English translation of Ekonomistas post.

In a recent speech, Deputy Governor Per Jansson presents old arguments that I have already responded to, in an attempt to defend the Riksbank’s monetary policy and to criticize its by now several critics. One might think that Jansson might use his time to learn from the Riksbank’s mistakes instead of trying to attack its critics. In any case, a close examination of the facts and data shows that Jansson’s defense of the Riksbank fails. 

Has the Riksbank aggressively with a high interest rate, tried to prevent household debt from rising?

Jansson argues that it is wrong to suggest that “the Riksbank has aggressively, with a high interest rate, tried to prevent household debt from rising.”

However, as figure 1 shows, the Riksbank in fact raised the real policy rate in 2010-2011 from minus 2.5 percent to plus 1 percent, a dramatic tightening of 3.5 percentage points (due to both the increase in the policy rate and the fall inflation). Meanwhile, the real interest rate in the euro area, the UK and the US stayed at rates around minus 2 percent or lower. If this is not an aggressive tightening, I do not know what an aggressive tightening is. (See details here.)

Jansson claims, oddly enough, that “The monetary policy conducted after the crisis has for the most part been governed by traditional monetary policy motives, that is, the expected development of economic activity and inflation over the coming years. Giving consideration to the risks associated with the developments in household debt and housing prices has been part of our assessment, but a relatively minor part.”

But the policy-rate hikes in 2010-2011 took place despite the fact that the inflation forecast was below target and the unemployment forecast was much above a long-run sustainable rate. This is defies all logic, if the aim is to meet the inflation target and to stabilize resource utilization, because a rate hike in that situation leads to less target achievement.

Figure 2 shows the inflation and unemployment forecasts of the Riksbank and the Fed in June 2010. It shows that these forecasts were quite similar. In this situation, the Fed chose to keep the policy rate between 0 and 0.25 percent and began to prepare QE2, the second round of quantitative easing. Does Jansson think that the Fed, as the Riksbank, instead should have raised the policy rate rapidly to 2 percent?

That the majority raised the policy rate despite the inflation forecast being too low and the unemployment forecast being too high was no secret. I pointed this out time and again at the policy meetings, as well as in an interview (in Swedish) in the Swedish newspaper Svenska Dagbladet on September 27, with the headline “Disaster path” (“Katastofbana”) referring to the Riksbank’s policy-rate path.

High growth has been mentioned as a reason for the rate hikes in 2010. But high growth was essential for GDP to recover as quickly as possible after the deep recession of 2009 (see this figure) and was no reason to raise the policy rate. At this stage, it was simply wrong to raise the policy rate to curb growth and the recovery.

That the reason for the rate hikes was concerns about household debt emerged only indirectly in the beginning, for instance, with the following cryptic sentence in the press release of July 1, 2010: “Another factor is that household indebtedness has increased significantly in recent years.” And in the minutes of the meeting held June 30, 2010, Governor Ingves said “that an interest rate increase was also a signal to avoid new financial imbalances from building up and that household indebtedness ought not to rise too much. Mr. Ingves pointed out that this was something he had noted on several earlier occasions. A low interest rate for too long could lead to a troublesome situation beyond the forecast horizon as a result of a credit expansion.”

But it was only in October 2012, in an op-ed (in Swedish) in Svenska Dagbladet, that Ingves spoke out and explained that the choice was between high interest rates and high unemployment: “Today’s high unemployment is a problem, but as Governor, I can not just act in the short term. I must also take responsibility for the long-term consequences of today’s monetary policy. And there are risks associated with excessively low interest rate over a long period of time that cannot be ignored.” (My translation form Swedish.)

Jansson also, as Ingves has done before, tries to claim that there were only small differences between the majority and the minority, that I just preferred a marginally lower policy rate and policy-rate path than the majority’s, and that I wanted to raise the policy rate almost as fast as the majority. The above headline on the front page of Svenska Dagbladet hardly gives the impression that it was a matter of marginal differences between me and the majority. But above all, Ingves and Jansson choose to ignore the fact that my policy-rate paths were only the first step towards a better monetary policy, not the only step. Furthermore, they choose to ignore that there were great differences of principle between the majority and minority regarding monetary policy. (Because they were at the meetings, they know of course that this was the case.) In this post, which Jansson not surprisingly chooses to ignore, I explain my reasoning, the policies my principles would lead to if I had had the majority with me, and how instead the majority came to turn monetary policy upside down.

The majority simply wanted to “normalize” interest rates and in practice elevated this to a separate target for monetary policy. But as I argued in the minutes of the meeting in June 2010:

The repo rate and the repo-rate path do not have any value in themselves, claimed Mr Svensson. They are not target variables. “A normal repo rate” is not a target for monetary policy. Mr. Svensson felt that there was no support in either the Sveriges Riksbank Act or the preparatory work for this Act for a high or low repo rate being a reason to allow poorer fulfilment of the inflation target or to accept poorer stability in resource utilisation. Instead, the repo rate should in all situations be set so that inflation and resource utilisation are stabilised as far as possible, regardless of whether this requires a high or a low repo rate and regardless of whether there is a crisis or not.

But the majority chose to try to normalize interest rates. As Karolina Ekholm and I had warned, this indeed led to abnormally low inflation and abnormally high unemployment.

Is my counterfactual analysis “an afterthought without basis in reality”?

Jansson also question my counterfactual analysis, which is done under the assumption that the policy rate had remained at 0.25 percent since the summer of 2010. His argument is that such a policy rate is “a rather odd reconstruction after the event.” The Swedish text of his speech even calls it “an afterthought without basis in reality.” But would it be unrealistic to try to meet the inflation target, which I discuss in this post after Jansson previously put forward similar arguments? (The post also notes that Jansson (and Skingsley) confuse two different kinds of evaluations of monetary policy, namely, ex ante, that is, in real time, and ex post, that is, with hindsight. Both types of evaluations have their place. The above discussion of interest rate hikes in light of the forecasts in June 2010 is an ex ante evaluation, while my counterfactual analysis is an ex post evaluation.)

Moreover, a policy rate of 0.25 percent or even slightly lower is exactly what the Fed has set since the summer of 2010, with similar inflation and unemployment forecasts as the Riksbank in June 2010. The Fed’s monetary policy has contributed to a significantly better outcome for inflation and unemployment and a somewhat better outcome for GDP than in Sweden. Does Jansson believe that Fed’s low policy rate is also an afterthought without basis in reality?

As is well known, I dissented against every single rate hike (and Karolina against every hike except one). We warned that, with the hikes, inflation would be too low and unemployment too high. We also warned that the Riksbank’s inflation forecasts systematically exaggerated the inflation risks, partly because the Riksbank’s forecast for interest rates abroad was unrealistically high. This is no reconstruction and afterthought. (On the Riksbank’s biased inflation forecasts, see also the interview with Karolina and the post (in Swedish) by Harry Flam.) Had the majority listened to us, there would probably not have been any policy-rate hikes, and the policy rate would have most likely remained at 0.25 percent.

Regardless of this, it is undoubtedly of interest that, according to the Riksbank’s own model, a policy rate at 0.25 percent since the summer of 2010 (the probable outcome if the majority had listened to Karolina and me) would have resulted in significantly better outcomes, with inflation significantly closer to the target and unemployment around 1.2 percentage points lower (equivalent to approximately 60,000 fewer unemployed people). This is discussed in an updated analysis here.

It is also of interest that the debt-to-income ratio would hardly have been much different. Housing prices and debt would have increased somewhat, but disposable income would probably have increased more, so the debt-to-income ratio might very well have been a bit lower. It is also of great interest that such a policy rate apparently would not have resulted in too high inflation or an overheated labor market. If anything, an even lower policy rate would have led to an even better outcome, for instance a policy rate at zero (which I dissented in favor of from July 2009 to February 2010).

Regardless of Karolina’s and my efforts, it must be seen as a major failure that the Riksbank’s monetary policy did not result in the policy rate being kept at 0.25 per cent instead of being raised.

What are the costs of the Riksbank’s monetary policy?

Jansson questions that the Riksbank’s monetary policy would have had a very negative effect on the Swedish economy, especially since the Riksbank “has not during this period hit the repo-rate brake, but rather eased up on the accelerator somewhat.” However, that the Riksbank hit the repo-rate brake interest is quite clear, when one sees in Figure 1 how the real policy rate went from minus 2.5 percent to plus 1 percent during 2010-2011.

Jansson further claims that “the fact that inflation has been surprisingly low is by the way not a specific Swedish problem, but very much an international phenomenon.” But it isi clear from this figure (from this post) that inflation began to fall in 2010 in Sweden while it stayed up in the euro area, the UK and the US.

Jansson compares the outcome in Sweden with that in some other countries, but the correct comparison is of course with what the outcome would have been with a more expansionary policy. Jansson unfortunately presents no own calculations, which would have been interesting. But it is certainly of interest that with a policy rate at 0.25 percent, according to the Riksbank’s own model, inflation would have been significantly closer to the target (table 1 of this post), the unemployment rate would have been about 1.2 percentage points lower (the above table and figures 4 and 5 in this post (in Swedish)), the employment rate would also have been about 1.2 percent higher, employment as a share of the population would have been nearly 1 percentage point higher, and GDP would have been about 2.4 percent higher (figure 1 in the same post).

To these costs can be added that the price level has been approximately constant during the last three years and has ended up about 6 percent lower than what households have expected. This in turn means that the real value of the debt households had in 2011 has now become 6 percentage points larger than households have expected and planned for. This is a substantial increase in households’ real debt burden (see the section “The effect of inflation below expectations” in this paper).

Overall, it is simply difficult to imagine a more counter-productive policy than the Riksbank’s, when with in Jansson’s words “the evaluation is based on a correct description of what has actually happened.”