Statistical Implications of Inflation Targeting

7 The U.K. Office for National Statistics and the Inflation Target

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

THIS CHAPTER discusses the responsibilities of the U.K. Office for National Statistics (ONS) with respect to the country’s inflation target. At its simplest level the relationship is straightforward: ONS is responsible for measuring inflation and a wide range of other economic statistics. ONS has no further input into the Bank of England’s inflation projections, over and above supplying statistics on what has actually happened, and has no role in defining the inflation target. The publication of ONS’s data also allows the public to judge the success of monetary policy.

In essence, the current processes for monetary policy in the United Kingdom might be depicted as a tripartite relationship between Her Majesty’s Treasury, the Bank of England, and ONS. The Treasury sets the inflation target, the Bank of England is responsible for meeting the target, and the Bank uses statistics, predominantly from the ONS, to inform the policy decision.

The first part of this chapter outlines these institutional relationships in more detail. One feature of the U.K. institutional model is the closeness of these working relationships: they are codified in a range of documents and are supported by regular, constructive meetings.

The second part of the chapter addresses ONS’s core business of data production directly with a discussion of the construction of ONS’s inflation statistics and the program of research on the methodology of the retail prices index (RPI).

Monetary Policy in the United Kingdom: The Institutional Arrangements

This section starts with an overview of monetary policy in the United Kingdom, illustrating how the focus on economic statistics has developed with changes in the monetary policy framework. This overview leads naturally into a discussion of the statutory basis of ONS, its relationship with the Treasury, and the monetary policy remit set for the Bank of England by the Chancellor of the Exchequer. The section finishes with a description of the relationship between ONS and the Bank of England.

Monetary Policy and the Link with Economic Statistics

The Bank of England has had responsibility for setting interest rates only since 1997; before that the Chancellor of the Exchequer was responsible for setting interest rates. There has always been a close relationship between the monetary policy framework and the demands made on economic statisticians, and the demands have increased with recent developments.

The U.K. money supply targets in the 1970s and early 1980s could be measured using data just from the Bank of England, although the policymakers also had the input of a range of macroeconomic statistics. With the switch to an inflation target in 1992,1 the focus on government statisticians increased, as the chief providers of macroeconomic data. Then in 1994 the move was made to start publishing the minutes of the regular meeting between the Chancellor of the Exchequer and the governor of the Bank of England to discuss the chancellor’s interest rate decision. This publication immediately increased the transparency of the process, focusing attention on the wide range of economic data used to make the decision.

Since 1993, the Bank of England has published an overview of inflationary pressures in its quarterly Inflation Report, providing useful background for economic commentators and observers. It includes consideration of monetary and financial conditions, demand and output, the labor market, and costs and prices.

The Inflation Report came into its own in May 1997, when the Treasury passed responsibility for setting interest rates to the Bank of England, and the Chancellor of the Exchequer granted the Bank operational independence. The remit from the chancellor to the governor, as chair of the Monetary Policy Committee (MPC), provides additional details, including those on the structure of accountability; it was sent in June 1997.

Since then the MPC of the Bank of England has used the Inflation Report2 to present its inflation projections and to share its thinking on monetary policy. The Inflation Report is a substantial resource, comprising some 60 pages of closely argued discussion of recent economic developments. As this report provides far more detail than before on the rationale for policy decisions, it has provided another step change in the degree of transparency of the process and has led to increased public scrutiny of ONS’s economic statistics and their coherence.

The Responsibilities of ONS

ONS was created in 1996 from the merger of the Central Statistical Office and the Office of Population Censuses and Surveys. ONS is a government department and is also an agency reporting to the Chancellor of the Exchequer.

ONS produces most official economic statistics, including the following:

  • the full range of price statistics, as outlined below;

  • national accounts and trade statistics;

  • short-term output indicators, notably the monthly indices of production and distribution output and the retail sales inquiry; and

  • labor market statistics.

In June 2000, a new framework was launched—National Statistics—following the government’s pledge in its 1997 election manifesto to provide independent national statistics. The scope of National Statistics is wide, including all statistics published by ONS, as well as nominated statistics from other government departments.

Responsibility for National Statistics is shared by three central players:

  • the national statistician, who is both the professional head of National Statistics and the director of ONS;

  • the Chancellor of the Exchequer, who is the minister for National Statistics; and

  • the Statistics Commission, which is independent of ministers and producers of National Statistics.

The framework document for National Statistics stipulates that the national statistician has “responsibility for the professional statistical quality of all the outputs comprising National Statistics,” inside and outside ONS.3 He is “accountable to the Chancellor for the performance of National Statistics and, with Heads of Profession [in Departments], for the discharge of the annual work plans approved by Ministers.”

The Statistics Commission’s role is to advise on quality assurance and priority setting for National Statistics and on the procedures designed to deliver statistical integrity. It produces an annual report4 that is laid before Parliament by the minister for National Statistics. Wide consultation with users is part of the process of setting priorities.

The framework document for National Statistics specifies slightly different responsibilities for the RPI than for other statistics. For all National Statistics apart from the RPI, the national statistician is responsible for “developing and maintaining statistical standards, definitions and classifications and promoting high quality statistical output through systematic evaluation and research.” However, for the RPI, “the National Statistician will take the lead in advising on methodological questions concerning the RPI but the scope and definition of the index will continue to be matters for the Chancellor of the Exchequer.”

HM Treasury: The Institutional Relationships

The aim of Her Majesty’s Treasury is “to raise the rate of sustainable growth, and achieve rising prosperity, through creating economic and employment opportunities for all.” The first of the nine objectives to achieve this goal is “maintaining a stable macroeconomic framework with low inflation.”5

The relationship between ONS and HM Treasury has two aspects:

  • the Treasury’s interest in economic statistics, deriving from its responsibility for macroeconomic policy; and

  • the fact that the Chancellor of the Exchequer is the minister for National Statistics, and, linked to this, ONS is one of the chancellor’s departments, along with other departments such as the Inland Revenue and Customs and Excise.

The working relationship between the Treasury and ONS is governed by the 1996 Concordat between the Office for National Statistics and HM Treasury, which outlines the aims and objectives of each, specifies the consultation arrangements between the two organizations, and provides for service-level agreements (SLAs). The key SLA relating to macroeconomic policy is the one on macroeconomic statistics,6 which is very similar to the SLA with the Bank of England. The SLA and related consultation arrangements are described below, in the section on the relationship between ONS and the Bank of England.

As it does with other smaller departments, HM Treasury also has an agreement with ONS under the public spending regime, specifying performance targets for ONS.7 The agreement states that “the Chancellor of the Exchequer will determine the policy and financial framework within which the ONS operates,” and that “the operational management of the ONS is delegated to the Director of ONS.”

When the Labour government gained power in 1997, one of its first acts was to grant operational independence to the Bank of England and establish an MPC with responsibility for setting monetary policy to achieve an inflation target. The government and Parliament remained responsible for setting and defining the objective of monetary policy. These arrangements were enshrined in law through the Bank of England Act (1998), which legislated procedure for MPC meetings, minutes, and appointments.8 The exact definition of price stability, and thus the nature of the inflation target, is communicated to the Bank through the remit, which is set by HM Treasury. The remit has not changed since its introduction in 1997.

The remit given to the MPC on interest rate policy is as follows. The target rate of inflation is 2.5 percent for the 12-month increase in the retail prices index excluding mortgage interest payments (RPIX). The MPC is charged with meeting this target at all times. It is a symmetric target. If inflation deviates by more than 1 percentage point above or below the target, the governor of the Bank of England must send an open letter to the Chancellor of the Exchequer explaining the following:

  • why inflation has deviated from target;

  • what action the Bank of England’s MPC intends to take to get it back to target;

  • how long it will be before inflation returns to target; and

  • how this meets the MPC’s remit as set out by the chancellor.

The 2001 Budget Report9 explains that this open-letter requirement helps to support the government’s wider economic policy objective of high and stable levels of growth and employment.

The monetary policy remit given to the Bank of England by the Treasury complements the Memorandum of Understanding between the Treasury, the Bank of England, and the Financial Services Authority in the field of financial stability.10 In addition, the Bank of England Act covers both of these Bank of England functions.

ONS and the Bank of England: The Wider Working Relationship

ONS and the Bank of England have developed a far closer relationship in recent years. Until the mid-1990s, the main working relationship was between statisticians in the two institutions—the Central Statistical Office (CSO; before the creation of ONS) and the Bank of England’s statistical division—although the Bank’s macroeconomists also took a close interest in ONS’s macroeconomic data. The focus of the statisticians was on the financial data the bank provided to the CSO to feed into the CSO’s compilation of macroeconomic statistics.11

When responsibility for setting interest rates was passed to the Bank of England in 1997, the relationship changed immediately. The Bank became one of ONS’s key customers, and ONS’s links with the Bank of England widened to include, in addition to statisticians, bank economists who analyze the economy and brief the MPC.

This relationship was formalized by the SLA between the Bank of England and ONS,12 which was signed in October 1999. It outlines the services ONS will provide to the Bank of England, and the obligations of the Bank of England in return. It was modeled on the SLA between ONS and the Treasury, and the contents are very similar.

The SLA stipulates the form and timeliness of the data supplied to the Bank of England, and requires statistics of “sufficient” quality and scope, consistent time series, and full public documentation of the sources and methods used in constructing key economic statistics. The Bank of England undertakes to inform ONS of its use and interpretation of ONS economic statistics.

The SLA also provides for regular meetings between ONS and the Bank of England, which has proved to be a good way to develop strong working relationships and a good sense of each side’s objectives as well as their constraints. When formulating spending priorities for economic statistics, ONS places significant weight on the views of the Bank of England.

ONS provides both the Bank of England and HM Treasury with regular updates on progress on the key development programs, and representatives of the Bank of England and the Treasury attend meetings of several of the project boards monitoring methodological change to provide feedback on user needs. The Bank of England and the Treasury both take a keen interest in the RPI research program, and tripartite meetings are held once a quarter to feed their views into the ONS staff’s “technical board.” In support, ONS has increased the number of economists on its staff in recent years; one of their roles is to help ONS understand the perspective of economist users of ONS statistics.

One feature of ONS’s close working relationship with HM Treasury, and more recently with the Bank of England, has been a strong focus on the need for timely economic data. The recent European “benchmarking” study comparing the timeliness of economic data in Europe with that in the United States found that U.K. short-term economic statistics were rather more timely than those produced by any other member state. The United Kingdom was the fastest country to publish 6 of the 12 indicators reviewed; for 10 of the 12 indicators, it was one of the three fastest countries. This timeliness seems likely to result from the very close working relationship between ONS and its key macroeconomic users, HM Treasury and the Bank of England.

Inflation Measurement in the United Kingdom

This section complements the discussion of institutional responsibilities with a brief overview of ONS’s work on prices measurement, focusing on the RPI. It starts by outlining the wide range of price statistics produced by ONS. This is followed by a discussion of the uses of the RPI over time, and then a summary of progress on the RPI development program.

ONS Price Statistics

ONS publishes an unusually wide range of price statistics. Those with the highest profile are the family of domestic retail price indices produced each month:

  • the “headline” RPI, which represents retail prices for the whole population, apart from the richest 4 percent and those pensioner households mainly dependent on state pensions;

  • the RPIX, which—as its name suggests—excludes mortgage interest payments from the RPI;

  • the retail prices index excluding mortgage interest payments and indirect taxes; and

  • detailed product breakdowns of the RPI.

Current press releases and further information can be found at

ONS also produces the harmonized index of consumer prices (HICP). This measure is designed for comparison between European countries. It is calculated differently from the RPI; the geometric mean rather than the RPI’s arithmetic mean is used to aggregate prices at the most basic level. A number of RPI series are also excluded from the HICP, most particularly those relating to owner-occupiers’ housing costs (for example, mortgage interest payments, house depreciation, council tax, and buildings insurance). In the United Kingdom, HICP inflation has consistently been lower than RPIX inflation. HICP inflation in July 2002 was 0.85 percentage points lower than RPIX inflation; of this, 0.4 percentage points were due to the formula effect, although this effect can vary over time. ONS is concerned about the size of this formula effect, and this is a major element of the research program described below.

The ONS complements these retail price series with a number of monthly producer price series, notably the following:

  • trade prices, broken down by continent and commodity type; the export prices are directly collected by ONS from U.K. manufacturers, and the import prices come from world market prices where the products are traded on a world basis (for example, some raw agricultural materials), some proxies, and directly collected prices otherwise (more information at;

  • producer input and output prices covering U.K. production for domestic markets (; and

  • experimental corporate services prices, which price a range of services used as inputs to other goods or services (

A range of deflators is currently produced as part of ONS’s comprehensive national accounts dataset. These build on the directly collected price indices, incorporating the effects of balancing adjustments where necessary. The ONS also produces the experimental final expenditure price index, which is a direct measure of economy-wide inflation. It covers government and investment expenditure as well as consumers’ expenditure. It was requested by the Treasury and the Bank of England, who use it as an additional measure of economy-wide inflation, alongside the implied deflators. Further information can be found at

ONS is not at present using a “stage of processing” framework to develop a complete set of price indices at the various stages of production. However, we are developing a consistent and detailed set of deflators as part of the major program to develop constant price input-output tables; doing so will provide a systematic structure and database to develop a stage of processing framework.

Uses of the Retail Prices Index

Of all ONS price statistics, the RPI has the longest history. Although there were occasional official comparisons of the prices of food in the late nineteenth and early twentieth centuries, the government first began a systematic, continuous check on the increase in the cost of living in 1914. This check was pushed as a way to help protect ordinary workers from what was initially expected to be the temporary economic circumstances of the First World War. The figures initially released related only to food prices, but after June 1916 they were expanded and calculated retrospectively to cover clothing, fuel, and some other items reflecting the principal expenses of a working-class family.

As noted above, the use of an inflation target for monetary policy is relatively recent. The RPI has a range of other functions, in particular the following:

  • many benefits are linked to the RPI, notably pensions;

  • the RPI is used to set the return on index-linked gilts, issued since 1981; and

  • the RPI is used in private sector contracts to specify benchmark price changes.

Because of these wider uses, it is ONS policy not to revise the RPI. With respect to index-linked gilts, the terms of the prospectus13 state that if, in the opinion of the Bank of England, the coverage or the basic calculation of the index is changed in a way that is “fundamental” and “materially detrimental” to the interests of holders of the particular index-linked stock, then HM Treasury is obliged to offer the stockholders the right to redeem their stock at the uplifted par value.14 The amount of principal due on repayment, and of any interest that has accrued, will be calculated on the basis of the index ratio applicable to the month in which repayment takes place. In practice, the Bank has so far not found any changes that have been fundamental or that would have been materially detrimental to the interests of the relevant stockholders.

The RPI is currently described as a “measure of price change.” A convenient way to understand the nature of the RPI is to envisage a very large shopping basket holding all the different kinds of goods and services bought by a typical household. As the prices of individual items in the basket vary, the total cost of the basket will change. The RPI is a measure of the change in this total cost from month to month.

As the balance of interests has moved, periodic reviews over the years have considered whether the basis for the RPI should be changed, notably to a cost of living index. However, the diversity of uses is a significant constraint on changing the rationale of the RPI, and the net result has been to retain the “measure of price change.”

When significant changes to the RPI have been considered, the Chancellor of the Exchequer has in the past called a Retail Prices Index Advisory Committee (RPIAC) to advise him on whether the changes were appropriate and how they should be handled. RPIACs have been staffed with representatives from the public and private sector, including a Bank of England official as a full member and an HM Treasury official appointed in a personal capacity, based on their relevant expertise. The secretariat was provided by the CSO, the forerunner of ONS.15

The RPI Development Program

Since moving to an inflation target, policymakers—first the Treasury and then also the Bank of England—have placed increasing emphasis on the accurate measurement of retail prices. The Boskin report on bias in the U.S. consumer price index (Boskin and others, 1996) crystallized such concerns, and ONS established a comprehensive program of research on the methodology underlying the RPI. The main areas of investigation are outlined below:16

  • Formula bias. Bias arising when the formula used to weight together prices is inappropriate. ONS is establishing broad principles for when the arithmetic mean should be used and when the geometric mean is more appropriate. ONS has also been investigating the impact on the formula effect of moving the month in which the RPI is rebased, such as from January to December, and the practical implications of any such move. The arithmetic differences observed need not be interpreted as bias, as the appropriate formula depends, among other things, on the definition and use of an index.

  • Quality bias. Bias occurring when insufficient account is taken of changes in the quality of goods. ONS is investigating a range of techniques for improving the measurement of quality change, including hedonics.

  • Substitution bias. This has two elements. The first is “product substitution bias,” which arises because there are lags before the price index adjusts for changes in purchasing patterns. In the United Kingdom, the RPI is rebased every year, so this bias will be relatively small compared to countries that rebase less frequently. The second is “outlet substitution bias,” which arises from lags in taking account of changes in the locations where people do their shopping. ONS has carried out a systematic rebalancing of the RPI sample, and outlets are reenumerated every five years on a rolling program. ONS is also currently investigating practical ways of measuring Internet shopping, and is also reviewing its general procedures for monitoring shopping trends to ensure that these are reflected in the RPI on a timely basis.

  • New goods bias. Bias occurring as a result of delays in including the prices of new goods in the index. ONS has estimated the effect of excluding new goods from the RPI in recent years, and it is negligible. The project is also tracking prices of products from the time they first enter the market, to check whether price falls have a more significant effect early in a product’s life.


The changes to the monetary policy framework in recent years provide far more information than previously on the rationale for interest rate decisions, which has naturally led to a higher public profile for economic statistics. In parallel, the three main parties in the monetary policy framework—HM Treasury, the Bank of England, and ONS—have developed clearer written agreements documenting their institutional relationships and responsibilities. This too has increased the transparency of the process. This chapter provides a comprehensive overview of all these agreements.

These agreements confirm the allocation of responsibilities between the three parties: ONS measures what has happened; the Bank of England uses these data, among others, to assess inflationary pressures and thus set interest rates to meet the inflation target; and HM Treasury sets this monetary policy remit. ONS consults the Bank of England and Treasury, as key users of its economic data, when setting its work priorities.

In response to the adoption of an inflation target, ONS is implementing an extensive RPI development program, is investigating all the main potential sources of bias, and has also reconsidered the design of the RPI in light of its evolving uses.

A version of this chapter previously appeared in ONS’s Economic Trends (No. 577, December 2001). [Permission by ONS is gratefully acknowledged, License #C02W0001128.] I am very grateful to ONS staff for their help with this paper, including Len Cook, John Kidgell, Dave Fenwick, Prabhat Vaze, and Geoff Tily. The author was Chief Economist at the U.K. Office for National Statistics at the time of the seminar.

Medium-term “monitoring ranges” for monetary variables such as M0 and M4 remained in place until 1997.

The Office for National Statistics Framework Document (1996) is being revised to take account of the Framework for National Statistics (2000). For both, see

For Statistics Commission Annual Report 2000–01, see

HM Treasury Departmental Report (April 2001),

Service Level Agreement between the Office for National Statistics and HM Treasury on Macro-Economic Statistics (1998).

Office for National Statistics Service Delivery Agreement (2000).

Budget 2001, Investing for the Long Term: Building Opportunity and Prosperity for All, Economic and Fiscal Strategy Report and Financial Statement and Budget Report.

This relationship is governed by the Firm Agreement between the Bank of England and the Office for National Statistics to Supply Data for Macro-Economic Statistical Purposes.

Service Level Agreement between the Office for National Statistics and the Bank of England on Economic Statistics (1999).

The terms governing index-linked gilts are not all the same; prospectuses for new issues of index-linked gilts were changed from March 1982 onward to allow for the possibility of switching to a substitute price index, which, as long as it did not result in material detriment, would avoid triggering the early redemption clauses.

The U.K. Debt Management Office, which has responsibility for gilts issuance, has recently launched a consultation on the redesign of the next tranche of gilts, including a review of the wording of the early redemption clause.

Further details on the construction of the RPI and its history are available in Baxter (1998).

For reports on this program, see Baxter (1997) and Baxter and Camus (1999).

    Other Resources Citing This Publication