Lars Svensson has argued that central banks should "target the forecast", which means setting the policy instruments at a position where the policy goal is equal to the policy forecast. He specifically referred to targeting the central bank's own internal forecast.
In contrast, I've argued that central banks should target the market forecast, that is, set policy at a position where the market expects success. The recent experience of the Bank of Japan provides a good illustration as to why it's better to target the market forecast:
Japan's headline inflation slowed in March from February, highlighting the central bank's struggle to hit its 2 percent price growth target after five years of heavy stimulus, keeping it under pressure to maintain an ultra-easy monetary policy. . . .
"We expect inflation excluding fresh food to average 1 percent in the current fiscal year, below the median forecast among BOJ board members of 1.4 percent," said Marcel Thieliant, senior Japan economist at Capital Economics.
"If we are right, the bank will have to reduce its inflation forecasts yet again over the coming months, underlining that policy tightening is a long way off."
Sources familiar with the central bank's thinking said it is likely to maintain the view that inflation will reach its 2 percent target next fiscal year, and will stay near that level the following year, when it issues new forecasts next week.
Those projections could increase the prospects of the BOJ deciding to debate whittling down its massive stimulus next fiscal year.
The Bank of Japan is under a lot of pressure to forecast an outcome that is relatively close to their target, especially for one or two years in the future (when policy is able to affect inflation.) The reason is simple; inflation targeting is the BOJ's primary job, and it looks bad if they were to continually forecast failure.
But this is precisely why I disagree with Svensson's approach to targeting the forecast. Because of these political pressures, we cannot rely on the BOJ to provide honest, unbiased forecasts of inflation. Almost all other forecasters predict that the BOJ will continue to undershoot its 2% inflation target for the foreseeable future. The BOJ has consistently been overly optimistic in its inflation forecast (much more so than the Fed), and I believe the current forecast will also prove to be too optimistic. Policy should be steered by market participants that are under no pressure to produce happy face forecasts.