Wickman-Parak: Inflation targeting and the financial crisis
It is usual at the beginning of a new year to look back over the past year. If we look back at economic developments during 2008, it would be an understatement to say that it was an eventful year. First we had soaring oil and other commodity prices, combined with rising cost pressures and ever declining economic activity abroad. At this point concern was expressed in the international debate that the economies could suffer stagflation, that is, a combination of rising inflation and an economic downturn. The anxiety about stagflation soon faded. When the crisis in the financial markets suddenly intensified in the wake of the Lehman Brothers bankruptcy and economic activity deteriorated substantially in most parts of the world, a new theme for the debate arose; deflation.
It is no surprise that situations like the present one cause people to wonder whether the monetary policy system is really strong enough to withstand these conditions, too. Let me first say that it is important to discuss this; new experiences and knowledge change the conditions for monetary policy. These things are not carved in stone! I also think that it is important to point out that the Riksbank is constantly working to improve its strategy and system and in this I include everything from forecasts and reports on which decisions are based to how we communicate our decisions and our strategy. This is a constantly ongoing process. Our present system has served us well during the crisis. It remains to be seen what lessons will be learned once the crisis is over and has been analysed.
At the same time, it is important that any discussion on potential changes is based on a correct description of the system and of how monetary policy is in fact conducted. If anything should need to be changed, the decision must be based on the right grounds. I also believe that one should be cautious with regard to remaking the foundations of our monetary policy regime, that is, the inflation target and how it is formulated – and this applies especially in a sensitive situation like the present one. But this of course raises questions regarding the present system and monetary policy strategy and why they are designed in this way.
Quantified inflation target as nominal anchor
In Sweden we had for a long time a fixed exchange rate as the norm for monetary policy and as the nominal anchor for the economy, that is, a means of anchoring inflation expectations so that they do not drift away. The idea was that a fixed exchange rate for the krona would mean that inflation in Sweden was adapted in the long term to inflation in the countries with currencies to which the krona was pegged. Confidence in the fixed exchange rate would mean that the expectations of inflation in Sweden would largely be the same as the expectations in the countries to which we were thus linked. But in Sweden, as in many other countries, it proved impossible to prevent inflation soaring with a fixed exchange rate regime. Other countries which tried to control inflation by means of targets for the money supply also faced problems.
The Riksbank announced its inflation target in January 1993, after the krona had been allowed to float earlier in the autumn with a great hullabaloo. The Bank then announced that the target for monetary policy would be to hold the annual change in the consumer price index at 2 per cent. The target would formally apply with effect from 1995. In this way Sweden became one of the first countries to choose to formulate its monetary policy in terms of an explicit inflation target.
Despite being aware that it would be difficult to attain the target exactly, the Riksbank chose to formulate the target with a specific figure for inflation. One of the most important reasons for this was to try to create an anchor around which inflation expectations could be stabilised. Clarity and openness were guiding principles for the inflation-targeting regime right from the start, and I believe that this has played an important part in building up confidence in the inflation target. This culture has therefore continued to permeate the Riksbank’s activities. It is also important so that the Riksdag, which is our principal, can evaluate the Bank and require accountability from us. This motive became particularly important when the new wording of the Sveriges Riksbank Act gave the Riksbank an Executive Board with an independent position in 1999.
Why 2 per cent?
The new wording of the Sveriges Riksbank Act applying from 1999 changed the objective for the Riksbank’s activities to maintaining price stability. The previous Act contained a more general formulation that the Riksbank was responsible for “foreign exchange and credit policies”. When the new wording was produced, it was pointed out in the preliminary work that in principle the Act could be very detailed with regard to the target for inflation and so on. But the legislators judged that this type of precision would not be sufficiently flexible and allow the Riksbank enough scope to adjust the target to different circumstances.
This raises the question of what price stability actually is and of whether 2 per cent inflation is the best target for the Riksbank. If one turns to academic literature on this subject, one will not find any clear answers. Everyone agrees that high inflation is detrimental to the economy. But it is possible to find both arguments in favour of stabilising inflation a couple of percentage points above 2 and arguments in favour of a target that is lower, perhaps even 0 per cent.
One argument for having an inflation target above zero is that it reduces the likelihood that the economy will suffer periods of deflation. If the target were to stabilise inflation around 0 per cent, inflation would probably fall below 0 during certain periods. That is, prices would periodically fall instead of increasing. There would then be a greater risk that inflation expectations – which should be around 0 per cent – would slide downwards so that households and companies would begin to expect a more lasting fall in prices. This could, for instance, lead to households and companies delaying their consumption and investment until later, when prices have fallen even further, thus leading to a decline in demand. This could in turn reinforce expectations that prices will fall and it could create a negative spiral that is difficult to reverse. Having an inflation target that is slightly above zero thus gives some protection against this.
So, the inflation target should perhaps be greater than zero, but not too high. A few years ago, professors Francesco Giavazzi and Frederic Mishkin made an evaluation of monetary policy in Sweden during the ten years we have had an inflation target. One of the questions they discussed was the size of the inflation target. As I have already implied, it was not possible to draw any more precise conclusions on an exact figure, but their conclusion was that an inflation target of between 1 and 3 per cent is a good choice. If it is lower than 1 per cent, it can have a negative effect on economic growth. However, if it is above 3 per cent, it is too high for most people to perceive prices as being stable. This is why countries with an inflation target have often chosen a measure of inflation of around 2 per cent, which was of course also the case in Sweden.
Must the inflation target be defined in terms of the CPI?
One question that often arises in the debate on monetary policy is whether it might not be better to replace the target for the CPI with a target for some other measure of inflation. One suggestion that came up during the autumn, for instance, when energy and food prices rose substantially, was that the inflation target should only apply to domestic inflation – that is increases in prices of goods and services produced in Sweden. Energy prices and many food prices are determined on the world market in any case and are outside the reach of the Riksbank. This is true, but the price of, say, rice in our shops is not the same as the world market price of rice. The prices of goods we import are also affected at the consumer stage by conditions in Sweden, for instance, via transport costs, price mark-ups by wholesalers and retailers and also by the value of the krona.
When the Riksbank raises the repo rate, total demand in the economy is dampened. This reduces the scope for price increases in all areas. And it is important that monetary policy focuses on keeping down the general price increases and that the inflation target is not redefined merely because some prices are rising more than others. Swedish households’ total purchasing power declines even when prices of imported products rise. So the risk of such price increases spreading to other prices and to wages will scarcely decline just because the inflation target is defined in terms of domestic prices.
It is also important to distinguish between the measure of inflation to which our target applies and measures that are useful from time to time to analyse the driving forces behind inflation. The Riksbank often highlights the development of other measures to describe developments in inflation and explain the monetary policy strategy in a better way. For example, last autumn one measure that was useful was CPIX excluding energy. This could illustrate that we were assuming that energy prices would push up inflation temporarily but that the inflation trend would not be affected.
But then is the CPI really the best measure for our inflation target? Different countries have chosen different solutions – although most countries define their inflation target in terms of the CPI or a modified version of the CPI. We Executive Board members have said that of course the Riksbank should not exclude the possibility of changing to another measure of inflation than the CPI if there are good reasons for doing so. But when discussing such a change I believe that it is important to remember the most important reason for introducing the inflation target, namely the need for a nominal anchor. This implied at that time, and in my opinion it still indicates, that the inflation target should be defined in terms of a price index that reflects normal purchases and which Swedish households are familiar with. In any case, one important criterion for an alternative measures must be that it can really function as an anchor for inflation expectations.
Are we Executive Board members "inflation nutters”?
Many proposals for changes in the inflation-targeting policy that have been put forward recently relate to the balance between stabilising inflation and stabilising the real economy. Sometimes they seem to assume that the Riksbank is not allowed to, is unable to, or quite simply does not want to give consideration to production and employment when making decisions on the repo rate. Let me say this as clearly as I can: they are wrong. When we make decisions on changing the repo rate and make forecasts for it we always consider how production and employment will develop.
To begin with, it is important to remember that there is generally no contradiction between low and stable inflation and good growth and employment. The whole point of having an inflation target is that it provides the economy with a stability that creates the right conditions for sustainable economic growth.
So it is correct that price stability is the overall objective of the Riksbank’s activities. Nor do we have any specific target for production or employment in the same way that we have a specific target for inflation. Does this mean then that we are by definition, and indeed should be, “inflation nutters” who only care about the inflation target when deciding on the interest rate?(1)
When they drew up the new wording for the 1999 Sveriges Riksbank Act the possibility of the Riksbank having a target for the real side of the economy was also discussed. For various reasons the legislators decided that it would not be appropriate. For example, both earlier experiences and economic theory show that an expansionary monetary policy that fuels inflation will not increase employment or lead to higher growth in the long term. Such a policy has on the contrary had negative consequences for households and companies. So, as monetary policy cannot be used to achieve a lasting higher level of employment or growth in the economy, it would not be appropriate to have specific targets for the real side of the economy. In the long term, monetary policy cannot affect the real economy and the overall objective of monetary policy should therefore be price stability.
However, monetary policy does have an effect on the real side of the economy in the short term, and in this way there is scope for monetary policy to help stabilise fluctuations in, for instance, production and employment. The legislator therefore considered that the Riksbank should have an obligation to support the general economic policy objectives laid down by the Riksdag, but to the extent that these did not conflict with the price stability target. They considered that this was a natural consequence of the Riksbank being an authority under the Riksdag, so this obligation did not need to be written into the act. There is thus no reference to this in the Sveriges Riksbank Act, although it is mentioned in the bill adopted by the Riksdag.
So how do we Executive Board members ensure that inflation will be on target at the same time as giving consideration to the real economy? Well, quite simply by not behaving as "inflation nutters” who concentrate blindly on how far inflation is deviating from the target. Normally, if inflation has deviated from the target we do not necessarily aim to bring inflation back on target as quickly as possible. In this way, we create the scope to conduct monetary policy so that fluctuations in the real economy are dampened. Monetary policy is flexible in this way.
At the same time, we of course cannot allow inflation to drift too far from the 2 per cent target, or to wait too long before bringing inflation back on target, as there would then be a risk that the inflation target would lose credibility. So, from the Riksbank’s point of view we have said that monetary policy is normally aimed at attaining the inflation target a couple of years ahead. This is our aim when forecasting the repo rate path. So an appropriate repo rate forecast is normally that inflation is close to the target a couple of years ahead, according to our forecast, without any major fluctuations in the real economy.
Monetary policy depends on the type of shock
This means that it is not possible to describe our strategy by any mechanical policy rule. To determine whether, and if so, how the repo rate needs to be changed now and in the coming period, it is necessary to analyse the causes behind the deviation of inflation from the target and also how the interest rate, inflation and the real economy interact. What is the driving force in the economy? What type of shock has led to inflation deviating from the target? How has it affected production, employment and so on? And how are they likely to develop in the future? As there is a time lag before the effects of monetary policy become apparent, the forecasts are of central importance.
One does not need to go further back in time than last year to find two entirely different examples of situations. The first example is the substantial increase in oil, electricity and food prices during the first half of the year, which pushed up inflation way above the target. This took place in a situation where cost pressures in the economy were rising. At the same time, the price increases meant that households’ purchasing power declined and there were clearer signs that economic activity was beginning to weaken in Sweden and abroad. In the background there was also the financial market turbulence. It was in this way an example of a situation where monetary policy must balance the development of inflation against the development of the real economy. For one thing, we must ensure that we retain the credibility of the inflation target and bring inflation back on target within a reasonable period of time. But at the same time, we must also ensure that the slowdown in production and employment is not too severe.
Given this balance, a majority of the Executive Board decided at the meetings in July and September that the repo rate needed to be raised. At the same time, what was presented then was by no means an extremely tight monetary policy – despite the fact that inflation was at its highest in over a decade and inflation expectations were clearly above the target level. As usual we looked both at how inflation and the real economy would develop. And the forecast for the repo rate presented was such that inflation was expected to be back on target within two years and GDP growth was expected to fall initially but to then rise to a good long-term level.
We know what happened next; the Lehman Brothers investment bank went bankrupt in mid-September, the financial markets ground to a halt and this changed the game rules. The circumstances for monetary policy changed and we were given an example of a different decision-making situation where both developments in inflation and developments in the real economy called for the same decision. During the autumn we cut the repo rate in three stages by a total of 2.75 percentage points. This meant a departure from the monetary policy we have normally conducted and which is characterised by more gradual changes in the interest rate. Before I talk about our most recent interest rate decision I shall discuss in slightly more detail how the financial crisis has affected monetary policy and how other measures by the Riksbank have been activated.
Monetary policy and the financial crisis
In principle, one can say that monetary policy is working as normal within the framework that has been in place since 1995. And unlike earlier crises, not least that at the beginning of the 1990s, we have in the storm now raging a credible nominal anchor for the economy in the form of an inflation target. I think this is worth pointing out. It is actually thanks to the credible inflation target that the Riksbank has the flexibility to forcefully counteract the negative effects of the financial crisis on production and employment.
Of course, the financial crisis means that the circumstances for conducting monetary policy are abnormal. For example, there have not been many occasions when the economy has ground to a halt in the way that has happened in recent months. This creates challenges for monetary policy.
One such challenge is that historical relationships in the economy have to some extent changed as a result of the crisis. One contributory factor to the repo rate needing to be cut so much during the autumn was that rising risk premiums in the wake of the financial crisis meant that the normal link between the short-term market rates and the repo rate was broken. This was indicated by the reactions to our reduction of the repo rate by a total of 1 percentage point in October. The market rates fell to a much lesser extent. On the other hand, the market rates followed the repo rate downwards in December.
Moreover, the opportunities for households and companies to borrow money have been restricted because access to credit has declined. The difficulties for banks and other lending institutions to find funding became acute when the financial crisis also hit Sweden with full force. They were forced to accept funding at shorter and shorter terms as these were the only markets that continued to function. In this situation the Riksbank intervened by, for instance, offering the banks loans at 3 to 6 months, that is, at much longer maturities than the daily and overnight loans normally offered. We took these measures to improve the functioning of the markets and to make it easier for the banks to continue to provide loans to households and companies. We will continue to offer this form of liquidity support as long as is necessary.
The extraordinary measures taken by the Riksbank thus aim to ease the supply of credit and improve the functioning of the market. Although the measures primarily concern safeguarding financial stability, they will also improve the ability of monetary policy to stabilise inflation and the real economy.
Substantial interest rate cut in December
On 3 December we made the individually largest change in the repo rate since the inflation target officially began to apply in 1995. The assessment of the Executive Board members was that cutting the repo rate from 3.75 to 2 per cent was necessary both to dampen the fall in production and employment and to attain the inflation target of 2 per cent.
The situation in the financial markets both in Sweden and abroad had not improved appreciably, despite central banks and governments around the world implementing forceful measures. The markets for medium and long-term borrowing were functioning poorly, and sometimes not at all, and liquidity premiums were still high.
The financial crisis had severely hampered access to credit and we could see clearly how this was reinforcing the downturn in the development of the economy. GDP had begun to fall in important parts of the world around us, economic indicators were at historically low levels and share prices had fallen drastically in recent months. Energy and food prices had fallen substantially since the peak in the middle of the year and this brought some relief for households. The downturn also contributed to a fall in inflation and in this way there was greater scope to counteract the negative developments by cutting policy rates, both in Sweden and abroad.
Our earlier forecast was that the Swedish economy would be tangibly affected by the financial crisis and that there would be a sharp downturn in international economic activity. In October we had revised down our growth forecasts for Sweden and abroad considerably and in December we revised the forecasts down by as much again; the economic slowdown had hit with even greater force than we had earlier expected. The weaker krona together with the fall in energy and food prices appeared to some extent to be able to counteract the severe slowdown in the economy. According to our forecast in December, Swedish GDP will decline by 0.5 per cent in 2009, but recover during 2010, when growth increases to approximately 2 per cent.
We also revised down the inflation forecast substantially, primarily because of falling energy prices and declining interest expenditure for households. According to our assessment, CPI inflation will on average be just over 1 per cent in 2009. However, inflation will then rise towards the target level at the end of 2010, as economic activity improves and the effects of falling energy prices and interest expenditure no longer affect the inflation figures.
The assessment in December was that resource utilisation would be lower than normal over the coming years, while inflation would undershoot the target for a while. Given this, we Executive Board members were agreed that a substantial and long-lasting cut in the repo rate was needed to avoid an unnecessarily long period of inflation undershooting the target and of weak real economic development. Nor were there any advantages to adjusting the repo rate in gradual stages. On the contrary, a large cut in the repo rate would in the opinion of the Executive Board have a faster effect than a smaller cut together with a downward adjustment in the forecast for the interest rate.
Uncertainty over what is around the corner
Another difficulty that monetary policy faces is that there is particularly great uncertainty over future developments right now. Our decision in December that the repo rate needed to be cut substantially was, as I see it, relatively simple. But at the same time it was very difficult to assess economic developments over the coming years and thereby the future development of the repo rate path.
Some of the difficulty lies in assessing the scope and effects of different fiscal policy stimulation packages abroad. However, the greatest difficulty lies in assessing how long the financial crisis will persist and what demands it will make on monetary policy. My assessment at the most recent meeting was that a cut of 1.75 percentage points and a repo rate around 2 per cent over the coming year was the closest we could come to a well-balanced monetary policy. But I considered the assessment to be very uncertain, particularly from 2010 onwards. Will the expected recovery in the economy manifest itself or will the downturn be more prolonged? We will probably all have to continue to wrestle with this question.
The plan is that the next decision on the repo rate will be made next month. We will then, as usual, take a stance on how the new statistics that have become available affect the prospects for inflation and economic activity and the repo rate forecast. At the same time, we will of course closely follow developments and be prepared to take action if and when necessary.
Some concluding reflections
The events of last autumn clearly illustrate that with a credible monetary policy framework it is possible to conduct a flexible monetary policy that can quickly be adapted to sudden changes in circumstances. The fact that the actual rate of inflation has been far above the target has not been an obstacle to significant and rapid cuts in the interest rate. The forecasts for both growth and inflation have been adjusted downwards considerably and have indicated scope to reduce the interest rate in order to bring inflation back to the target a couple of years ahead. Monetary policy is and must continue to be forward looking.
As I suggested earlier, the theme of deflation has begun to appear in the international debate. I commented on this at our latest monetary policy meeting. My view was that the threat of deflation should not be exaggerated and that there are major differences compared to the 1930s, the period that people often refer to in this context. One important difference is that central banks and governments around the world are taking strong measures to support the financial system and stimulation packages have been launched to counter the economic downturn.
We forecast that the rate of inflation will decline rapidly in 2009 and this will probably mean that people will continue to talk about deflation. If so, it is important to remember why inflation is falling. A large part of the explanation comprises falling energy prices and falling interest expenditure; that is factors that are favourable to growth. And it is also important to remember that a temporary fall in the rate of inflation towards zero, or even below zero, should not be confused with a prolonged process with falling prices that affects long-term inflation expectations. In the latter case, consumption and demand may come to a halt, which could further reduce inflation expectations and thus affect demand in an ongoing negative spiral. Our forecasts, and those of most other experts, indicate that growth will recover in 2010 and that inflation will then increase. Expectations of generally falling prices over a long period of time should not thrive in such an environment.
We expect the financial crisis to peter out during the course of the year that has just begun. However, the analysis of what has happened and of what needs to be done in terms of regulation will continue. The discussion of the consequences for monetary policy and the monetary policy framework will also continue. One question that will probably arise again is whether, and if so how, central banks should take the development of credit and asset values such as property prices into account. This has been a controversial issue in the international debate in recent years. In the light of the events of the autumn, this issue will probably become topical again.
Note:
1) Mervyn King, Governor of the Bank of England, coined the term "inflation nutter" as a description of a central bank that only concerned itself with the development of inflation
