The Bank of Canada has been very successful in keeping inflation on target. Which is what the Bank of Canada was supposed to do.
But keeping inflation on target has failed to prevent recessions caused by deficient aggregate demand. Which is what keeping inflation on target was supposed to do.
The problem is not the Bank of Canada. The problem is the Bank of Canada's inflation target.
The first graph shows the level of total CPI, compared to a 2% trend line.
1. Over the longer term, while total CPI kept very close to the 2% trend line, core CPI drifted below the 2% trend line. That is exactly what is supposed to happen, if the Bank of Canada was doing its job right. If total CPI is trending higher than core CPI, the Bank of Canada's job is to keep total CPI growing at 2%, and let core CPI grow more slowly than 2%.
2. In the shorter term, in 2008, we notice a sharp rise then fall in total CPI, relative to the trend, while core CPI keeps growing very steadily. That is exactly what is supposed to happen, if the Bank of Canada was doing its job right. It uses core inflation as a short-term "operational guide", rather than the more volatile total inflation.
The Bank of Canada does not have a crystal ball. But over the last few years its lack of a crystal ball didn't seem to make much difference. Even with the benefit of hindsight, we can say that the Bank of Canada made almost no mistakes in doing what it needed to do to keep inflation on target.
The Bank of Canada kept inflation on target, almost exactly as it was supposed to.
But keeping inflation on target wasn't good enough to prevent the 2008 recession.
Price level path targeting wouldn't have made any difference either. Because what the Bank of Canada actually did under inflation targeting was almost exactly the same as if it had been successfully targeting the price level path. Inflation targeting is only different from price level path targeting if the Bank makes repeated mistakes in the same direction, so that those mistakes cumulate under inflation targeting but are corrected under price level path targeting. But the Bank did not make repeated mistakes in the same direction.
You can't see any signature of the 2008 recession in the above two graphs. You can easily see the signature of the 2008 recession in the next graph.
The third graph shows Nominal GDP, compared to a 5% trend. In a counterfactual world, where the Bank of Canada was supposed to be targeting the level path of NGDP to follow that 5% trend line, what would we say? We would say that monetary policy was a little too loose in the years leading up to 2008, and then suddenly became much too tight in 2008, and stayed too tight thereafter.
In that counterfactual world, where the Bank was supposed to be targeting NGDP, but actually allowed NGDP first to rise a little above target, then suddenly fall well below target, we would all be looking at that third graph, and we would all be blaming the Bank of Canada for the 2008 recession.
It is as easy for the Bank of Canada to target NGDP as it is to target CPI inflation. There is no theoretical reason to believe that, if we had told the Bank of Canada to target NGDP at 5%, rather than CPI inflation at 2%, the Bank of Canada could not have kept NGDP as close to the trend line as it actually kept CPI close to the trend line. Because:
1. Monetary policy can ultimately target nominal variables but not real variables. NGDP is a nominal variable, just like CPI.
2. NGDP responds more quickly to monetary policy than does the CPI. We know this because the very reason the Bank uses a 2 year horizon to bring inflation back to target is precisely because it believes that overly aggressive monetary policy to bring CPI inflation back to target more quickly than 2 years would cause excessive short-term volatility in Real GDP. And NGDP is simply the product of Real GDP and the price level.
3. If people knew that the Bank was targeting NGDP, and expected NGDP to return to trend line following a shock, those expectations would act as an automatic stabiliser for NGDP. For example, if NGDP fell below trend, people would expect NGDP to grow faster than 5% for the next few years, and that combination of higher expected inflation and/or higher expected real growth would cause aggregate demand to increase today and NGDP to increase today, even if the Bank did nothing today. Or even if the Bank were unable to do anything today, because nominal interest rates were at the Zero Lower Bound.
The decision 20 years ago to have the Bank of Canada target inflation was a mistake. Inflation targeting was a better policy than what went before, but not as good as the path not taken. We should have adopted NGDP targeting instead. We thought that if the Bank succeeded in keeping inflation on target there would be no recessions due to deficient aggregate demand. We thought that recessions due to deficient aggregate demand would only happen if the Bank made a mistake, and let inflation fall below target, because it lacked a crystal ball. We were wrong. We know that now.
The graphs for CPI, either total or core, show no signature of the 2008 recession. The graph for NGDP shows a very clear signature of the 2008 recession. Targeting NGDP is no harder, and probably easier, than targeting inflation.
"It ain't broke; don't fix it"? The above graphs show that inflation is the dog that didn't bark to warn us of the 2008 recession. Inflation targeting is broke; we need to fix it. NGDP targeting is a better policy.
(Thanks to Stephen Gordon for the graphs.)
(BTW, I'm not able to post or respond to comments as frequently as I used to. That's because I'm temporarily back in administrative harness as associate dean, which is keeping me very busy doing lots and lots of all the little things that associate deans do. Sorry.)