Statistical Implications of Inflation Targeting

5 The Experience of Statistics Sweden

Carol Carson, Claudia Dziobek, and Charles Enoch
Published Date:
September 2002
  • ShareShare
Show Summary Details

In January 1993, the governing board of the Riksbank (the Swedish central bank) decided to adopt an explicit inflation target. The board specified that, from 1995 onward, the rate of inflation should be limited to 2 percent a year with a tolerance interval of plus or minus 1 percentage point. Price stability as a primary objective of monetary policy is established by the Riksbank Act of 1999. The act also granted the Riksbank a high degree of independence from political influence.

The inflation target in Sweden is expressed as the change in the official consumer price index (CPI). The CPI is compiled and disseminated by Statistics Sweden and dates back to 1954. It is published monthly, has a short time lag, and is rarely subject to revision. Statistics Sweden has a Council for the Consumer Price Index that interprets the basic principles for the CPI and also advises on methodology issues. The Riksbank is represented on the council.

Since 1993, the attention paid to the CPI and the statistics that influence the development of CPI has increased notably. The Riksbank, other policymakers, and actors on the financial markets have become highly active users of statistics. They have become more knowledgeable about what methodology is used and how the compilation is done. Statistics Sweden has supported this development of competence among users and has also examined their need for more and improved statistics. A release system that guarantees contemporaneous dissemination of identical information to all interested parties has been developed as a complement to regular dissemination via a website and in press releases.

In a monetary policy regime with an inflation target, the same kind of accurate and timely economic statistics is needed as with any other type of monetary policy regime. In addition, it is important to compile statistics on underlying or core inflation, inflation expectations, and asset prices. So far, monetary policy with an explicit inflation target has been rather successful in Sweden. However, improvements still need to be made in the statistical basis (in particular in terms of coverage and timeliness) for monetary policy. In the European Union (EU), there is an urgent need to harmonize and improve the coverage and timeliness of EU/Economic and Monetary Union (EMU) economic statistics, while preserving accuracy.

Institutional Setting

Since 1995, Sweden has been a member of the EU). However, Sweden is not yet a member of the third stage of EMU. According to documents signed when Sweden became a member of the EU, Sweden is obligated to join the EMU but has some room to decide on the timing of this change. So far no date has been announced for deciding the issue of membership in the EMU.

The Swedish central bank, the Riksbank, is an institution under the Swedish Parliament, not under the Swedish government. Over the past 10 years, the Riksbank has become more and more independent of the government and the Parliament.

Because Sweden is a member of the EU but not the EMU, the Riksbank takes part in the European System of Central Banks (ESCB), but it has no decision-making power when it comes to the monetary policy of the European Central Bank (ECB). The ESCB is responsible for producing statistics relevant to the conduct of European monetary policy. To this end, the Riksbank provides the ECB with relevant statistics for Sweden.

In Sweden, there is a long-standing tradition of a government made up of relatively small ministries and independent authorities. Statistics Sweden is one of the independent authorities. The yearly turnaround is about $80 million, half of which comes from the Parliament and half from commissioned work. The staff totals about 1,300, of which more than half have a university degree.

Compiling price statistics is one of the responsibilities of Statistics Sweden. For this we get about $2.3 million a year from the Parliament. The task includes the CPI, the harmonized index of consumer prices (HICP), producer price indices, and purchasing power parity statistics.

The Instruction for Statistics Sweden, which was drawn up by the government, states that there shall be a Council for the Consumer Price Index. This council consists of experts on price statistics from the Ministry of Finance, the Riksbank, government authorities, and universities. It is charged with deciding on methodological issues related to the CPI. The chairman of the council is the Deputy Director General of Statistics Sweden.

Swedish Inflation Targeting

Since 1992, Sweden has conducted monetary policy within the framework of a floating exchange rate regime. During most of the twentieth century, however, Sweden had a fixed exchange rate, which had occasionally been devalued. After a devaluation in the early 1950s, the Swedish krona had a fixed rate vis-à-vis other compatible currencies until this system broke down in the early 1970s. The Swedish krona then for a few years followed the German deutsche mark, and after that a basket of currencies, until it finally was pegged to the European Currency Unit (ECU). The krona was devalued three times in 1976–77 and twice in 1981–82.

Following a boom in the late 1980s, the Swedish economy in the early 1990s entered the deepest recession since the 1930s. Within a couple of years Sweden lost about 12 percent of its jobs, unemployment rose to levels unthinkable a few years earlier, the government deficit increased to more than 12 percent of GDP, the banking system was on the edge of a total collapse, and the main short-term refinancing rate was raised to 500 percent during a few days in September 1992. In November 1992, the krona was forced off the peg to the ECU. Since then the krona has been floating.

In January 1993, the governing board of the Riksbank decided to adopt an explicit inflation target. The board decided that, from 1995 on, the rate of CPI inflation should be limited to 2 percent a year with a tolerance interval of plus or minus 1 percentage point. The reasons for choosing CPI as the target variable, for specifying the objective as 2 percent, and for allowing a variation around the target will be discussed in the next section.

During the 1990s, the Riksbank became gradually more independent. In 1990, a former minister of finance replaced the undersecretary of state of the Ministry of Finance as the chairman of the governing board. In 1991, when there was a change in the political majority in the Parliament following a general election, the governor of the Riksbank stayed in office, which was not the case in earlier similar situations. In the mid-1990s, preparations and negotiations were under way to establish a new broad-based constitutional base for the Riksbank. In 1997, this work resulted in a bill to the Parliament suggesting a new legal act, which was accepted and, after a general election, confirmed by the Parliament in 1998.

In January 1999, the present Riksbank act came into force. The new act was designed to give the Riksbank independence from political influence, establish a primary objective for monetary policy (price stability) with legal backing, and ensure accountability on the part of the Riksbank for achievement of its policy objective. The new act also gave the Riksbank the task of promoting a safe and efficient payment system. Furthermore, the management structure of the Riksbank was changed. Within the framework of the present exchange rate regime, which is decided by the government, the responsibility for monetary and exchange rate policies was transferred from the governing board, which is appointed by the Parliament, to a new body, an executive board. The governing board now has a supervisory function and appoints the members of the executive board.

Price stability facilitates the role of the payments system, reduces uncertainty in enterprise and household investment decisions, and prevents arbitrary redistribution of income and wealth. Price stability is accordingly viewed as an appropriate objective for monetary policy.

A matter that is discussed in the monetary policy literature is whether the target should be the rate of inflation or the price level (see Svensson, 1999). Both possibilities have benefits and the choice is not self-evident. The Riksbank has chosen to define the objective in terms of the annual rate of change in the price level.

There is generally considered to be a rather long lag before monetary measures affect inflation. That means that policy has to be forward looking. Forecasts of central macroeconomic variables, inflation in particular, therefore play an important role in the work of the Riksbank.

The results of the forecasting rounds at the Riksbank are published quarterly in the Inflation Report, which is released four times a year, normally in March, June, October, and December. The aim of the reports is to present a basis for the decisions on monetary policy that are made by the Riksbank, to disseminate knowledge about the judgments that are made, and to allow interested parties to follow and understand the bank’s monetary policy. Through the reports, the Riksbank also wants to stimulate discussion about monetary policy issues.

Why the Consumer Price Index Is the Target Variable

The Swedish inflation target is expressed as the change in the official CPI, which is compiled and disseminated by Statistics Sweden. The advantages of the CPI are several. It is widely used and recognized as a measure of inflation by economic agents and the general public. The CPI has a long history; it has been available monthly since 1954. It is published monthly with a short time lag (about 13 days after the reference month) and is rarely subject to revision. It is also a determinant of the domestic purchasing power. Consumer goods and services comprise a dominant part of the end use. Consumer prices include price movements from earlier stages and directly influence the welfare of households.

A problem with using the CPI as a target variable for monetary policy is that it contains prices that are outside the control of the Riksbank (indirect taxes and subsidies) and prices that are influenced by monetary policy in a special way (mortgage interest costs). Another problem is that the CPI, like other price indices, includes prices on commodities that may give rise to one-off effects on the price level—for example, petroleum products and coffee. The advantages of the CPI as the target variable have, however, been judged to outweigh these disadvantages. To get around some of the problems, the Riksbank has chosen to disregard the effects of fluctuating factors, such as changes in indirect taxes and interest costs, on inflation when making monetary policy decisions.

As has already been said, the objective of monetary policy in Sweden is to limit the annual increase in the CPI to 2 percent, with a degree of tolerance of plus or minus 1 percentage point. There is no clear evidence that the optimal level for an inflation target is exactly 2 percent. There are reasons for not having too low an inflation target. One reason is that there are indications that in many countries the CPI tends to overestimate inflation by a measurement bias. Statistics Sweden, however, has found that the Swedish CPI may both overestimate and underestimate inflation. Second, a low inflation target may cause problems if nominal wages display downward rigidity. Third, the fact that nominal interest rates are nonnegative is an argument for a positive inflation target.

Because monetary policy cannot control future inflation fully, inflation will have to be allowed to fluctuate around the targeted value. A degree of tolerance can be interpreted as a confidence interval in the statistical sense, implying that inflation may lie outside the interval for a certain percentage of time. In its analyses and explanations of CPI inflation, the Riksbank uses both the consecutive 12-month CPI inflation rates and the moving average of 12-month inflation rates.

Since 1995, both the 12-month figure for CPI inflation and the moving 12-month average for CPI inflation have been outside the target range during three periods. In the second half of the 1990s, the inflation rate was close to zero for a considerable time, largely because falling interest rates and house mortgage costs had a downward effect on CPI. Transitory effects from altered indirect taxes and subsidies pulled in the same direction. During August and September 2001, the CPI inflation was over 3 percent.

At the end of 1997, the Swedish government appointed a special committee to review the Swedish CPI. Its work was reported to the government in November 1999. Among other things, the committee studied the usefulness of the CPI as a target variable for monetary policy. The committee stated that the current CPI has disadvantages as a target variable, foremost as a consequence of the treatment of mortgage interests in the owner-occupied housing item. The committee considered an index that is not directly influenced by changes in the nominal interest rate to be more suitable as a basis for stabilization policy.

The committee discussed several alternatives for developing such an index. The first was a new CPI in which a real mortgage interest is held constant. This proposal has, however, received a great deal of criticism, and there are reasons to consider it unrealistic. The second alternative was to use the European Union-constructed HICP. In the current version of HICP, housing costs in cooperatives and capital costs in owner-occupied houses are not included. There are plans, however, to include them. The advantage of using HICP is that it can be compared directly with price measures in other EU countries and with the target variable for the ECB. A disadvantage is that its future design is still uncertain.

The Riksbank asked the committee to try to specify an indicator of core or underlying inflation that would be broadly accepted and published monthly by Statistics Sweden. The committee said that a simple form of constant-tax index appeared to be the best alternative. The constant-tax index now produced by Statistics Sweden for the Riksbank has been a good trend indicator for the CPI excluding mortgage interest. The committee, however, believed that these types of constant-tax indices should be used as supplements to the official target variable, rather than as replacements. Other examples of measures that also might be used as indicators are indices excluding goods with strong price fluctuations, indices focusing on the impact on inflation from domestic sources, and indicators based on specific economic models for price and production change.

The Riksbank’s Need for and Use of Statistics

Because the official CPI is the target variable for Sweden’s monetary policy, it is of utmost importance for the Riksbank to follow and analyze the development of the CPI. However, neither the Riksbank nor any other user of the CPI—or of other short-term statistics of importance to the financial markets—receives notice of the outcome from Statistics Sweden before the official release.

The CPI is constructed to measure, on a monthly basis, the average development of prices for private consumption as a whole. The prices are those that consumers actually pay. The sample of goods and services measured is updated every year. The weights are also updated on an annual basis. In that way, changes in the structure of private consumption are taken into account. The Riksbank is represented on the Council for the Consumer Price Index and thus is able to offer its opinion on the interpretation of the basic principles for the CPI and to advise Statistics Sweden on methodology issues. The council, however, never participates in and is never asked for its opinions on the production and release of CPI results.

Starting with the publication of the CPI for August 1998, Statistics Sweden has calculated and published two measures of underlying inflation. These measures, called UND1X and UNDINHX, are defined by the Riksbank and published monthly. The work of Statistics Sweden is financed, on a commission basis, by the Riksbank, which asked the agency to make the calculations because it is an independent body and experienced in handling the CPI. In the opinion of the Riksbank, despite its use in formulating the target for inflation, the CPI as a target variable is problematic. The Riksbank has therefore made it clear that there may be reasons to base operational monetary policy on alternative measures of inflation.

UND1X and UNDINHX exclude temporary effects that enter into the CPI but are normally irrelevant or unimportant for monetary policy. In both measures of underlying inflation, the CPI domain is modified to exclude mortgage interest on owner-occupied houses. In UNDINHX the domain is further reduced to exclude goods that are mainly imported. Both indices are adjusted for changes in indirect taxes and subsidies. Thus alterations in indirect taxes and subsidies, except those related to wages and salaries, are excluded from price movements.

Results of analyses of the development of the CPI and the measures of underlying inflation, as well as their effects on the stance of monetary policy, are made public in the Inflation Reports. In these, the Riksbank presents an assessment of the most likely development of the most important determinative factors for inflation in coming years.

Economic development in other countries, especially in the United States and in the euro area countries, is a critical element in anticipating inflation in Sweden. Price developments, the situation and development of labor markets, production, and consumer and company confidence and expectations are variables that are studied and used to determine the state of international economic development, which in turn will influence the Swedish economy during coming years.

In the Inflation Reports, the condition of the Swedish economy is analyzed under headings such as monetary conditions, inflation expectations and import prices, demand and supply conditions in the Swedish economy, deregulation, political decisions, and temporary effects.

In these parts of the reports, statistics on the following are presented and commented on: interest rates, exchange rates, monetary aggregates, Swedish import prices, crude oil and petrol prices, retail trade and household consumption, quarterly national accounts aggregates, government finance, production and order inflow in manufacturing industry, foreign trade, investments, construction and construction prices, rent, asset prices, employment and productivity, wages and salaries, unit labor cost, capacity utilization, and so forth.

The experience of Statistics Sweden is that the Riksbank has become an even more well-informed and active user of statistics since the decision was made to base monetary policy on an inflation target. In different ways, not just as a member of the Council for the Consumer Price Index, the Riksbank has built up a thorough knowledge of methods and calculations that go into formulating the CPI.

In an indication of the Riksbank’s serious and active interest in high-quality statistics, at the end of 2000 the Bank decided to commission Statistics Sweden to conduct two studies: a feasibility study on the possibility of shortening the production time for quarterly national accounts, and an analysis of differences between Sweden and the United States concerning possibilities for measuring the development of productivity.

The Riksbank also decided to financially support ongoing work at Statistics Sweden in three areas: improving the quality of the producer price index system (including export price, import price, and home market price index series) by intensifying the completion of the sample of goods and companies; optimizing samples in different ways; and examining possibilities for introducing a probability sample. Statistics Sweden will also report and analyze methods used to introduce new goods, replace goods, and take account of quality changes. Statistics Sweden is very satisfied with this cooperation.

Statistics Sweden has also been commissioned by the Riksbank to survey actors on the financial markets, representatives of employer and employee organizations, and chief buyers (procurement officers) of large Swedish enterprises concerning their views on expected inflation rates and expected rates of annual wage increases during the coming one to five years. Actors on financial markets are also asked to state their opinions on the Riksbank’s repurchase interest rate in 3, 12, and 24 months and the expected level of the Swedish krona. The surveys are conducted four times a year, in time for the Riksbank’s work on the Inflation Reports.

In September 2000, a Commission on the Review of Economic Statistics was appointed by the government. The chairman of the commission is the Director General of Statistics Sweden. The commission’s task is to analyze Swedish economic statistics in an international perspective and to propose how economic statistics can be adapted to meet expanding and new needs. As its initial step, the commission reviewed user views on present Swedish economic statistics and on possible improvements.

The Riksbank being a user of special importance, the commission asked the Riksbank to report in writing on its need for improved and new economic statistics. In its letter to the commission, the Riksbank states that monetary policy depends on economic statistics of high quality. The quality of measurements of historical price developments is of central significance, as the Riksbank’s aim is to maintain price stability and its decisions on monetary policy depend on the ability to estimate the future development of prices. High-quality statistics in several areas are therefore required.

According to the Riksbank, the production time for quarterly national accounts should be shortened. An observed bias in the first estimates of annual GDP growth should be analyzed, and deviations between supply and use in the calculations should be reduced. There is an urgent need for deflators for the services industries, especially for the telecom industry. The reliability of index series within the producer price index system should be strengthened. The information technology sector should be systematically covered by statistics. Statistics Sweden is urged to calculate historic time series for strategic variables and make them comparable; several changes in classifications and methodology make comparisons over time difficult. The Riksbank also points to the need for improvement in inventory and investments statistics, for more comparison between the results of different statistical products (such as employment), and for more and better presentations of and comments on statistical results.

Other Users

Since the deregulation of the credit market in the middle of the 1980s, financial markets have developed substantially in Sweden. A system of financial intermediaries and auxiliaries has been established and commercial banks have set up organizational units with traders of financial instruments. The influence of economic policy on interest rates, stock prices, exchange rates, and so on has forced actors on the financial markets to analyze and predict policy actions by, among other things, following economic statistics. Thus, a new and important user group for economic statistics has been established in Sweden.

With the independence of the Riksbank and the linking of monetary policy to inflation and CPI, actors on financial markets have become even more eager to watch, analyze, and even predict outcomes of price statistics. The interest of the actors on the financial markets is directed to the same statistical variables that the Riksbank analyzes and forecasts. Thus, they now show a great interest in methodology and the quality of relevant statistics.

Furthermore, because the way statistical results are released is of great importance, representatives of the actors on the financial markets took the initiative to establish the computer system that is still used—in a modernized form—to disseminate statistical results to thousands of users at once via intermediaries. Since its establishment, the “data shooting” system has been financed, on a commission basis, by some news agencies acting as intermediaries.

The knowledge of how statistical calculations are done is used to improve analyses and forecasts, especially for the CPI. Some of the actors have invested substantial resources in preparing the ground for good forecasts. For example, a commercial bank has created a network of retail traders who are questioned every month about price developments for selected goods and services. This information, together with collected information on prices on other goods and services, forms a calculation system, the purpose of which is to estimate a CPI figure a few days before Statistics Sweden releases the official CPI figures.

Production and Dissemination of Relevant Statistics

The Swedish CPI is based on the concept of standard of living, which states that an index should monitor the change in the cost of maintaining a fixed standard. Compensation is declared to be the main purpose of the CPI. The CPI is thus used to determine adjustments in pensions, tax rates, transfers within the private sector, prices in long-term contracts, and interest on index-linked bonds.

The CPI is also used for transforming nominal value changes into volume changes, such as deflating retail turnover and private consumption in the national accounts. The CPI is not, however, designed with any of these particular purposes in mind. Rather, it is intended to be a general measure of change in the cost of living.1

The Swedish CPI is a chain index with annual links going from December of one year to the following December, which are multiplied to calculate the long index series. At present, 1980 is the reference year of the index. For each new link, weights are recalculated based on new information.

There are basically two different modes of price collection for the Swedish CPI. For most services and some goods, staff members at the central office of Statistics Sweden collect the prices, either by telephone or by a small-scale mail survey using a shuttle form, and enter them directly into the computer. This procedure is usually referred to as the central price collection. Data on rents are collected by a large-scale mail survey.

For most goods and some services there is a local price collection. In these cases, product specifications are established centrally. All in all, there are some 200 such representative products in the CPI. Prices for the representative products are collected by price collectors who visit the outlets directly. The outlets are visited on a random day in the week in which the fifteenth of the month occurs. Some 20,000 prices in 700 outlets are observed in this way.

With respect to sampling, the CPI exhibits a mixed picture. Probability sampling is a preferred method and the predominant technique is order probability proportional to size. In some areas other methods are used, with cutoff sampling and quota sampling being the most common.

A special problem is how to measure the price development of housing. In Sweden, the owner-occupied housing sector consists only of single-family houses so a user-cost approach has been adopted. In the present version of this approach, several components are included, such as mortgage interest, depreciation, repairs, insurance, water and sewage, and property tax. Starting in 2001, the total weight for owner-occupied housing has been based on direct estimates of different components of user costs from a survey of owner-occupiers.

A study of bias components and their influence on the total CPI bias was made in 1999. The total bias was found to be very close to zero for the long-term index, and was estimated to be 0.13 percentage points for the short-term index, which determines the published 12-month inflation rate. The estimates refer to averages over several years. For a single year they might be markedly different. The uncertainty of the assessment itself, which is mainly due to conceptual problems concerning owner-occupied housing, quality, and new products, should be emphasized.

The Swedish CPI is published in a press release and on the website of Statistics Sweden. The date and time of day are decided and published beforehand. At the exact moment of the release, the results are sent out by a computer system specially made for a couple of news intermediaries, which in turn disseminate the results to a large number of customers. The press release contains the results of the calculations for the CPI, the HICP, the net price index, and indices on underlying inflation.

The results are also made public in Statistical Reports, which are available in print and on the website. They contain the same information as the press releases, supplemented by tables on index numbers and weights. The results for the different index numbers are also contained in the databases of Statistics Sweden, with free public access through the website.

As was indicated in previous sections, a great many statistics are used to analyze the economic situation and to make the forecasts on which decisions about monetary policy are based. Statistics Sweden produces a large portion of these statistics. The most important statistics are quarterly national accounts, quarterly financial accounts, statistics on prices, foreign trade statistics, short-term statistics on manufacturing industry and services, employment and unemployment statistics (labor force surveys), and statistics on wages and salaries.

Aside from national and financial accounts, which are based on primary statistics, most short-term statistics are sample surveys, with data collected directly from enterprises or individuals. These statistics are mainly produced on a monthly basis. They are released according to the same procedures as were described for the CPI. The results of the CPI as the target variable for monetary policy and the statistics that influence price development constitute critical information for the actors interested in the actions of the Riksbank and other economic policymakers. Therefore, the results have to be released and disseminated to all interested parties at exactly the same time. As was mentioned earlier, nobody outside the producers at Statistics Sweden gets the results before the point-of-time of release.

A European Perspective

Short-term economic statistics are released much later in Europe than in the United States. This is of great concern to the European Commission, the ECB, and the Economic and Financial Affairs Council (ECOFIN). In June 2000, the ECOFIN Council invited the European Commission (Eurostat), in close cooperation with the ECB, to establish an EMU Action Plan identifying areas where urgent progress should be made. This plan is now being gradually implemented.

In September 2000, the Statistical Program Committee (SPC), at the initiative of Statistics Sweden, set up a task force, chaired by Svante Öberg from Sweden and co-chaired by Photis Nanoupolos from Eurostat, to conduct an EU-U.S. and an intra-EU benchmark study to analyze the differences and their causes and to present proposals for action. The task force presented its findings in September 2001 in a report to the SPC, which endorsed the report and its recommendations, with some elaborations.

By tradition and in accordance with the treaty establishing the European Community, the EU statistical system is very different from the U.S. statistical system. While the U.S. system is geographically centralized at the federal level and thematically decentralized, the EU statistical system is geographically decentralized, although in each country it is more or less centralized around the national statistical office and the central bank. National authorities produce national statistics, while at the European level, Eurostat and ECB aggregate national figures to produce EU/EMU statistics. This is the case also for price statistics.

One of the conclusions reached by the task force is that the production of short-term economic statistics, including price statistics, cannot be centralized, as it is in the United States. One might think that, since the ECB is conducting monetary policy with the objective of keeping inflation low in the EMU area, price statistics for the EMU area as a whole would be enough. However, the ECB is not the only institution concerned with price stability. Fiscal policies, collective wage bargaining, competition policies, and so forth also affect prices. Therefore, we need price statistics at the national level as well. As a result, the task force tried to find ways to speed up short-term economic statistics that could satisfy both national and European needs.

Concluding Remarks

Accurate and timely economic statistics are a necessity for all kinds of economic policies. Most important are quarterly national accounts and short-term economic indicators on demand, employment, prices, and costs. In Europe, Eurostat and national statistical agencies are in the process of harmonizing and speeding up the production of these statistics. Today they are less timely and coherent than in the United States. This is a big problem for the ECB, which is in charge of monetary policy for most of the EU.

Regarding monetary policy with an inflation target, a few additional requirements must be met. First, it is necessary to collect statistics on underlying or core inflation, where the effects on the CPI (or some other price variable on which monetary policy is focusing) of such things as changes in indirect taxes and prices on gas and other commodities with volatile prices are deleted. Otherwise, monetary policy might be misled by price changes that it should not react to.

Second, statistics on inflation expectations are needed. These are difficult to measure. People in general do not have a very precise picture of price changes over the past 12 months or what to expect in the coming 12 months. Certain groups, however, such as price setters in companies, labor market partners, and actors in financial markets, have a more precise opinion on these matters. Inflation expectations from prices of different types of assets on capital markets can also be deducted. In particular, the spread of inflation expectations is an interesting variable for monetary policy.

Third, statistics on asset prices are a necessity. Prices on financial assets such as shares and bonds are available in real time on financial markets. Prices on real estate, owner-occupied houses and dwellings, and so forth are not available in an orderly fashion in Sweden or in many other countries or even the EU. Furthermore, to analyze the link between asset prices and consumer demand or reactions by the business sector, asset prices would need to be combined with households’ and companies’ balance sheets.


The methods and calculations of the Swedish CPI are exhaustively described in The Swedish Consumer Price Index: A Handbook of Methods (Statistics Sweden, 2001).

    Other Resources Citing This Publication