinely reached double digits, New Zealand brought it down and managed to keep it there. Other countries noticed. Canada adopted an inflation target in 1991, followed by the United Kingdom in 1992, and during the next two decades the approach spread across most advanced economies. The United States was among the last to make the commitment formal. The Federal Reserve finally did so in January 2012, when Ben Bernanke's FOMC identified two percent as the rate "most consistent over the longer run" with its mandate. The European Central Bank similarly spent years targeting inflation "below, but close to, 2%" until a 2021 review recast the objective as a straightforward, symmetric two percent. In little more than a generation, a number improvised in Wellington had become something close to a constitutional principle for the global economy.
There are legitimate arguments for keeping inflation low without trying to eliminate it completely. Official inflation measures probably exaggerate the underlying increase in prices to some degree, meaning a reported two percent rate may correspond to something slightly lower in reality. A modest amount of inflation can also make labor markets more adaptable by allowing real wages to change without requiring employers to impose the nominal pay cuts workers strongly resist. And keeping inflation above zero gives central banks more room to reduce interest rates during downturns before they run into the effective lower bound. Those considerations support a small positive inflation rate. They do not, however, explain why the ideal figure should be two percent instead of one and a half, or three. For decades, economists have developed increasingly sophisticated explanations for a number that was chosen before the theory existed — almost as though the academic justification were written after the verdict had already been reached.
Every so often, a prominent economist calls attention to that awkward history. Olivier Blanchard, who was then the IMF's chief economist, suggested a four percent target in 2010 and has continued to make the case for three percent. His reasoning is that the final leg of disinflation, particularly the move from three percent to two, can impose disproportionate economic costs in the form of weaker growth and higher unemployment, while producing benefits that are difficult to identify with precision. The usual response is not that two percent has been proven to be economically optimal. Instead, policymakers argue that moving the target would damage credibility. Once markets see the number change, they may assume another adjustment could come later, weakening the anchor itself. That leaves us with an unusual arrangement: two percent is defended less because it is demonstrably the correct number than because people have come to believe it is the number. The belief creates the credibility, and the credibility makes the target effective. A number chosen arbitrarily can therefore work better than a supposedly superior number that nobody accepts.
That credibility question has become especially relevant this month. U.S. inflation came in at 3.4 percent in July — considerably nearer the Federal Reserve's goal than it has been in much of the recent past, but still 1.4 percentage points above it. Debate over how aggressively policymakers should continue pushing inflation lower has even surfaced within the Federal Reserve, where three governors broke ranks in the same direction at one meeting, the first time that had happened in a decade. The dispute is ultimately about how much economic pain should be tolerated in pursuit of a threshold whose origins trace back to a politician making an offhand remark on a television broadcast. There is something almost religious about that transformation: a highly precise number can acquire authority it never originally possessed, and once institutions build their policies around it, questioning its foundation becomes increasingly difficult.
The most generous interpretation is that the lack of a rigorous origin story was never necessarily a weakness. Modern economies depend on coordination, and coordination often requires a common reference point — a number that households, businesses, investors and governments can all use when making plans. Its value comes less from being mathematically perfect than from being widely shared. Two percent prevailed because it was available, easy to remember and attached to a catchy phrase and deadline, not because an economic model had produced it. Seen that way, New Zealand's improvisation was not a defect that somehow grew into today's monetary system. It was the moment the system acquired a focal point. Someone had to choose a number that others could organize around. Nearly four decades later, central banks around the world are still honoring a promise Roger Douglas never intended to make in quite the way it turned out — and the remarkable thing is that the promise appears to work.