New Ekonomistas blog (in Swedish). Here is an English translation:
The Riksbank has conducted a monetary policy that has led to far too low inflation, far too high unemployment, and to a somewhat higher (not lower) debt ratio compared to if the policy rate had been left at 0.25 percent from the summer of 2010 until now. This is not a good result.
Inflation in Sweden is since a year far below the target, and unemployment is far above any reasonable estimate of a long-run sustainable rate. Target achievement for monetary policy is thus bad.
An obvious question in the light of the bad target achievement is whether monetary policy could have been conducted in a different way, such that target achievement would have been better. How would target achievement have been with a more expansionary policy during the last few years? This question can be answered with a so-called counterfactual analysis with the Riksbank’s model Ramses.
From the policy meeting in June 2010, the majority of the Executive Board raised the policy rate steadily at each policy meeting, from the 0.25 percent level up to June 2010 to 2 percent in July 2011, an increase by 1.75 percentage points. In an article, I have shown that these increases began in spite of the June 2010 forecast for inflation being below target and the forecast for unemployment being above any reasonable estimate of a long-run sustainable rate (se figures 1 and 2 in the article, or figures 1 and 2 in the minutes from the policy meeting in February 2013). From December 2011 the majority of the Board reduced the policy rate to 1 percent in December 2012. The policy rate has since been kept at 1 percent.
What would have happened if the policy rate instead had been kept at 0.25 percent up to now. The inflation had of course been higher, and unemployment would have been lower. But by how much? This is shown in this four-panel figure (click here for a bigger figure).
The upper left panel shows the actual (red line) and counterfactual (blue line) outcome for the policy rate in percent. The right panels show the corresponding actual (red) and counterfactual (blue) outcomes for CPIF inflation and unemployment in percent. (CPIF inflation is CPI inflation exclusive of the direct effect on housing costs of changes in mortgage rates.) We see that for the low policy-rate path, CPIF-inflation would have stayed close to the 2-percent inflation target, instead of falling to 1 percent and lower. Target achievement for inflation would have been as good as possible. At the same time, unemployment would have fallen and been about 1.2 percentage points lower, around 7 percent. About 60 000 fewer people would have been unemployed. Target achievement for unemployment would have been much better.
These calculations are of course uncertain, but they are not more uncertain than other monetary-policy analysis. They clearly indicate the order magnitude we are talking about, and how much better the situation would have been now if the policy rate had not started to be raised in the summer of 2010.
The conclusion from the analysis is thus that the actual monetary policy has led to substantially lower inflation than the target and substantially higher unemployment than a policy that would have kept the policy rate unchanged at 0.25 percent.
The Riksbank has more recently justified the tight policy by maintaining that a lower policy rate would have increased the household debt ratio (debt relative to disposable income) and would have increased any risks connected with the debt. But, as I have shown in a previous Ekonomistas blog (translated into English here) and in a paper, this is not true. A lower policy rate would have led to a lower debt ratio, not a higher one. This is because a lower policy rate increases the denominator (nominal disposable income) faster than the numerator (nominal debt). Then the debt ratio falls.
In the lower left panel I show what would approximately have happened with the debt ratio. Instead of 174 percent of disposable income, it would have been 171 percent. (See the paper about my lessons from six years in the Riksbank for details.) If you believe that any risks vary positively with the debt ratio, any risks would thus have been lower. However, 171 compared to 174 is too small a difference to materially affect any risks. But it is definitely not a matter of a substantial increase in any risks, as the majority of the Board seems to have believed.
In summary, the Riksbank has conducted a monetary policy that has led to far too low inflation, far too high unemployment, and a somewhat higher debt ratio compared to if the policy rate had been left at 0.25 percent from the summer of 2010 until now. This is not a good result.
By the way, the latest report The Swedish Economy by the National Institute of Economic Research includes a very interesting special study, ”The Riksbank has systematically overestimated inflation,” which may be important in this context. In an analysis of the Riksbank’s inflation forecasts, the NIER shows that Riksbank forecasts have systematically overestimated inflation. The NIER concludes that “[t]he Riksbank’s overestimation of inflation has contributed to overly tight monetary policy with higher unemployment and lower inflation than would have been the case if, on average, its inflation forecasts had been on the mark.”
Why the majority of the Executive Board so systematically has exaggerated inflation risks so systematically is a question that may be worth returning to.