Search:

Keyword:

Navigation

Content

1997 CPI Revision Advisory Committee

Background to the Household Economic Survey
Calculation and Structure of the Consumers Price Index
Glossary
Housing and Mortgage Interest in the Consumers Price Index
How the Consumers Price Index fits into the Statistical System
Managing Measurement Limitations in the Consumer Price Index
Methods of Outlet Selection and Weighting in the Consumer Price Index
Obtaining Representative Prices in the Consumer Price Index
Quality Change and Assessment in the Consumers Price Index
Regional Coverage in the Consumer Prices Price Index
Report of the Consumer Price Index Revision Advisory Committee
Services in the Consumer Price Index
The Treatment of Seasonal Commodities in the Consumers Price Index
What Should the Consumer Price Index Measure?

Background to the Household Economic Survey

Executive Summary

The Household Economic Survey (HES) is an annual survey primarily designed to collect itemised household expenditure for calculating the expenditure weights of regimen items in the Consumers Price Index (CPI). This paper outlines the use of the HES in the CPI, outlines other uses of the HES data and provides general background information about the survey.

Attached is the Msword 6 file of Background to the Household Economic Survey

If you are unable to open this document refer to technical help.
(59Kb)

contents

Calculation and Structure of the Consumers Price Index

Executive Summary

A Consumers Price Index (CPI) has been produced in New Zealand since early this century, initially with emphasis on its use as a tool for wage rate adjustment but more recently for a wider variety of uses.

There have been three distinct phases in the historical development of the CPI. The most recent significant change to the conceptual base of the CPI occurred in the third phase in 1974, when there was a change from a consumption approach to an acquisition based approach.

A variety of index formulae can be used to produce a price index. The most commonly used formula is the Laspeyres index, which measures the change in the cost of purchasing the same basket of goods and services in the current period as was purchased in a specified base period. The New Zealand CPI uses a price relative variant of the basic Laspeyres calculation.

New Zealand's CPI has a wide geographic coverage compared to most other countries. The target population is private New Zealand resident households living in permanent dwellings in the two main islands. Expenditure data are mainly derived from the annual Household Economic Survey, while price data are mainly collected in selected urban areas.

The CPI is released quarterly, with data published at a national level and for selected urban areas.

This paper, Calculation and Structure of the Consumers Price Index, details the historical development of the CPI, and the calculation and publishing of the data released. Issues relating to the structure and frequency of the CPI are highlighted in Section 6. The 1998 revision follows the five-yearly schedule of reviews currently in place, and provides an opportunity to address these issues.

Issues and Recommendations for the Revision Advisory Committee

Issues

1. Should the calculation of the New Zealand Consumers Price Index continue to be based on the Laspeyres price index formula?
2. Are five-yearly major revisions adequate? At what level should changes to regimen weights be introduced? Should annual reviews continue to be used to reweight more frequently below the published regimen level?
3. How much change in the population distribution, outlet mix and commodity purchases can be accepted within the life of any particular Consumers Price Index, before revision is needed?
4. Is the current quarterly release of the Consumers Price Index adequate or should Statistics New Zealand move towards releasing either a monthly Consumers Price Index or a monthly indicator series?
5. Should the regimen continue to be published at the existing level of detail?

Recommendations

1. That the New Zealand Consumers Price Index continue to be based on the Laspeyres price index formula.
2. That the New Zealand Consumers Price Index should be revised every three years with more frequent reweighting below the published regimen level occurring in each annual review.
3. That Statistics New Zealand should continue to publish a quarterly Consumers Price Index and should not move towards publishing a monthly Consumers Price Index or a monthly indicator series.
4. That the Consumers Price Index regimen should be published to section level detail.

Attached is the MSWord 6 file of Calculation and Structure of the Consumers Price Index

If you are unable to open this document refer to technical help.
(404Kb)

contents

Glossary

This is a list of technical definitions used in the Consumers Price Index Revision papers.

Attached is the MS Word 6 file of Glossary

If you are unable to open this document refer to technical help.
(179Kb)

contents

Housing and Mortgage Interest in the Consumers Price Index

Executive Summary

This paper outlines the various approaches to home ownership in consumers price indexes. It describes which approaches are applied by the main OECD countries and the reasons advanced by them for their choice. In turn the paper describes New Zealand's evolution to its present practice. The paper reviews the rationale for the present treatment and discusses whether it is still appropriate or whether and on what grounds any changes should be considered.

Issues

1. Should shelter costs be included in the CPI?
   There is general international acceptance amongst index statisticians that shelter is an integral component of a consumers price index. Most OECD countries do include shelter costs in their indexes.
2. What is the role and uses of the CPI ?
   This has particular relevance with respect to the treatment adopted for home ownership costs. Is the primarily purpose of the index as a measure:
   • of household inflation in terms of price level change of the goods and services that they actually purchase; or
   • to be used as a yardstick in determining adjustments of incomes and benefits to compensate for changes in prices; or
   • to be used in economic analysis such as a deflator of the Private Final Consumption Expenditure component of the National Accounts; or
   • to represent economy wide inflation.
3. The primary role of the CPI should determine the conceptual basis on which the index is compiled. This has particular relevance to the treatment adopted to measure changes in the costs associated with home ownership. Therefore should the index adopt:
   • a consumption approach which relates to what households are estimated to consume.
   The consumption approach excludes house purchases (and therefore any of their investment component) as an explicit regimen item. However there is a choice between a rental equivalence or depreciation based notional measurement for measuring changes in the cost of consuming owner occupied shelter, or
   
   • an expenditure approach which relates only to actual and not to notional transactions.
   The expenditure can be based on either actual outlays or acquisitions. This approach means a decision on whether to include or exclude house purchases explicitly from the index. The decision is dependent on the view held regarding inclusion in the index of items which contain an investment component.
4. Should mortgage interest be included in the CPI?
   All the alternative measures with the exception of the rental equivalence measure include mortgage interest as an explicit item within the index.
5. How should mortgage interest be measured in the index ?
   There are three main approaches: simple revaluation; debt profile; and acquisition.

Recommendations

1. That the treatment of home-ownership in the New Zealand Consumers Price Index be chosen in accordance with generally recognised concepts, methodologies and practices for the construction of consumers price indexes; and be consistent with the main uses of the index and be able to command wide public acceptance.
2. That the acquisitions variant of the expenditure approach be adopted for the treatment of home-ownership if the committee considers that the main purpose of the consumers price index is as a measure of price level changes of the goods and services actually purchased by households.
3. That house purchases be explicitly included in the regimen of the index as these represent an important item of expenditure in terms of the goods and services actually bought by the household sector in the market place.
4. That the acquisition approach continue to be applied in the measurement of the change in the price of mortgage interest and other credit services as this most closely represents the debt servicing experience of private New Zealand households.
5. As an alternative measure of housing costs Statistics New Zealand should develop an experimental "housing index" based on the rental equivalence approach.

Attached is the MS Word 6 file of Housing and Mortgage Interest in the Consumers Price Index

If you are unable to open this document refer to technical help.
(103Kb)

contents

How the Consumers Price Index Fits into the Statistical System

Executive Summary

The Consumers Price Index (CPI) measures the change in the price of goods and services purchased by private households in New Zealand. It is put to a variety of uses, ranging from a general measure of inflation to a measure of the purchasing power of personal income.

The CPI is one of a number of price indexes designed to measure inflation in particular sectors of the economy. With the recent liberalisation of the market and the resulting changes in the economic structure of New Zealand, it continues to be a high-profile statistic of interest to an increasingly diverse audience. Historically the CPI was used primarily as a tool for wage adjustment by the Arbitration Court. (For a detailed history of the CPI in New Zealand, refer to the paper Calculation and Structure of the Consumers Price Index). However, because of the CPI's early availability, the confidence others place in it, and its high quality, it is now often used where, conceptually, another measure may be more appropriate.

This paper provides background information on the position of the CPI in the existing statistical system, and gives a general overview of the CPI and other price measures currently in use. It places these in a conceptual framework to show where the CPI and other indexes fit into a national accounting framework.

Attached is the MS Word 6 file of How the Consumers Price Index Fits into the Statistical System

If you are unable to open this document refer to technical help.
(90Kb)

contents

Managing Measurement Limitations in the Consumers Price Index

Executive Summary

This paper is prepared primarily to provide information and discussion on the reliability of measurement in the Consumers Price Index (CPI). It includes any factors considered to contribute to bias in the CPI, but also covers other elements of measurement reliability.

This paper explores the factors that may limit the reliability of the measurement of price change as represented by the CPI and the steps taken to manage these limitations.

There are three aspects to any statistic. These are:

• the situation that is required to be measured;
• the model designed to measure the situation; and
• the actual measure achieved.

The limitations to measurement reliability fall into two categories:
- those involved with the calculation of the actual measure
- compared to the model; and - those which arise when the model is used to measure a particular situation.

The types of measurement limitation in each of these two categories and the importance of these categories are examined.

The differences in the limitations to measurement reliability for different uses of the CPI are discussed. The effect of the five documented index biases listed below are examined:

• Commodity Substitution Bias
• Outlet Substitution Bias
• New Goods Bias
• Elementary Index Bias
• Quality Adjustment Bias

Other forms of limitations involved in the construction, collection and processing of weight and price information are examined. The effect of these limitations on the reliability of the CPI is discussed.

A summary of the results from overseas studies into measurement limitations is included in an appendix. The appendices contain more technical material and the paper is designed to be read without reference to them.

The conclusion is drawn that the most important source of limitation to address is the proper treatment of changes in quality. Other forms of limitation can be minimised or taken into account when interpreting the CPI. Changes in quality limit the measurement reliability of any price index, regardless of the formula used.

Issues for the Revision Advisory Committee

1. Does Statistics New Zealand provide enough information and assurance on the quality of the Consumers Price Index?
2. Are the steps taken by Statistics New Zealand to maintain the quality of the Consumers Price Index adequate?
   Should the basket of goods be selected and weighted with greater frequency? Should procedures be set up to identify and introduce new goods sooner into the Consumers Price Index basket? Should the annual review be extended to cover more areas? Should the New Zealand Consumers Price Index be annually chain linked (this is not current practice)?
3. Are the quality assessment and control methods used by Statistics New Zealand adequate?
   Are the methods used to identify and assess quality change in the Consumers Price Index sufficient for current needs? Are there any areas where quality changes are being missed? How much emphasis should Statistics New Zealand give to new methods of quality adjustment and on-going assessment of quality adjustment?
4. Should an attempt be made to comprehensively measure the level of accuracy in the New Zealand Consumers Price Index?
   Measures of accuracy range from monitoring of adherence to good practices to applying a comprehensive framework of causes of error on an occasional basis. How far should Statistics New Zealand proceed in measuring the accuracy of the New Zealand Consumers Price Index? What guidance can the committee give to the balance of resources allocated to the production of the Consumers Price Index and the integrity of the Consumers Price Index? Statistics New Zealand has made no attempt to measure the possible bias in the New Zealand Consumers Price Index. Would such a measure be of use in limiting bias in the Consumers Price Index? What priority should such a measure be given versus attempts to limit the bias? How much will this cost and who should fund any such study? Is such a study possible?

Recommendations

1. Statistics New Zealand should publish documentation on Sources and Methods of quality adjustment in the New Zealand Consumers Price Index.
2. Statistics New Zealand should review practices and extend the annual review to reflect changing expenditures patterns between major revisions.
3. Statistics New Zealand develop experimental indexes using alternative quality adjustment methods.
4. Statistics New Zealand should give more emphasis to measuring the level of accuracy in the New Zealand Consumers Price Index, given the precision required for policy setting in the current low inflation environment.

Attached is the MS Word 6 file of Managing Measurement Limitations in the Consumers Price Index

If you are unable to open this document refer to technical help.
(247Kb)

contents

Methods of Outlet Selection and Weighting in the Consumers Price Index

Executive Summary

The New Zealand Consumers Price Index (CPI) aims to measure changes in the prices of goods and services purchased by private households in New Zealand. In order to measure the price changes experienced by this population, its purchasing patterns have to be incorporated into the design of the price survey procedures. The spending pattern of this group determines the appropriate selection of outlets from which prices are collected.

The process used to select representative outlets must take into account the range of commodities sold by an outlet, the market share represented by that outlet and by its outlet type, the variability of price movements for the commodities purchased at that outlet, and the cost-effectiveness and ease of collecting information from that outlet.

Each outlet which is priced within a region is representative of a particular outlet type. Where possible, surveyed outlets are weighted so that any price movements recorded contributes to the CPI in proportion to the share of total expenditure which they represent.

This paper describes how outlets are selected from the reference population, and the methods used to combine prices from different outlets for each commodity.

Issues and Recommendations for the Revision Advisory Committee

1. Should new types of outlets not currently included in pricing specifications be introduced into the Consumers Price Index between major revisions?
   Currently new outlet types are not introduced into the Consumers Price Index between major revisions, and only minor or necessary changes are made to the outlet sample outside five-yearly revisions. The introduction of new outlet types in a more timely manner would facilitate the collection of more accurate price data. It would require careful monitoring of the impact of new outlet types on consumer behaviour.

Recommendations

1. New types of outlets not currently included in pricing specifications should be introduced into the Consumers Price Index between major revisions where accurate price history data and market share information is available.
2. Statistics New Zealand continues to accept the use of purposive sampling but recognises significant benefits in statistical sampling where it can be applied at reasonable cost.

Attached is the MS Word 6 file of Methods of Outlet Selection and Weighting in the Consumers Price Index

If you are unable to open this document refer to technical help. (47Kb)

contents

Obtaining Representative Prices in the Consumers Price Index

Executive Summary

The retail environment in New Zealand today, is both dynamic and competitive offering the consumer a range of goods and services more extensive than in any other period of New Zealand history.

The role of the statistician in calculating the CPI is to measure as accurately as possible, the price movements of this range of goods and services purchased by private New Zealand households. The task and how accurately it is performed depends on a number of variables including, sample frames and size, geographic location, quality control and the representativeness of the items priced.

This paper is one of ten papers written for the CPI Revision Advisory Committee. It is concerned with the pricing strategies and mechanisms currently employed by Statistics New Zealand to obtain accurate weighted average prices and touches on possible future developments for responding to changing consumer and market behaviour. Conceptual issues such as generic pricing and possible future developments particularly with reference to modelling prices and price collection are also covered.

The selection of outlets and outlet weights are not covered in this paper but in Regional Coverage in the CPI.

Discussion Issues for the Revision Advisory Committee

1. Imputation is used for the estimation of prices when acceptable prices are not available. In the Consumers Price Index, the main reason for imputing prices is unavailability e.g. when goods and services are temporarily out of stock. Currently, Statistics New Zealand imputes prices by carrying forward the previous available price for a limited period. However, it is possible to estimate prices directly particularly for seasonal commodities. Changes in methods of imputing prices will depend on the outcome of any studies undertaken and available funding at the time.
2. The practice of calculating national average prices (with no corresponding regional average prices) could be extended to calculating a number of enterprises which price their goods and services regionally or nationally and where the same specials and discounts throughout the country. How national prices are applied by each such enterprise would need to be monitored?
3. Currently Statistics New Zealand prices over 700 distinct items throughout the country. The potential for reducing the number of items that are priced will be investigated.
4. Currently Statistics New Zealand introduces new items into the Consumers Price Index during the five yearly major revision. However, Statistics New Zealand would like to introduce new items into the Consumers Price Index between major revisions.
5. The advantages and disadvantages of purposive and probability sampling have been outlined in the paper. The 1991 Consumers Price Index Revision Advisory Committee recommended that Statistics New Zealand should continue to use purposive sampling for selection of prices in the Consumers Price Index.

Recommendations

1. That Statistics New Zealand should investigate the possibility of introducing new imputation techniques to estimate prices when they are not available.
2. That the direct measurement of national average prices only (with no corresponding regional average prices) be extended, where practical, to other items in the Consumers Price Index regimen.
3. That where possible Statistics New Zealand reduce the number of items priced for the Consumers Price Index without compromising the quality of the price measures in the Consumers Price Index.
4. Statistics New Zealand should, when appropriate, introduce new items into the Consumers Price Index between major revisions.
5. That Statistics New Zealand will continue to use purposive sampling for the selection of items prices in the Consumers Price Index.

Attached is the MS Word 6 file of Obtaining Representative Prices in the Consumers Price Index

If you are unable to open this document refer to technical help.
(146Kb)

contents

Quality Change and Assessment in the Consumers Price Index

Executive Summary

This paper is prepared as a supplement to the paper, Managing Measurement Limitations in the Consumer Price Index, also prepared for the 1997 Revision Advisory Committee. The main purpose of this paper is to provide additional information about quality adjustment and techniques used for controlling quality.

To do this, the paper defines quality change and explores types of quality changes and quality control methods. It investigates some of the problems that may be encountered when making adjustments for quality change. The paper also looks at the quality control methods currently used in Statistics New Zealand and some possible future improvements which could be made if additional resources were to be available.

Attached is the MS Word 6 file of Quality Change and Assessment in the Consumers Price Index

If you are unable to open this document refer to technical help.
(125Kb)

contents

Regional Coverage in the Consumers Price Index

Executive Summary

This paper discusses issues of population and regional coverage in the New Zealand Consumers Price Index (CPI). Population coverage concerns who is represented by the price index or any sub-population index (ie the reference population). Regional coverage deals with the methods used for constructing regional indexes and geographic coverage of the locations which should be included in price surveys.

Throughout its history the reference population of the New Zealand CPI has been more extensive than in the consumers price indexes of Australia, Britain, Canada and the United States. Non-private households are the only segment of the New Zealand-resident population excluded from the reference population.

Although the New Zealand CPI aims to represent a large proportion of the population, CPI pricing surveys have largely been restricted to outlets in main urban areas. Price movements in these outlets have been assumed to represent those experienced by all New Zealanders.

The issue of how widely prices should be surveyed depends on the purpose of the CPI; whether regional indexes are desired; price movement differences between and within regions; the costs and benefits of different levels of index accuracy; and public confidence in the CPI as an index which is representative of their experiences. Similar considerations apply to sub-population indexes.

Issues for Discussion

1. Is the current coverage of the Household Economic Survey adequate for calculating the Consumers Price Index regimen weight?
   
   At present the Consumers Price Index reference population is private New Zealand-resident households (this excludes non-private and non-resident households).
2. Is the current coverage of the fifteen urban areas adequate for the collection of prices for the Consumers Price Index?
3. Should Statistics New Zealand:
• move to calculating the national index in a way which gives wider geographic coverage of prices with the possibility of extended outlet and geographic coverage using probability sampling; and
• as a separate exercise calculate (annually or more frequently) sound regional price indexes that include national prices only where they apply to all New Zealand?
4. Regional population weights are used to combine regional price data in proportions approximating the relative expenditure in each Consumers Price Index region. Should these be updated:
   • five yearly when the final population census results are available; and/or
   • annually when population estimates are available?
5. The Superannuitants Price Index (SPI) is produced as a research index. Further development of the SPI or any other sub-population indexes is unlikely unless specific funding is obtained.

Recommendations

1. That the current coverage of the Household Economic Survey for calculating the Consumers Price Index regimen weights is adequate.
2. That the current coverage of the fifteen urban areas is adequate for the collection of prices for the Consumers Price Index. However, Statistics New Zealand could, given specific funding, investigate the following:
• calculating the national index in a way which gives wider geographic coverage of prices;
• the use of probability sampling in selecting outlets or representative geographic regions; and
• as a separate exercise calculate (annually or more frequently) sound regional price indexes with no national prices except where they apply to all New Zealand.
3. Regional population weights are used to combine regional price data in proportions approximating the relative expenditure in each Consumers Price Index region. Should these be updated:
• five yearly when the final population census results are available; and/or
• annually when population estimates are available.
4. Development of the research index Superannuitants Price Index (SPI) is unlikely unless specific funding is obtained.

Attached is the MS Word 6 file of Regional Coverage in the Consumers Price Index

If you are unable to open this document refer to technical help.
(187Kb)

contents

Report of The Consumers Price Index Revision Advisory Committee 1997 (G28A)

The committee reviewed the use, methodology, calculation, and release to users of the New Zealand Consumers Price Index. The committee, in its deliberations, took into account the wide ranging uses of the CPI and, more particularly, the changing economic environment in which it is used. As a consequence, some fundamental changes in the approach to the Consumers Price Index have been recommended. The committee was mindful of the need to maintain public confidence in the CPI and the recommendations for change are intended to do this. Although there aresome substantive conceptual changes recommended, the majority of the recommendations reconfirm the essential soundness of the current CPI and the way it is calculated.

If you do not have the Adobe Acrobat Reader you may download the reader to view or print the contents of this file.
(125Kb)

contents

Services in the Consumers Price Index

Executive Summary and Issues

This paper is prepared primarily to provide information and discussion on services in the CPI, focusing on current issues in the services area. The opportunity exists for public submissions to be made to Statistics New Zealand who will, in turn, make these available to the Revision Advisory Committee (RAC).

This paper explores the following issues:

• Gambling can be viewed as either a donation, investment or paying for an entertainment. In the light of no empirical evidence to determine the proportion of gambling expenditure that represents entertainment, should Statistics New Zealand include gambling in the Consumers Price Index if reliable weightings and price methodologies can be established.
• Should Statistics New Zealand include interest payments on student loans in the Consumers Price Index regimen?
• Should Statistics New Zealand create a separate section in the Consumers Price Index regimen structure to separately distinguish telecommunications services from other household services?
• If a robust price survey methodology can developed should Statistics New Zealand include overseas package tours and accommodation expenses prepaid in New Zealand in the Consumers Price Index?
• Should the credit services group be extended to incorporate bank maintenance charges, bank transaction and other fees and bank overdraft interest charges provided that reliable weighting and pricing methodologies can be developed and implemented?

Recommendations

1. It is recommended that interest payments on student loans be included as a separate regimen item in the Credit Services group of the Consumers Price Index at the next revision in 1998.
2. It is recommended that a separate section be established within the index regime structure to separately distinguish telecommunications services from other household services.
3. It is recommended that overseas package tours and accommodation expenses prepaid in New Zealand be included in the Consumers Price Index provided that a robust price survey methodology can developed.
4. It is recommended that credit services be extended to incorporate bank maintenance charges, bank transaction and other fees and bank overdraft interest charges provided that reliable weighting and pricing methodologies can be developed and implemented.

Attached is the acrobat file of Services in the Consumers Price Index

If you are unable to open this document refer to technical help.
(125Kb)

contents

The Treatment of Seasonal Commodities in the Consumers Price Index

Executive Summary

There are a number of commodities in the New Zealand Consumers Price Index (CPI) which are subject to seasonal fluctuations in their purchases, whether due to supply or demand. Some of these have seasonal variation in their prices which can influence the CPI and affect its value as an inflation indicator. This paper outlines the treatment of seasonal commodities in New Zealand's CPI, and discusses issues on alternative treatments and how these can be used in the production and publication of the CPI.

Particular attention is paid to the treatment of seasonal fresh fruits and vegetables, a known problem area. Statistics New Zealand has given special treatment to these items for some time. Last revision, a new method, the fixed basket with regular reviews of seasonal adjustment factors, was introduced. It is now time to reflect on this change in method.

The treatment of seasonal commodities is an area of price index practice that is always open to debate. There is no defined "best practice" internationally for guidance in treating seasonal commodities in New Zealand. Various practices used in overseas statistical agencies are outlined in the paper.

Issues like revising seasonally adjusted data, the use of trends, and how to publish the resultant indexes arise when the treatment of seasonal commodities is considered. Those issues are discussed here.

Issues for Discussion by the Revision Advisory Committee

1. Should Statistics New Zealand attempt to adjust seasonal commodities in the Consumers Price Index?
2. Should Statistics New Zealand extend the seasonal adjustment to other commodities in the Consumers Price Index that exhibit a seasonal pattern?
3. Should Statistics New Zealand investigate calculating a Consumers Price Index without seasonal treatment for use as an analytical series?

Recommendations

1. The Consumers Price Index should be published with the removal of normal seasonal variations.
2. That Statistics New Zealand continues seasonal adjustment of fresh fruit and vegetable items in the Consumers Price Index.
3. That, in principle, Statistics New Zealand should extend seasonal adjustment to other commodities in the Consumers Price Index that exhibit a seasonal pattern.
4. That Statistics New Zealand investigates calculating a Consumers Price Index without seasonal treatment for use as an analytical series.

Attached is the MS Word 6 file of The Treatment of Seasonal Commodities in the Consumers Price Index

If you are unable to open this document refer to technical help.
(309Kb)

contents

What Should the Consumers Price Index Measure?

Executive Summary

The Consumers Price Index (CPI) is designed to measure inflation as it affects households. It measures changes in the prices of the goods and services which households purchase.

It is an "expenditure" based index not a "consumption" based index. This means that with respect to houses and consumer durables the CPI measures, each quarter, changes in the actual prices paid for these goods, not an estimate of change in the value of the services delivered to their owners.

The CPI is a "base weighted" index. It measures the change in price from quarter to quarter of the "basket" of goods and services purchased by households in a past base period (currently the base period is the December quarter 1993).

The CPI is the most prominent measure of inflation in New Zealand. It is used for a variety of purposes, and it is not always ideal for any one of those uses. The design of the CPI must balance the different demands arising from the different uses of the index. It must also balance what is conceptually desirable with what is practical and efficient.

It is not clear that some of the changes suggested to the current design of the CPI will make it any more useful.

The CPI is used as a general measure of inflation in the New Zealand economy. In this role it helps set inflationary expectations in labour, commodity and finance markets. In this role it is used by the Reserve Bank as the starting point for constructing a measure of inflation suitable for monetary policy and for monitoring the requirements of the Policy Targets Agreement (PTA).

As a general measure of inflation the CPI is limited in that it only relates to the household sector. Inflation affects a much wider range of market transactions, including expenditure by government, business and the private non profit sectors. Likewise, monetary policy impacts more widely than just on the household sector.

The ideal measure of inflation for monetary policy purposes would be one which measures inflation across a wider range of sectors. It would exclude notional measures such as imputed rent. It would be produced monthly rather than quarterly to contain inflationary speculation.

It would be preferable to develop a purpose built index with these features for monetary policy, rather than compromise the other uses of the CPI.

A general inflation measure can not be easily derived from the CPI. There are no obvious changes to the form of the CPI that would make it less than just a rough approximation to the ideal measure.

The CPI is also used to adjust benefits to compensate for the impact of inflation.

The design of the index depends on the compensation objective:

• Is it maintenance of purchasing power?
• Is it to offset an increase in the "cost of living"?
• Is it maintenance of a "standard of living"?

For each purpose the CPI has its limitations.

"Cost of living" maintenance requires maintenance of a constant level of utility or satisfaction from the consumption of goods and services. CPIs are at best only approximations of a true cost of living index. Conceptually, the adoption of "rental equivalence" or "user cost" methods for the treatment of housing and other durables would make the CPI more closely approximate a cost of living index.

The current expenditure based index design takes account of the impact of price change on the purchasing power of each dollar of income, but a more complete measure, such as a real disposable income index which takes account of changes in incomes, tax and benefit elegibility criteria is a more suitable measure of changes in peoples standard of living.

It is the erosion of peoples capacity to buy goods and services, including their capacity to purchase houses, rather than consume them, which is arguably what "the impact of inflation" means to households. An expenditure based index is a better fit for this purpose.

Issues

1. What should the Consumers Price Index be a measure of?
   What the Consumers Price Index measures determines some of the conceptual framework of the Index. An index to measure cost of living would have a different framework to an index to measure general inflation.
2. Should the Consumers Price Index be a broad measure suitable for a range of uses?
   To what extent should the Consumers Price Index try to accommodate the range of uses and situations as a single index? What guidelines, such as international best practices, should Statistics New Zealand follow in determining its own methods?
3. Is the Consumers Price Index the correct measure for the monitoring of monetary policy?
   Is the coverage of the Consumers Price Index, specifically the household sector, wide enough to allow monitoring of the effects of monetary policy? Would wider, general, or economy wide measures of inflation be more appropriate?
4. Does the current suite of indexes produced by Statistics New Zealand cover the dynamics of the New Zealand economy?
   Statistics New Zealand produces a range of indexes covering different areas of the New Zealand economy. The Consumers Price Index, Producers Price Indexes, Labour Cost Indexes, Overseas Trade Indexes, Capital Goods Price Indexes and Implicit Price Deflator are all measures of price change in different areas of the New Zealand economy. Is this range of measures adequate for users? Are there measures that need developing to fill any gap in the coverage of the economy, or gaps about the impact of inflation on households?

Recommendations

1. That the Consumers Price Index remains a price index of goods and services purchased by private New Zealand households. The weights for the index should continue to be calculated using an expenditure approach and the Consumers Price Index should continue to be calculated using an acquisition approach.
2. No single use of the Consumers Price Index should direct its form.
   Statistics New Zealand should continue to give most emphasis to selecting technically valid price index methods.
3. The Consumers Price Index should not have as its stated purpose the measurement of the change in the cost of living.
   The Consumers Price Index is intended to be of sufficient quality to meet the broad range of needs it can be validly used for. For those purposes that are to be given less emphasis in a future Consumers Price Index, new indexes should be established. These should be supporting measures to the Consumers Price Index and released at the same time.
4. That the need for a general or economy wide measure of inflation, to be used in the context of monetary policy requirements, be specifically addressed.
5. The real disposable income indexes be re-instated as a high priority.

Attached is the MS Word 6 file of What Should the Consumers Price Index Measure?

If you are unable to open this document refer to technical help.

(101Kb)

back to top

For enquiries and information requests

This page last modified on: 15 December 2003