With the Bank of Canada’s current five-year mandate expiring at the end of 2011, and with the institution enjoying as high a profile as ever, thanks to an uncertain economy and the leadership of its governor, Mark Carney, it’s only right that there should be some public debate about its next mandate.
The bank currently targets the rate of CPI inflation—the growth rate of the consumer price index—and aims to keep it close to 2% annually. Over the past 20 years, it has done so with great success; inflation has varied modestly around target, but the average over two decades has been remarkably close to 2%.
Though perhaps leaving it a little late to weigh in, at the end of October the House of Commons finance committee approved a motion by Scott Brison, the Liberal party’s finance and revenue critic, to broaden the bank’s mandate. Members of the committee are presumably open to many suggestions, but the only two it mentioned explicitly—targeting nominal GDP growth and targeting full employment—would not be improvements. In fact, they would both make Canadian monetary policy worse, not better.
Let’s begin with nominal GDP— the dollar value of the goods and services produced in the economy. The annual growth rate of nominal GDP is the sum of the growth rate of real GDP and the rate of inflation. Many outcomes would be consistent with, say, a 5% growth target for nominal GDP—1% real plus 4% infl ation, 2% real plus 3% inflation, and so on. This would appear to give the bank more flexibility in its actions, because it would place less emphasis on keeping inflation near a constant target.
Aside from the greater variability in inflation that would result—and this is no small thing for firms and workers trying to determine appropriate wages and prices—an important problem with this idea is that we do not live in a world in which the trend rate of real GDP growth is constant. With the aging of Canada’s baby-boom generation, the next 30 years will see a gradual reduction in the growth rate of real GDP by about one percentage point, caused by the boomers’ eventual retirement. If monetary policy began targeting nominal GDP growth, this slowdown in real GDP would force a gradual increase in average infl ation by about one percentage point. But this higher infl ation would not be playing any useful role in the economy; it would just be higher inflation, along with all of its associated costs—especially relevant for the growing numbers of retired Canadians with un-indexed incomes.
An even worse idea is for the Bank of Canada to target “full employment.” This is a slippery concept, and not directly observable in the data. Look all you want on Statistics Canada’s website and you won’t find it. You’ll find data for real GDP, nominal GDP and diff erent measures of infl ation, but you’ll find nothing for the full-employment levels of output, employment or unemployment. For economists and policy-makers, full employment exists when all resources like land, labour and capital are being “fully” utilized. Our theoretical models are typically based on the idea that inflation will be stable only when real GDP is at its full-employment level, and so we use these models and lots of data and fancy statistical techniques to “back out” estimates of the (unobservable) level of full-employment GDP. But since there can be lots of disagreement about our macroeconomic models, there can also be lots of disagreement about the estimated level of full employment. This imprecision makes it a bad choice for a central-bank target: How could we ever judge the success of policy actions?
An even more important reason for not targeting full employment is that the bank could never really expect to succeed. After decades of economic shocks and policy mistakes, and plenty of learning from those errors, central bankers the world over have come to recognize two principles that now form the foundation of their policies. First, high inflation is costly for individuals and firms. Second—and this is the real kicker—inflation is the only macro variable that central bankers can influence in a sustained and systematic way. Monetary policy affects lots of things for a while, but one of the more important insights about macroeconomics is that these short-run effects get unwound over time, leaving the long-run effects of monetary policy to fall only on inflation. Once you accept this inconvenient truth, it makes a lot of sense for the Bank of Canada to target inflation since it actually can control it over the long run—but not much sense to target things like real GDP or the level of employment. Such “real” variables owe their long-run changes to factors well beyond monetary policy, like tax and labour-market policies, the rate of technological progress, rates of fertility and labour-force growth, and a country’s openness to international trade.
Arguing that the Bank of Canada should not target full employment is not to say that the bank doesn’t care about the level of output or unemployment. It cares deeply. Indeed, since its view of the economy is based on the idea that infl ation is most stable when real GDP is at its full employment level, the Bank of Canada is constantly engaged in the exercise of estimating this slippery concept, and using its imperfect estimate (together with reams of other economic data) to determine its best decisions regarding short-term interest rates. But the bank knows very well that its policy actions have a tighter long-run connection to infl ation than to any other macroeconomic variable.
Canada’s inflation-targeting system has been a great success over the past 20 years, with inflation lower and more stable than at any time in the past half-century. The House of Commons finance committee is nonetheless to be commended for examining the bank’s mandate and thinking about possible improvements. But the two options they have so far mentioned are easy to dismiss. Canada can do much better.
Chris Ragan is a McGill University economics professor, David Dodge Chair in Monetary Policy at the C. D. Howe Institute.