New Ekonomistas post (in Swedish). Here is an English translation.
The Riksbank has systematically neglected the inflation target by keeping inflation significantly lower than the target of 2 percent. This Ekonomistas post summarizes the real-economy consequences of this that I have covered in several previous posts. Since expectations of inflation in the short and longer run have been close to the target, average inflation has fallen significantly below expected inflation. This causes real consequences in the form of higher unemployment, an unexpected and unwelcome increase in the real value of household debt, and an unwelcome transfer of wealth from households to banks. The Riksbank thereby counteracts not only the work of the Government and the Riksdag to achieve the most important objective of economic policy, full employment, it also counteracts Finansinspektionen’s (the Swedish Financial Supervisory Authority) work to maintain financial stability and consumer protection in financial area.
As I have shown in a previous Ekonomistas post (English translation here), the Riksbank has systematically deviated from the inflation target by keeping average inflation significantly below the target of 2 percent for the annual increase in the Consumer Price Index, the CPI. Inflation measured by real-time data has on average been 1.4 percent since 1995, when the inflation target began to apply. From 1997 and onwards 5-year moving averages have always fallen below 2 percent. At present the 5-year moving average is 1 percent, the lowest level since 2002. (See this picture for details.)
An average since 1995 of 1.4 percent thus implies an average deviation below the target of 0.6 percent. What are the real-economy consequences of this?
- An important fact here is that inflation expectations from 1997 onwards have been well anchored to the target, in spite of average inflation falling below the target. That average inflation has become lower than expectations causes real-economy consequences for unemployment and household debt.
Figure 1 shows 5-year moving averages of inflation expectations according to Prospera and CPI inflation. We see that the moving averages of expectations of inflation in 1, 2 and 5 years ahead are close to and mostly slightly above 2 percent. The moving averages of CPI inflation are in contrast distinctly below 2 percent. Actual inflation has thus on average been distinctly below expected inflation. This is what makes inflation below the target having real effects.
A first consequence of average inflation below the target is that average unemployment has become higher. That inflation expectations have been anchored to the inflation target also when average inflation has fallen below the target implies that the long-run Phillips curve, the long-run relation between inflation and unemployment, is no longer vertical but downward-sloping. As I have shown here it has a slope of about 0.75 (see also this figure). With this slope, 0.6 percentage points lower average inflation implies that average unemployment becomes about 0.6/0.75 = 0.8 percentage points higher than if average inflation had equaled the target. 0.8 percentage points higher unemployment since 1997 is a very high real cost of neglecting the inflation target.
During the last few years, the Riksbank’s monetary policy with the interest-rate increases that began in the summer of 2010 has lead to particularly low inflation and particularly high unemployment. The unemployment rate might have been as much as 1.2 percentage points lower compared with if the policy-rate had been kept at 0.25 percentage points instead of being increased from the summer of 2010 onwards. (See this figure and this paper.)
That the Riksbank’s monetary policy has led to higher unemployment than if inflation had been kept on target is today pretty well known and accepted. That this also means that most estimates of the long-run sustainable unemployment rate has a bias of perhaps 0.75 percentage points does not become common knowledge yet. (About this, see this attachment to the minutes of the monetary policy meeting in July 2012.)
Higher real debt
A second consequence of the Riksbank neglecting the price-stability objective is that the real value of households’ debt has become substantially higher than if the Riksbank had achieved the inflation target. This consequence is pretty obvious, once one has thought it through, but it does not seem to have become widely known and understood yet.
If inflation is below the target, the real value of a given nominal debt over time becomes higher than if inflation had been 2 percent, since the actual price level becomes lower than if inflation had been 2 percent. But the important thing here, and what causes the problem, is that the actual price level becomes lower than the price level that the borrower had anticipated. Expectations in Sweden of inflation in the short and long run have, as shown in figure 1, been very well anchored on the inflation target, even though average inflation has been significantly lower. This means that the actual price level has become lower than the price level the borrower previously anticipated.
A lower price level than anticipated means that the real value of any given nominal debt becomes higher than anticipated. (This is an example of what Irving Fisher called “debt deflation,” see this Vox column). The households’ debt-to-income and loan-to-value ratios then become higher than what the household previously had anticipated. The households’ net worth and net worth-to-total assets ratio become lower than anticipated. The older a given nominal debt, the higher than anticipated its real value and the debt-to-income ratio, and the lower the borrower’s net worth and net worth-to-total assets ratio. The higher real debt also corresponds to an unanticipated wealth transfer from borrowers to lenders, from households to the banks.
- This can be understood as the surprisingly low price level resulting in an unanticipated capital loss for the borrower and a corresponding capital gain for the lender. It is the surprise, the surprisingly low price level, that is the problem.
If inflation expectations had been 1.4 percent and equal to average inflation, the lower inflation than the target would certainly still have led to a slower real amortization of a given loan, but the borrower would have been compensated by nominal mortgage rates having been lower.
- One can also understand it as the real interest rate ex post (the nominal rate less actual inflation) for a given nominal loan being higher than the real rate ex ante (the nominal rate less expected inflation). The real rate ex post is the actual cost of the loan. The real rate ex ante is the expected cost of the loan.
This can be made more concrete in the following way. Assume that the mortgage rate before tax is 6 percent and thus (1 – 0.3) x 6 = 4,2 percent after tax (I assume a deductible tax rate of 30 percent). Assume also that inflation expectations are 2 percent. The real after-tax mortgage rate ex ante is then 4.2 – 2 = 2.2 percent. Assume now that actual inflation becomes 1.4 percent. The real after-tax mortgage rate ex post then becomes 4,2 – 1.4 = 2,8 percent, 0.6 percentage points higher than the ex ante rate. A 0.6 percentage points higher real rate over a long period is a lot.
You can translate this into a corresponding increase in the nominal mortgage rate before tax. In order to have an after-tax real mortgage rate that is 0.6 percentage points higher, you need a 0.6/(1 – 0.3) = 0.86 percentage points higher before-tax nominal mortgage rate. It is as if the borrower would believe that the mortgage rate is 6 percent before tax when it actually is 6.86 percent. If the lender, the bank, would secretly add 0.86 percentage points to the mortgage rate, and secretly add the increased debt service to the original debt, almost everyone would consider it scandalous. In practice, the Riksbank is doing the same by keeping average inflation below the target and the expectations.
The Riksbank frequently says that borrowers shall amortize their loans. But, with som irony, the Riksbank, by keeping the inflation rate at a lower level, actually counteracts the real amortization that an inflation of 2 percent implies (2 percent inflation means that the real debt is halved in 35 years).
- By neglecting the inflation target the Riksbank thus contributes to the real debt becoming larger than the borrowers have anticipated. Over time this sums to considerable amounts.
As I have shown in this Vox column, a loan taken out in the beginning of 2003 now has a real value that is about 9 percent higher than if the Riksbank had kept inflation on target (see this figure). Thus, this borrower now has correspondingly higher debt-to-income and loan-to-value ratios, and corresponding lower net worth and net worth-to-total assets ratio. The borrowers balance sheet is definitely weaker. The bank’s claim on the borrower is also 9 percent larger in real terms.
The Riksbank’s neglect of the price-stability objective, when inflation expectations have been anchored to the target, has led to real-economy consequences for unemployment and household real debt. Average unemployment becomes higher. The last few years, monetary policy may have increased unemployment particularly much. The Riksbank thereby counteracts the Government’s and the Riksdag’s work to achieve the most important objective of economic policy, full employment.
Riksbank monetary policy has also lead to an unanticipated capital loss for indebted households and thereby to an unanticipated and unwelcome increase in the real value of households’ debt and debt-to-income ratio, an unwelcome weakening of households’ balance sheets, and an unwelcome transfer of wealth from indebted households to banks. The Riksbank thereby counteracts Finansinspektionen’s (the Swedish FSA) to maintain financial stability and consumer protection within the financial area.