1SOURCES AND METHODSCONSTRUCTION PRICE INDICES

Statistics Directorate, Organisation for Economic Co-operation and Development, Paris

Statistical Office of the European Community, Luxembourg

2FOREWORD

This publication is the result of joint work between the Statistics Directorate of the Organisation for Economic Co-operation and Development (OECD) and the Statistical Office of the European Communities (EUROSTAT). Theobjectives of this collaboration were to outline conceptual problems in the development of construction price indices,based on a comparative description of the methodologies currently used by Member countries of the OECD and theEuropean Union.

Such information will enable national statistical institutes, or other organisations compiling construction priceindices, to compare their methodology and data sources with those currently used in other countries. It will alsoprovide a range of options for countries in the process of creating their own indices, or overhauling existing indices.At present there is very little multi-country documentation available on construction price indices.

The level of activity in the construction sector is one of the key determinants of the level of short-term economicactivity in Member countries. The demand for reliable construction price indices arises from the need to assess realchanges in the output from these activities which cannot be derived solely through reference to regular building andconstruction statistics. These indices have a wide range of applications including deflation of components of nationalaccounts, adjustment of construction contracts and leases, and as a basis for indexation for insurance purposes.

At the moment, there is considerable variation in the concepts underlying the compilation of construction priceindices by Member countries, as well as in the price components and methodologies used. This variation is the resultof the different administrative and legislative environment, and differing physical characteristics, such as geographicsize, population density, climate, etc. in which organisations undertaking construction activity operate in eachcountry. Another factor are the different uses of the construction indices compiled.

The Statistics Directorate of the OECD and EUROSTAT are greatly indebted to the statistical offices of OECD andEuropean Union Member countries for their co-operation. Without this assistance, it would not have been possible toproduce this publication.

This publication is presented on the joint responsibility of the Secretary General of the OECD, and the Director-General of EUROSTAT.

Louis Kincannon Yves FranchetDirector Director-GeneralStatistics Directorate EUROSTATOECD

5TABLE OF CONTENTS

Page

Introduction ................................................................................................................................ 7

Section 1: Basic Concepts......................................................................................................... 9

A. Relationship of producer price indices to construction price indices ............................... 9Terminology used in this publication............................................................................... 9Elements of construction industry prices ......................................................................... 10Construction price indices ............................................................................................... 10

B. Main types of construction price indices .......................................................................... 12Input price indices ........................................................................................................... 12Output price indices ......................................................................................................... 12Sellers price indices ....................................................................................................... 12Relation between the three types of indices .................................................................... 13Types of construction price indices compiled in OECD and European Union Membercountries ........................................................................................................................... 13

C. Typology of methods used by OECD and European Union Member countriesfor compiling construction price indices .......................................................................... 17

Prior breakdown methods ..................................................................................................... 17Standard factorsComponent cost method

Subsequent breakdown methods ...................................................................................... 18Quoted pricesSchedule of pricesMatched modelsBuilding volume or areaHedonic method

D. Outline of processes in developing a construction price index ........................................ 20

E. Uses of construction price indices..................................................................................... 21Use of construction price indices for deflating components of national accounts ........... 22

Section 2: Construction Price Index Compilation Issues ...................................................... 24

A. Important considerations in the compilation of construction price indices ...................... 24Diversity of construction activity ..................................................................................... 24Changes over time ............................................................................................................ 24Selection of appropriate prices ......................................................................................... 24Range of items for inclusion ............................................................................................ 25Ensuring geographic representativeness of the index ...................................................... 26Types of construction firms to include............................................................................. 26

B. Particular problems/issues for transition countries ........................................................... 26

6TABLE OF CONTENTS (cont.)Page

C. Sources of information used to compile construction price indices ................................. 27

D. Major elements of construction price indices ................................................................... 28Construction type/activity coverage ................................................................................. 28Geographic coverage........................................................................................................ 29Items in the index ............................................................................................................. 30Weights............................................................................................................................. 30Basis of prices .................................................................................................................. 31Price collection ................................................................................................................. 32Index review..................................................................................................................... 33Index formulae ................................................................................................................. 34Frequency of compilation/publication.............................................................................. 34

Section 3: Description of Country Indices .............................................................................. 35

A. Explanatory notes ............................................................................................................. 35

B. Input construction price indices .................................................................................... ... 37Australia ........................................................................................................................... 37Austria .............................................................................................................................. 39Belgium ............................................................................................................................ 42Canada.............................................................................................................................. 43Denmark ........................................................................................................................... 46Finland.............................................................................................................................. 49France ............................................................................................................................... 51Greece............................................................................................................................... 55Iceland .............................................................................................................................. 57Ireland............................................................................................................................... 58Italy ............................................................................................................................... 60Japan ............................................................................................................................... 64Mexico.............................................................................................................................. 65New Zealand..................................................................................................................... 67Norway ............................................................................................................................. 69Portugal ............................................................................................................................ 71Spain ............................................................................................................................... 73Sweden ............................................................................................................................. 76Turkey .............................................................................................................................. 78United Kingdom............................................................................................................... 79

C. Output construction price indices ..................................................................................... 81Australia ........................................................................................................................... 81Austria .............................................................................................................................. 84Canada.............................................................................................................................. 88France ............................................................................................................................... 90Germany ........................................................................................................................... 93Greece............................................................................................................................... 97

7TABLE OF CONTENTS (cont.)Page

Luxembourg ..................................................................................................................... 99Netherlands....................................................................................................................... 101New Zealand..................................................................................................................... 104Norway ............................................................................................................................. 106Sweden ............................................................................................................................. 109United Kingdom............................................................................................................... 111United States..................................................................................................................... 116

D. Sellers construction price indices .................................................................................... 118Canada.............................................................................................................................. 118Spain ............................................................................................................................... 120United States..................................................................................................................... 122

Section 4: Country Summary Tables ...................................................................................... 124

Table 1: Main data sources used............................................................................................ 124Table 2: Methods for deriving weights.................................................................................. 128

Section 5: International Classifications Relevant to the Compilation of Construction Price Indices............................................................................................................. 132

A. International Standard Industrial Classification (ISIC), Rev. 3........................................ 132B. Statistical Classification of Economic Activities in the European Community (NACE), Rev. 1 ................................................................................................................ 133C. Provisional Central Product Classification (CPC) ............................................................ 134D. Classification of Types of Construction ........................................................................... 137

Bibliography ............................................................................................................................... 138

Glossary ...................................................................................................................................... 139

6ABBREVIATIONS

CPI Consumer price index

ESA European System of National andRegional Accounts in theEuropean Community

EUROSTAT Statistical Office of the EuropeanCommunities

GDP Gross domestic product

GFCF Gross fixed capital formation

ISIC International Standard IndustrialClassification

NACE Statistical Classification ofEconomic Activities in theEuropean Community

OECD Organisation for Economic Co-operation and Development

PPI Producer price index

SNA System of National Accounts

VAT Value added tax

7INTRODUCTION

This publication provides an outline of the sourcesand methods used by OECD and European UnionMember countries in the compilation of priceindices for construction activity. Its objectives are:to provide information on techniques used on acountry by country basis; a conceptual frameworkfor the compilation of construction price indices;and indications of best practice in the compilationof these indices. Such information will enableOECD and European Union Member countries tocompare their methodology/data sources with thosecurrently used in other countries. It will alsoprovide a range of options for countries in theprocess of creating their own indices, oroverhauling existing indices. These options may beeither adopted in their entirety, or modifiedaccording to their own circumstances.

The term construction covers a wide variety ofactivities, these include the construction ofdwellings, non-residential buildings, and civilengineering works such as roads, bridges, dams,etc. Construction activity also encompasses repair,renovations, rehabilitation and maintenance ofexisting structures, etc.

The diversity of construction activity is the cause ofone of the major problems in the compilation ofconstruction price indices, that of comparability.The items comprising an index and their relativeweights are the result of different norms andstandards that apply in each country. For example,when comparing indices compiled by differentcountries for housing construction, account needsto be taken of differences in climate, culture, levelof affluence, legislative standards, physicalcharacteristics (e.g. geographic size, populationdensity, terrain) etc. These factors even influencewhat is included in a house, and therefore therange of items included in a price index of houseconstruction. There is considerable variationbetween countries in the inclusion/exclusion ofitems such as land, utility connection, fixtures,transport of materials, architects fees, etc. fromtheir indices.

At present there is very little multi-countrydocumentation available on construction priceindices. This publication complements twoprevious publications produced by the OECD

in 1994, Consumer Price Indices - Sources andMethods, and Producer Price Indices - Sources andMethods, and a publication produced byEUROSTAT in 1995, Industrial Trends: NationalMethods.

In February 1996 EUROSTAT hosted a seminar onconstruction price indices. Papers from thisseminar providing detailed descriptions ofmethodologies for the compilation of these indicesfor some European Union countries will bepublished in Methodological Aspects ofConstruction Price Indices in late 1996.

The demand for price indices for constructionactivity arises from the need to assess real changesin the output from these activities (i.e. to create aconstant value series) which cannot be derivedsolely through reference to regular building andconstruction statistics.

Construction price indices are used in guaranteedvalue clauses in rental, leasing, and other contracts;adjustment of sales contracts for buildings underconstruction; and as a basis for indexation forinsurance purposes. They are also used to deflatenational accounts estimates of output ofconstruction activities, and gross fixed capitalformation in residential construction.

In summary, construction price indices are used totrack changes/trends in the cost (or price) ofconstruction. They do not provide information onthe current market value of construction work,earning capacity, or rental values.

This publication comprises three sections.

Section 1 outlines basic concepts and describes therelationship between producer price indices andconstruction price indices, elements of constructionindustry prices, the main types of constructionindices, and a typology of methods used by OECDand European Union Member countries in theircompilation.

Section 2 outlines the major problems involved incompilation of construction price indices, sourcesof information used, and describes some of themajor elements in their compilation. The latter

8includes construction activity coverage, items in theindices, the calculation of weights, basis of prices,the alternate approaches to price collection, andissues relating to area coverage. Examples of bestpractice are also provided.

Section 3 briefly describes each of the three mainconstruction price indices compiled in OECD andEuropean Union Member countries. These are:input price indices, output price indices, andsellers price indices. Where appropriate, moredetailed descriptions are provided for somecountries to illustrate important aspects of a numberof the concepts outlined in Section 1, andcompilation issues described in Section 2.

9SECTION 1. BASIC CONCEPTS

A variety of tools are used to measure pricechanges taking place in an economy. Theseinclude consumer price indices (CPIs),producer price indices (PPIs), price indicesrelating to specific goods and/or services, andGDP deflators.

CPIs are designed to measure changes over time inaverage retail prices of a fixed basket of goods andservices taken as representing the consumptionhabits of households.

PPIs provide measures of average movements ofprices received by the producers of commodities.In principle, PPIs exclude transport costs and

consumption taxes. Producer price indices are nota measure of average price levels, or of the costs ofproduction. Moreover, PPIs do not includecommercial mark-ups. Though the scope of PPIsvaries, they are generally calculated on the basis ofthe total turnover of a definable industry such asmanufacturing, agriculture, or mining.

In broad terms, construction price indices providemeasures of changes in the prices of either theinputs to, or outputs of, construction activity.However, terminology used in the context of priceindices for construction activity varies betweencountries. There is also considerable variation inthe inclusion/exclusion of items such as transportcosts, consumption taxes, fittings, etc.

A. RELATIONSHIP OF PRODUCER PRICE INDICES TO CONSTRUCTION PRICE INDICES

Terminology Used in this Publication

For the purposes of this publication it is necessaryto have a clear understanding of terminology usedin the context of construction price indices. AGlossary is provided in this publication, however anumber of key terms are also described below.These relate to the concept of the constructioncontractor, the client, purchaser, and final owner.

It should be emphasised that depending on theinstitutional framework operating within theconstruction branch of an individual country, oreven for different construction projects within acountry, these three concepts may relate to eitherseparate entities, or the one organisation/individual. For example, a construction contractorbuilding houses may also be the entity that took theinitiative for the construction work. Similarly, theclient may also be the final owner.

Construction contractor A firm which undertakes works as part of aconstruction project by virtue of a contract with aclient.

Client (Maitre douvrage) Is the natural or legal person for whom a structureis constructed, or alternatively the person ororganisation that took the initiative of theconstruction.

Purchaser, final owner Is the person or organisation that pays the finalsellers price. In some instances, this person ororganisation may be the same as the client (referabove).

10

Elements of Construction Industry Prices

From the supply side, the price of the output ofconstruction activity is a function of the followingfactors:

Direct inputs: These include materials, labour,energy, etc. Direct inputs generally vary inproportion to output.

Indirect inputs and overheads: These includedepreciation, administrative expenses, etc.These are generally fixed and do not varydirectly with the volume of output.

Productivity: Refers to the efficiency withwhich inputs are converted into outputs (e.g.through new technical solutions, increasedlabour productivity, or more effectiveorganisation of work).

Profit: Is a residual determined by the salesprice, and combinations of the three precedingitems. Profit varies widely and may benegative.

The output price for a construction project maychange for any one or more of the followingreasons:

widening or narrowing of profit margins due tochanges in market conditions (i.e. irrespectiveof changes in costs);

increases or decreases in the prices of directinputs; and

changes in productivity resulting in changes inthe quantity of direct inputs per unit of output.

On the demand side, the price actually paid by thepurchaser or final owner of the output of theconstruction activity includes a number ofadditional cost elements paid by the purchaser.These include the price of the land, costs ofobtaining planning permission, taxes andconnection fees, insurance, professional fees (legal,architects, engineers), real estate agent fees, landregistry charges, etc.

The requested price index cost elements fordeflating components of the national accounts aredescribed in Section 1E below.

Construction Price Indices

Most of the information used in the compilation ofconstruction price indices are derived from thesupply side of the industry (i.e. from constructionfirms, sub-contractors, materials supply firms, etc.).

However, a unique feature of construction activitythat impacts on the compilation of constructionprice indices is that in most situations thecompleted building or construction is not producedand sold by one construction contractor alone1.Normally, the client (or architect charged by theclient with the responsibility of supervising theconstruction) concludes contracts with a number offirms. Most of these are predominantly part of theconstruction branch, however they may also belongto other branches of the economy (e.g. steelconstruction, manufacture of fixtures, engineering,etc.).

The client (or supervising architect) invitesconstruction contractors (who in turn may invitesub-contractors) to undertake work at a building orconstruction site. The work to be done is referredto as work category. If the offer is accepted thework is performed and supplied to theclient/architect as a product. The work categoriescorrespond to the goods or products observedin other price statistics.

From the perspective of the production performedby a construction contractor, the prices in questionmay be either the prices of the various inputs to theconstruction process paid by the constructioncontractor, or the prices received by theconstruction contractor from the client for theoutput of the construction contractor. The latter areproducer prices and come close to the concept of aPPI (i.e. in the context of the construction industrythe prices received by the producers of workcategories). The construction contractors salesprices of individual work categories from theconstruction sector are in most cases also thepurchase prices of the client.

1 German Federal Statistical Office, Wiesbaden,Studies on Statistics, No. 10 - Comments on theRevision of Statistics on Construction Prices,November 1959.

11

An example of an input index is the index ofbuilding costs compiled in Finland which monitorschanges in the prices of 95 cost items using priceinformation obtained from construction enterprisesand price lists. On the other hand, the Austrianresidential and non-residential building output priceindex records changes in the price of residentialbuildings by monitoring changes in 82representative construction operations involved intheir construction.

There are a number of difficulties in compilingconstruction price indices using practices followedin the compilation of PPIs. These include:

The product building or construction or itscomponents are not produced and sold by theconstruction industry or its various branchesalone.

Sub-contractors selling their output toestablishments and enterprises of theconstruction industry.

The construction branch does not only producenew buildings or construction work. It is alsoinvolved in maintenance and repair work,conversions, extensions, demolition, etc. Inaddition, the turnover of the construction branchcontains turnover from other branches (e.g.transport, equipment hire, etc.). These types ofactivities are less readily observed.

As a result of these factors it is difficult to combineindividual series of price relatives for constructionwork at new buildings in the construction industryor its individual branches, as sufficient weightinginformation is seldom available. This isparticularly the case for classifying turnover onnew construction, repairs and maintenance, etc.

However, unlike PPIs, construction price indicesare not based on the total turnover of aninstitutionally definable sector, but on certainproduct categories. It is therefore difficult toproduce a PPI for the construction industry. Mostof the indices produced by OECD and EuropeanUnion Member countries are indices of a number ofbranches participating in the erection of buildings,or other construction activity.

As mentioned above construction price indices mayalso be compiled from information on the demand

side of the industry. This could entail collectinginformation on the price actually paid by thepurchaser or final owner of the output of theconstruction activity, as in the case of the Canadiannew residential housing price index. Alternatively,it could involve gathering information on a numberof additional cost elements that need to be includedin order to arrive at the actual price paid by thepurchaser or final owner. As mentioned above,these include the price of the land, costs ofobtaining planning permission, taxes andconnection fees, insurance, professional fees, realestate agent fees, land registry charges, etc.

In summary, construction price indices may bedescribed as indices compiled from:

prices paid by the contractor for inputs to theconstruction process; or

the price received for the completed output ofconstruction activity paid by the client; or

the selling price including all of the demandside cost elements paid by the purchaser or finalowner.

12

B. MAIN TYPES OF CONSTRUCTION PRICE INDICES

The methods used to compile construction priceindices vary significantly between OECD andEuropean Union Member countries. IndividualMember countries also use a variety of methods,data sources, etc. for the different constructionprice indices they produce.

Three main types of construction price indices arecompiled in OECD and European Union Membercountries: input price indices, output price indices,and sellers price indices.

Input Price Indices

Input price indices measure changes in the price ofinputs to the construction process by monitoringseparately the cost of each factor. This generallyentails the compilation of a weighted index of thecosts of wages and materials.

Initially, representative object (e.g. a dwelling of aspecific type, size, style, etc.) is taken and thequantity of labour hours and materials needed forits construction calculated. These quantities areperiodically multiplied by the corresponding pricesand the outcome totalled.

Input price indices should not be used to provideinformation on price movements for finishedconstruction work as they generally do not reflectthe whole range of influences that impact on marketprices2. These include changes in productivity,profit, and trade margins of the constructioncontractor, and changes in actual market conditions.

Input price indices only provide a reflection ofchanges in the prices of construction inputs. Theindices produced are production cost rather thanproduction price indices.

As a result, the real trend of building costs maydiffer considerably from the trend compiled solelyon the basis of wages and material costs. An input

2 University of Geneva, Laboratory of AppliedEconomics, Geneva - Indices of ConstructionPrices, A Methodological Inventory, LynnMackenzie, April 1994.

cost index is likely to overstate the price rise ofcompleted construction work as it ignores gains inproductivity reflected in price reductions.

Output Price Indices

Output price indices measure changes in the pricesof what is produced by entities engaged inconstruction activity.

Output price indices cover most of the itemsnormally built into the price paid by purchasers orclients to entities involved in producing thecompleted output of the construction activity.These generally include materials, labour,equipment hire, land preparation costs,bathroom/kitchen fittings, overheads, profits, andtrade margins.

Several different techniques are used to include allthese components. One method involves theinclusion in the index of all (or as many as possibleof) the individual factors involved in theconstruction of a dwelling, non-residential building,etc. These include overheads, profits, trademargins, and any other costs paid by the client orpurchaser to the builder. An alternative methodentails basing the index on the prices of actualfinished constructions. Both methods are describedin detail below in the typology of methods used tocompile construction price indices.

Sellers Price Indices

Sellers price indices measure changes in the pricesof construction output paid by the purchaser orfinal owner of the output of construction activity.

The term sellers price is used to distinguish itfrom the purchasers price as defined in theSystem of National Accounts (SNA). The latter(which is discussed in more detail in Section 1Ebelow) excludes the land component in theownership transfer.

Sellers price indices described in this publicationcover the total sales price of completedconstruction, including not only the cost of labourand materials, but also land, direct and indirect

13

selling expenses, and sellers profits. House priceindices compiled in the United States and Canada,and the dwelling price index compiled in Spain, areconceptually broader in item coverage than almostall input and output indices compiled in otherOECD and European Union Member countries.

The resultant indices compiled in these countriesfor houses include most factors which influencemovements in home prices. These include bothsupply factors such as wage rates, material costs,and productivity, and demand factors such asdemographic changes, incomes, and the availabilityof mortgage finance. These indices are the closestapproximations in item coverage to the actual pricepaid for construction output.

Some of these countries also compile sellers priceindices excluding the land component.

The inclusion in the index of all the cost elementspaid for by the final owner of the construction(particularly the land, finance costs, sellingexpenses) conceptually brings a sellersconstruction price index close to being a consumerprice index.

Relation Between the Three Types ofIndices

The item imposition of the three types ofconstruction price indices is illustrated in thefollowing diagram.

Input Price Index Output Price IndexSellers PriceIndex

Elements Paid by Elements Paid by Elements Paid byContractor Client Final Owner

Materials Materials MaterialsLabour Labour LabourPlant &

Equipment Plant &

Equipment Plant &

EquipmentTransport Transport Transport

Energy Energy EnergyOther Costs Other Costs Other Costs

ContractorsProfit Margins

ContractorsProfit Margins

Productivity ProductivityOverheads Overheads

VATLand

ArchitectsFees

Other CostsClients Profit

Margins

In practice, there is some degree of overlap in theitem composition of the three types of constructionprice indices described above. For example, outputprice indices may include professional fees (such asarchitects, lawyers, engineers) to the extent thatthey are initially paid by the construction contractorand subsequently included in the output price paidby the client. The same principle applies to othercost elements. As will be further discussed inSection 3 the development of reliable constructionprice indices requires a thorough understanding ofall stages of the construction process from thematerials supply through to the sale of thecompleted product to the final owner.

Types of Construction Indices Compiled inOECD and European Union MemberCountries

There is considerable variation in the type andcomposition of construction price indices compiledby OECD and European Union Member countries.Just over half of the more than 60 constructionprice indices described in Section 3 below are inputindices. All except three of the remaining indicesare output indices compiled using either of the

14

methodologies referred to above. The threeexceptions are sellers price indices for housescompiled in the United States and Canada, and thedwelling price index compiled in Spain.

There is a continuum in the range of items includedin input and output construction price indices. Forinput indices some countries such as Belgiuminclude only labour and material inputs. Othercountries, such as Japan also include someinstallation costs (e.g. water/gas/electricity supply,bathroom and kitchen fittings, etc.). At the otherextreme countries such as Finland include a widerange of additional cost items such as transport ofmaterials to the site, equipment hire, site

preparation, conveyancing fees, outside fittings,etc.

There is a similar range of item inclusion/exclusionin output price indices reported by OECD andEuropean Union Member countries, though almostall include estimates for trade margins, overheads,and profits. The main differences between thecountries are in the inclusion/exclusion ofprofessional fees and conveyancing fees.

The types of construction price indices compiled byOECD and European Union Member countries,together with the compilation methodology, andfrequency of compilation are provided in thefollowing table.

Construction Price Indices:Types of Indices Calculated, Methods Used, and Frequency of Compilation

Country Index Title Type of Index Methods Used* Frequency ofCompilation

Australia Civil Engineering Input Standard factors QuarterlyConstruction and renovation of privately builthouses

Output Matched models Quarterly

Construction of other dwellings and otherbuildings

Proxy output Component cost Monthly

Austria Housing and estate building costs index Input Standard factors MonthlyCost index for road construction Input Standard factors MonthlyCost index for bridge construction Input Standard factors MonthlyResidential buildings price indices Output Component cost QuarterlyOther building price indices Output Component cost QuarterlyOutput price index for road construction Output Component cost QuarterlyOutput price index for bridge construction Output Component cost QuarterlyOther civil engineering work Output Component cost Quarterly

Belgium Composite construction price index Input Standard factors AnnuallyCanada Residential/Non-residential building input

costsInput Standard factors Monthly

Electric utility construction price index Input Standard factors Bi-annuallyTelecommunications plant Input Standard factors AnnuallyConstruction union wage rates Input Standard factors MonthlyApartment building construction Output Component cost QuarterlyNon-residential building construction priceindex

Output Component cost Quarterly

New housing price index Sellers Matched models MonthlyDenmark Regulating price index for residential

building constructionInput Standard factors Quarterly

Regulating price indices for civil engineeringworks

Input Standard factors Quarterly

* A description of these methods is provided in Section 1C below

15

Finland Index of building costs Input Standard factors MonthlyBuilding renovation costs index Input Standard factors Monthly

France Building construction sector indices (BTindices)

Input Standard factors Monthly

Price indices for civil engineering sector (TPindices)

Input Standard factors Monthly

Construction price indices for residentialbuildings (ICC)

Output Schedule of prices Quarterly

Germany Conventional construction price indices Output Component cost QuarterlyStandard house price indices Output PPI Bi-annually

Greece Input materials for the construction of newresidential buildings

Input Standard factors Quarterly

Work categories for the construction of newbuildings

Output Component cost Quarterly

Iceland Index for privately built apartments Input Standard factors MonthlyIreland Housebuilding cost index Input Standard factors Monthly

Wholesale prices index of materials Input Standard factors MonthlyCapital goods price index Input Standard factors Monthly

Italy Costs index for a residential building Input Standard factors MonthlyCost index for a building for industrial use Input Standard factors QuarterlyCost index for stretches of road Input Standard factors Quarterly

Japan Construction price index Input Standard factors MonthlyLuxembourg Price index for the construction of residential

and semi residential buildingsOutput Component cost Bi-annually

Mexico Price index for social housing Input Standard factors MonthlyNetherlands Price indices for social rented housing Output Component cost Monthly

Hedonic price index for the low rentresidential building sector

Output Hedonic Monthly

NewZealand

Input index for the construction sector Input Standard factors(largely)

Quarterly

Output index for the construction sector Output Component cost(largely)

Quarterly

Capital goods price index for theconstruction sector

Output Component cost(largely)

Quarterly

Norway Cost indices for residential buildings Input Standard factors MonthlyConstruction cost indices for civilengineering works

Input Standard factors Quarterly

Construction price index for detached houses Output Hedonic QuarterlyPortugal Construction Cost Indices Input Standard factors MonthlySpain Construction cost index Input Standard factors Quarterly

Average prices of dwellings per squaremeters

Sellers Building volume orarea

Quarterly

Sweden Factor price index for residential buildings Input Standard factors MonthlyFactor price indices for repair andmaintenance of multi-dwelling, andagricultural buildings

Input Standard factors Monthly

Output price indices for houses andapartments

Output Hedonic Monthly

Turkey Building construction cost index Input Standard factors Quarterly

16

Country Index Title Type of Index Methods Used Frequency ofCompilation

UnitedKingdom

Construction materials cost index Input Standard factor Monthly

Construction industry price index of averageearnings

Input Standard factors Monthly

Public sector housebuilding price index Output Component cost QuarterlyPublic sector non-housing price index Output Schedule of prices QuarterlyRoad construction tender price index Output Schedule of prices QuarterlyCommercial and industrial building tenderprice index

Output Schedule of prices Monthly

Output price indices Output Derived from above QuarterlyUnitedStates

Price index for highway construction Output Component cost Quarterly

Cost index for large projects Output Component cost QuarterlyPrice index for new one family houses Sellers Hedonic Monthly

17

C. TYPOLOGY OF METHODS USED BY OECD AND EUROPEAN UNION MEMBERCOUNTRIES FOR COMPILING CONSTRUCTION PRICE INDICES

Several different methods are used by OECD andEuropean Union Member countries to compileconstruction price indices. Each method aims atminimising the problems of comparability betweendifferent constructions and comparisons over time.

Most commonly, these entail the selection of one ormore standard (or model) types of construction onthe assumption they are representative of allconstruction projects for that type of constructionactivity (house, apartment, factory office building,etc.). The model constructions may actually exist,or be fictitious. Each model construction is definedin great detail in order to be able to obtain theprices (either actual or notional) of all or some ofthe work entailed in their construction. Prices arethen obtained from a number of sources. Theseinclude surveying construction companies,professional associations, administrative sources,etc.

The use of model types of construction projects ismore difficult for civil engineering, wherediversity, between even the same category ofprojects, makes the selection of standardconstructions more difficult3. Methods used for thistype of construction activity include analysis of thecomposition of expenditure by civil engineeringcontractors on major construction projects,expenditures being classified according to thepurpose of the construction project, e.g. roads,bridges, dams, etc.

Seven construction price index compilationmethods used either currently, or in the recent past,by Member countries are described below. In orderto demonstrate the main principles more effectivelythese descriptions only contain the main elementsof each methodology. In practice, the more than60 construction price indices described in Section 3involve numerous variants and combinations of theseven compilation methodologies described. Thecompilation methods (primarily) used for each of

3 University of Geneva, Laboratory of AppliedEconomics, Geneva - Indices of ConstructionPrices, A Methodological Inventory, LynnMackenzie, April 1994.

the construction price indices calculated in Membercountries are listed in the table in Section 1B above.

Prior Breakdown Methods4

The starting point for prior breakdown methods is alist of carefully specified factors or components,from which the total input or output costs of abuilding or construction project are built up.

Standard Factors: This method is mainly used forthe compilation of input price indices. For anygiven year a representative construction (or smallnumber of projects) is selected and the quantities ofeach factor used to build it (e.g. materials, labour,transport, machinery, etc.) evaluated.

Changes in the costs of construction are determinedby monitoring the cost of each factor. Therepresentative building or construction choseninitially is used only to establish the weights.

This methodology, which yields a construction costindex rather than a construction price index, is usedby several OECD and European Union Membercountries. As mentioned earlier there isconsiderable variation in the range of itemsincluded.

Component cost method: This approach is used forthe compilation of output price indices. It regardsconstruction output as bundles of standardisedhomogeneous components. These componentscorrespond to the supply of standard operations (cf.Glossary). Examples would be: the supply andlaying of so many square metres of roofing tiles;installation of a hot water tank of a given capacity;construction of so many square metres of brick walletc.

Price indices are compiled using the prices of thesehomogenous components. A representative

4 Institut National de la Statistique et des EtudesEconomiques (INSEE), Paris, Bloc-Notes 4/1990- Frances Cost of Construction Index, Vincent leCalonnec, 1990.

18

construction (or number of projects) is also chosen.However, the actual work entailed in its completionis broken down into precisely defined standardservices or components.

A number of representative construction firms thathave recently performed any of these services aresurveyed to determine the price they have actuallyagreed or invoiced for these services. A priceindex is then created for each standard component.

These indices are then aggregated for the buildingsinitially defined as the benchmarks. The actualbuildings are used only to define a selection ofservices and the corresponding weights.

A variant of this approach involves the re-specification of a number of the representativeprojects on a cyclical basis.

The difference between an index based on inputfactors and one based on standard components isessentially one of degree, since components areonly factors at a more advanced stage ofproduction. However, the standard component costindex also incorporates productivity gains andchanges in profit margins, as it reflects not only thecost of the factors, but also the price of the finishedproduct paid by the customers of the constructionfirm. It will therefore include more of the priceelements listed above that determine the prices offinished construction work.

Another advantage of this method is that the pricesobtained are for components which remaincomparable over time. Fluctuations due todifferential quality or execution are eliminated.However, while components are morehomogeneous than completed buildings it isunlikely that they will be completely identical fordifferent completed buildings.

If there is a permanent change in the way acomponent is performed its definition is changed,and the new price series is spliced to the old one bymeans of chaining taking account of changes inquality.

Selecting truly representative standard components,and determining weighting coefficients thataccurately reflect construction techniques and theway buildings, etc. are usually constructed duringthe base year is the hardest and most important part

of compiling the index. However, every standardcomponent does not have to be priced separately.For example, if the price change characteristics ofbrick laying and paving are similar then prices ofonly one of (similar) standard components need tobe incorporated in the construction price index.

An advantage of both the standard factor andstandard component cost methods is that they canproduce different indices merely by changing theweighting of the indices for each component. Forexample, indices could be compiled by type ofwork, by trade, or as in the example of theconventional construction price indices compiled inGermany, by type of building.

Subsequent Breakdown Methods

Subsequent breakdown methods involve the use ofsamples of either actually completed, or fictitiousconstruction projects. They entail the collection ofprices effectively charged (or what would becharged) by the builder and incorporate changes inproductivity and profit margins. These methodsstart from the completed building or projectwhich are then broken down into component parts.

These methods are used primarily for thecompilation of output price indices. The hedonicand matched model methods outlined below arealso used by Member countries to compile sellersprice indices.

Quoted prices: In this method the problem ofcomparability of components between constructionprojects is overcome by having respondents quoteprices for a standard construction output product(house, apartment, bridge, school, etc.) whosespecifications are kept constant from one period tothe next.

For example, the detailed specifications for a house(of specific type, style, size, etc.) are distributed toa sample of building firms. Bids or quotes are thencollected for the entire building as specified, aswell as the electrical work in the specifications, thebrick work, etc. Respondents are asked to bid asthough they were tendering for real work, takingaccount of prevailing market conditions and costs.The bids for each component are averaged andweighted up to an overall price, and a price index iscomputed.

19

The standard construction (house, apartment,bridge, school, etc.) is updated periodically toreflect changes in materials, styles, etc. to ensurethat it is typical of those being constructed at thetime.

A major problem with this method is that it isdifficult for firms to take the process seriously.They are not bidding for real work and there is nobargaining involved.

Schedule of prices: This method entails theselection of a representative sample of constructionprojects either taking place, or completed, in agiven geographic area, over a specified period oftime. The cost of each technical component(derived from blueprints, work specifications, etc.)of a given construction in the sample is priced as atthe base reference date. This involves the use of aschedule of prices containing the price of eachcomponent of the construction at the base perioddate. Examples of this method are the constructionprice indices for residential buildings compiled inFrance, and several output price indices compiledin the United Kingdom.

By aggregating the prices for all the components atheoretical average price of the entire constructionis obtained as though it had been undertaken at thebase reference date. The general weighting isobtained from statistics on current construction. Aprice index is then obtained by calculating the ratioof the current actual price of the sampledconstruction to the recalculated price at the basereference period (derived from the sum of itscomponents compiled from the schedule of prices).

The role of the schedule of prices is to define aprice structure, not the average level of eachcomponent. The composition of the schedule ofprices may not necessarily reflect the averageprices for the base period. Obtaining such anaverage would require a very large sample.However, the relative structure of the prices on theschedule reflects market conditions at the timeperiod in question.

Matched models: Involves the specification of asample of representative construction projects (ormodels), the matching of these model projectsagainst actual projects carried out by contractors ina specified period, the collection of prices for the

matched projects for each period, and the weightingtogether of price movements for each project.Constant quality is maintained by calculating pricemovements on a matched sample basis (i.e. theprice movements between adjacent periods arebased on the same model projects each period).

This methodology is used for the calculation ofprice indices for standard project homeconstruction, as in the case of the price indices forthe construction and renovation of privately builthouses compiled in Australia. The composition ofthe list of construction projects or models to bematched against actual construction projects issubject to continuous review, and is determined bythe continued availability of price informationrelating to a particular model specification. If thespecification of an individual model changessignificantly, or if a price is unable to be obtained,then that model is excluded from the calculation ofprice movement.

Price information for the actual finishedconstruction is obtained for each period from asample of builders/construction firms, real estateorganisations, government agencies, etc. Theyrelate to actual sales transacted during the period.

Building volume or area5: For this index the cubicmetre is used as a common denominator tocompare the costs of a recent construction to costsin a base period. An index of the value per cubicmetre is calculated and adjusted for differences involume, quality, period, and region. Theseadjustments enable a price index to be prepared.However, the use of this method requires theconstruction activity included to be homogenous.The index is valid only for each reference building,etc. as it was originally built.

This method is no longer used by any Membercountry in its pure form, however building volumeis used in some countries (by the Netherlands for

5 Central Bureau of Statistics Netherlands,Voorburg - A Hedonic Approach to OutputIndices for Construction, George van Leeuwen,February 1995.

University of Geneva, Laboratory of AppliedEconomics, Geneva - Indices of ConstructionPrices, A Methodological Inventory, LynnMackenzie, April 1994.

20

example) as one of the building characteristics inthe regression analysis used in the compilation ofthe hedonic construction price indices describedbelow.

A variant of this method used in the compilation ofan output price index for dwellings in Spain usesthe average price per square metre as one of thevariables collected in determining the averageprices of monitored dwellings.

Hedonic method: Regression techniques may beused to construct hedonic indices to measurepurchasers preferences for the differentcharacteristics of construction work. Thisapproach, which is used in the compilation of someof the price indices compiled in the Netherlands,Sweden and the United States, starts from thepremise that each construction is a combination ofcharacteristics, each of which has an implicit price.This price is set by the market and is reflected inthe over-all prices for which different combinationsof these characteristics are sold, and where differentvarieties of the same construction type, each withits own peculiar combination of characteristics, co-exist.

In addition to total price a limited number (7 to 15)of characteristics (qualitative or quantitative) aretaken into account. For a house, possiblecharacteristics include: floor area, number offloors, type of garage, method of heating, numberof toilets, etc.

Using econometric techniques (regression andcovariance analysis) weight of each of thesecharacteristics in determining the price is estimated.These weights can then be used to factor out thatpart of the price change in the next month which isdue to changes in the characteristics of the house,etc. sold. The coefficients of the regressions arecalculated first by means of total construction priceinformation and their characteristics in the currentyear, and on the basis of information on the sametype for a base period.

Each construction during the current period ispriced at what it would have cost during thereference period, once its characteristics are known.Indices are calculated from current prices and thenaggregated using either Laspeyres formula if thevalues of the characteristics are taken from the baseperiod, or Paasche if they are taken from thecurrent period.

The problem with this method, which is based onjust a few elements of construction, is that it is lessdiscerning than the schedule of prices method indetecting qualitative change. If constructionquality is enhanced the improvement will tend to beunderestimated and thus the actual increase inprices overestimated.

Furthermore, the use of these techniques requiresthe statistical agency to have access to trainedeconometricians, as well as specialist knowledge ofthe construction industry.

D. OUTLINE OF PROCESSES IN DEVELOPING A CONSTRUCTION PRICE INDEX

As with both CPIs and PPIs, the development andcompilation of price indices for constructionactivity is a complex procedure consisting of a longand varied set of operations. The usefulness of theconstruction indices compiled also depends onhaving a clear understanding of the purposes of theindices, and the characteristics of the constructionindustry in the country where it is located. Thesecharacteristics include:

the range of construction activities conductedthroughout the country;

construction techniques commonly used foreach type of construction activity, together withan idea of the rate of change in techniques used;

types of entities/organisations undertakingconstruction activity, and their characteristics(e.g. size, industry concentration, etc.);

administrative arrangements for themaintenance of building/construction standards;

administrative arrangements for governmentauthorisation of individual constructionprojects.

21

These aspects must be determined beforecommencement of work on the creation of theindex. Relevant characteristics of the countryexternal to the construction industry also need to beidentified. These could be economic, demographic,geographic, or administrative.

The major processes in the development andcompilation of construction price indices using themodel price methodology outlined above are:

Selection of a small, representative group ofrecently constructed buildings, civil engineeringprojects, etc. as models. The number of modelsrequired depends on the range of constructionactivity to be included in the index, diversity ofthe specifications for each type of activity, andregional diversity .

Specification of the hundreds of detailed tasksor component trades in the construction of thesemodel projects. These are prepared usingarchitectural drawings and specifications. Alsoinvolves the development of components for thegeneral requirements (overheads and profitmargins) of the main construction contractor.

Selection of a sample of components. Theselection of components within each trade areais based on both money value and the coverageof significant materials and/or productsinvolved. A goal might be to select componentswhich cover at least 70 per cent of the totalvalue of the trade.

Development of specifications for eachcomponent to include quantities involved andbase-weight unit prices. Specifications must beexact to avoid the risk of varying interpretationby different respondents.

Selection of a sub-sample of subcontractors andgeneral contractors in the appropriate

geographic areas from whom prices arecollected. An important goal is to selectcontractors who are actively engaged inbuilding sample components and can reportprice quotes based on recent experience. Somerespondents might be able to supply quotes oncomponents included in more than one model.

Collection of periodic reports for a sample ofthese components from subcontractors. Theseshould be based on current prices they charge(including overheads and profit) for thecomponent they supply. Price collection may bedone by telephone or mail, generally after aninitial personal visit to gain co-operation anddiscuss reporting problems. The prices ofelectrical and mechanical equipment can beobtained from manufacturers of the equipment.

Calculation of a price index for the constructionas a weighted combination of these componentprices. This is done by multiplying new pricequotations by base period weights, andcomparing the result to base period modelprices.

Development and implementation of an on-going process of index review to revise the listof model projects, weights, component items,respondents, etc.

It could take as long as year to set up the models,enlist respondents, and begin to collect data.

Further information on each of these processes isprovided below in Section 2 of this publication.

E. USES OF CONSTRUCTION PRICE INDICES

Construction price indices are primarily used foranalysis of price movements and price formation inthe construction industry, for price escalationclauses in construction contracts, and for deflationof components of the national accounts.

More specifically, the primary uses are6:

6 German Federal Statistical Office, Wiesbaden,Detailed indices for Construction Prices andConstruction Work, Series 4, May 1994.

22

Measuring the changes of prices of constructionmaterials for construction work.

In developing a program of projects, preparingestimates, comparing estimates with bids, andscheduling projects within funding limits it isnecessary to have some way of judging pricemovements. The aim is to express physicalvolumes of work needed for future constructionwork in value terms.

Studying the impacts of changing prices overthe total construction cost and selling prices ofthe construction work.

Measuring the expenditure of consumedmaterials at constant prices.

Estimating the short-term evolution of prices.

To determine replacement values for insurancepurposes.

The use of construction price indices (wherequality and other changes in the pricedetermining characteristics of the constructionoperations observed have been eliminated) canhave considerable impact if they are used todetermine replacement values. If constructionwork of the original quality is no longersupplied because of substantial changes inmaterials, techniques, etc. the replacementvalues obtained from the use of constructionprice indices may be considerably less than theamount actually required to be spent on thereplacement.

Realising price-index readjustments ofconstruction contracts.

Planning the production of materials andchecking the efficiency of entrepreneurial units.

Deflating components of the national accounts.In addition to the compilation of nationalaccounts at current prices there is a necessity inhaving constant price measures that separate theeffects of price and volume increases (ordecreases). This necessity is particularly strongin countries experiencing high inflation.

Use of Construction Price Indices forDeflating Components of NationalAccounts

National accounts constant value estimates areusually made by deflating estimates in currentvalues by appropriate price indices at as fine a levelof detail as possible and adding up the results. Theaim of such calculations is to enable the analysis ofperiod-to-period changes, by separating GDP andits sub-aggregates into distinct price and volumecomponents.

One of the main problems in the compilation ofprice indices for national accounts deflationpurposes is that of ensuring that an appropriateprice index is used for the national account itembeing deflated. In the context of construction priceindices the issue is primarily one of matching thenational account item being deflated with the itemcoverage of the price indices used. The formerprimarily involves the deflation of constructionindustry output and/or gross fixed capital formation(GFCF). The deflation of these components of thenational accounts by OECD and European UnionMember countries often entails the use of differentconstruction price indices.

The System of National Accounts (SNA) makes thedistinction between two types of prices for valuingoutput: basic prices; and producers prices.

The basic price is the amount receivable by theproducer from the purchaser for a unit of goodor service produced as output minus any taxpayable, and plus any subsidy receivable, onthat unit as a consequence of its production orsale.

The producers price is the amount receivableby the producer from the purchaser for a unit ofa good or service produced as output minus anyVAT, or similar deductible tax, invoiced to thepurchaser.

Both prices exclude transport charges invoicedseparately by the producer.

The preferred basis of valuation for deflatingconstruction industry output is at basic prices,though producers prices may be used whenvaluation at basic prices is not feasible. TheEuropean System of National and Regional

23

Accounts in the European Community (ESA)specifies that only basic prices may be used.

Care should be taken to ensure consistency betweenthe valuation of industry output and the valuationof the price index used for deflation.

For all transactions on the uses of goods andservices such as final consumption, intermediateconsumption, and capital formation) both the SNAand the ESA specify purchasers prices as thepreferred basis of valuation.

The purchasers price is the amount paid by thepurchaser, excluding any deductible VAT orsimilar deductible tax, in order to take delivery of aunit of a good or service at the time and placerequired by the purchaser. This price includestransport charges paid separately by the purchaserto take delivery.

Therefore, on the expenditure side a purchasersprice index is required for the deflation of items ofconstruction falling within GFCF. For new fixedassets these include not only all transport andinstallation changes but also the costs incurred inthe transfer of ownership, in the form of fees paidto surveyors, engineers, architects, lawyers, estateagents etc., and any taxes payable on the transfer.

Output construction price indices compiled byMember countries include a number of costelements to enable the index to approximate basic,producers, or purchasers prices as defined in theSNA. The essential differences between the threeprice valuations being the inclusion or exclusion oftaxes, subsidies, and transport charges.

Due to the lack of suitable available output priceindices some countries use input price indices fordeflation purposes. Some use these indices makingallowance for productivity change, variations inprofits, and incorporation of overheads.

Sellers price indices which include cost itemsoutside the scope of the production account forconstruction, and GFCF (in particular the cost ofthe land) may be less suitable for national accountdeflation purposes.

24

SECTION 2. CONSTRUCTION PRICE INDEX COMPILATION ISSUES

A. IMPORTANT CONSIDERATIONS IN THE COMPILATION OF CONSTRUCTION PRICE INDICES

Most of the procedures involved in thedevelopment and calculation of construction priceindices (outlined in Section 1D above) are similarto those used in the compilation of CPIs and PPIs.However, listed below are a number of problemareas that are either unique to construction activity,or where particular attention is required whenconstruction price indices are developed. Theseissues are described in more detail in Section 2D.

Diversity of Construction Activity

To portray price changes by means of a price indexit is important for the data to relate to types of workthat remain constant. The main problem in thecompilation of price indices for construction arisesfrom the fact that the products/outputs ofconstruction activity (houses, apartments, factories,schools, bridges, dams, etc.) are seldom exactlycomparable with one another. Differences arisedue to variations in the plans/specifications,production location (building site), etc. for eachproduct. Civil engineering and non-residentialbuilding outputs are particularly heterogeneous.

The range of compilation methods used tominimise the impact of this diversity are outlined inSections 1B and 1C above. These include pricingcomponents rather than the entire constructionproject and the selection of individual types ofwork (or model) that regularly occur.

Another factor to consider is the type ofconstruction activity to include in the scope of theindex. Both the components, and the relative use ofthese components (weights) vary significantlybetween each type of construction activity. Eachconstruction price index relates to predeterminedtypes of construction activity and care is required toavoid the use of indices for unrelated types ofconstruction work.

Changes Over Time

Given that both the inputs to, and outputs ofconstruction activity constantly change, anotherproblem in compiling construction price indices isthat of ensuring comparability over time. Factorsthat influence such change include size, finish,style, equipment used, fashions, etc. This problemdoes not exist to the same extent for consumeritems where a price index can be compiledimmediately for the entire period over which suchproducts, their quality unchanged, are sold.

To ensure that the values reported for a particularprice series will show only pure price changes allthe factors contributing to a price level (the so-called price determining characteristics) must bekept constant as long as possible.

One method commonly used by OECD andEuropean Union Member countries to compare twoor more construction projects (which are alwaysdifferent), involves breaking the projects down intoelementary components that are more comparableover time.

Selection of Appropriate Prices

The selection of the appropriate price is also animportant consideration in the development of aconstruction price index. The following types ofprice data are generally available for constructionactivity7:

tender prices prices at the time of award of contract invoice prices

Obtaining the right transaction price as opposedto list prices, the lowest price, etc. is critical. Listprices for example need to be adjusted for:

7 Austrian Central Statistical Office, Vienna,Statistische Nachrichten - Index of BuildingPrices.

25

cash discounts; competitive discounts; seasonal discounts; trade discounts (i.e. a discount made to a certain

class of buyer); quantity discounts.

A difficulty with one type of model pricing is thatcontractors may be asked to provide hypotheticalprices, what they would charge, rather than actualtransaction prices. It is also difficult to ensure thatthe prices compiled reflect current marketconditions and include a realistic allowance foroverheads and profit.

Finally, as with other types of price indices, it isalso difficult to separate changes in price fromchanges in quality. The latter may occur when abuilder or subcontractor alters the way thatcomponent is supplied, for example through the useof a new material or new equipment.

A prerequisite for dealing with this issue is to havestaff with detailed knowledge of the constructionindustry to ensure that price information providedby material suppliers, construction contractors, etc.can be evaluated for reasonableness, and who canask probing questions where necessary. Pricecollection staff must also be particularly alert tochanges in the quality of components.

Range of Items for Inclusion

As mentioned in Section 1B above there isconsiderable variation in the type and number ofitems included, even for the same types of indices,compiled by OECD and European Union Membercountries.

The construction of a building or civil engineeringproject often involves thousands of complexoperations and components. In developing aconstruction price index a decision is required onthe number and types of items to include to ensurethat the resulting index will accurately reflect pricechanges. The choice of items to be included in aparticular index is strongly influenced by the typeof index being compiled (input, output, sellers),and its purpose. For example, if the primarypurpose of the index is to deflate components of thenational accounts, the item coverage of the index

should match the scope/content of the nationalaccount component being deflated.

Another consideration in compiling each type ofconstruction price index is the need to distinguishbetween the items paid by the producer (thecontractor) and those paid by the purchaser or finalowner.

Items considered for inclusion or exclusion in anyspecific index are profits, overheads, margins, theprice of land, site preparation, architects andsupervisors fees, transport, legal fees, certain typesof installations (central heating, elevators, cabletelevision systems, garages, etc.). TheEUROSTAT publication Industrial Short TermIndicators - Methodology outlinesrecommendations, rules and regulations concerningthe composition of short term indicators (includingconstruction price indices) for European UnionMember countries.

To compile indices comparable between EuropeanUnion member countries EUROSTAT recommendsthe inclusion/exclusion of the following elementsfor the three types of construction price indicesdescribed in Section 1 of this publication.

Input: Should include at least the prices of rawmaterials and wages paid by the constructioncontractor, though not changes in thecontractors productivity and profits, or theprice of land, architects fees, etc. Input indicescan also include other input factors such asequipment, transport or energy.

Output: Should include prices paid by thecontractor for all factors contributing to theconstruction activity paid by the client. Theseshould include those listed previously forinclusion in input indices, plus changes inproductivity, profits. Output indices should notinclude elements not normally paid by theclient. These could include the price of land,architects fees, etc.

Sellers: Should include all the elementsincluded in both input and output price indicesplus the price of land, and other elementsincluded in the total sales price of the completedconstruction paid by the purchaser or finalowner.

26

Ensuring Geographic Representativenessof the Index

Regional differences can have a major impact oncosts, prices, size, style (single dwelling versusmultiple dwelling constructions, low versus highdensity dwellings, etc.), construction materialsused, methods used, etc. It is particularly importantto identify any rural/urban differences. If an indexis to be applied nationally there is a need to ensurethat regional differences are represented in the pricecollection process, and that the index is compiledwith the appropriate weights reflecting thesedifferences.

Types of Construction Firms to IncludeConstruction activity may be conducted byspecialised businesses predominantly engaged inconstruction, either as prime contractorsresponsible for the entire project, or sub-contractorsinvolved in specific (and specialised) trades.Construction activity may also be carried out bybusinesses classified to other industries, or bygovernment. At the other extreme, constructionactivity (particularly house construction) may beundertaken by individuals for their own use (i.e.owner-builders).

The materials, construction techniques used, andeven the type of construction activity undertaken,may vary according to the type of entity involved inthe construction activity.

In transition economies there is a particular need toensure that data collection is extended to the newlyemerging private organisations (and newconfigurations of old enterprises), as well as state-owned enterprise. In other words, price data mustbe obtained from organisations representative ofthe whole range of construction activity. Forexample, selection of only the largest enterprisescurrently operating might result in the emergingprivate sector (which may only comprise smallerentities) being under-represented, or omittedaltogether.

Related to the issue of the type of constructionfirms, etc. to include in the collection of priceinformation is inclusion or exclusion of foreigncontractors, or domestic contractors workingabroad. Consideration needs to be given as towhether or not the price information obtainedreflects the price change characteristics of thecountry in which the index is being compiled.

B. PARTICULAR PROBLEMS/ISSUES FOR TRANSITION COUNTRIES

Many transition countries are currently developingtheir own construction price indices, or plan todevelop these indices in the near future. In additionto the problems faced by market economiesoutlined above, transition countries need to take anumber of particular issues and concerns unique totheir situation into consideration. These include:

The inclusion of construction activities in theprice indices that are experiencing particularlydramatic growth during the transition period.These include the construction of hotels, shops,and private residential buildings. Whilst smallin numbers these types of construction could besignificant in terms of the gross output of theconstruction industry. At the same, work onmajor new state projects has almost beencompletely suspended in many transitioncountries.

Movement away from the dominance of theconstruction industry by large state enterprises

by the entry of a larger number of smaller andperhaps more specialised constructionenterprises, e.g. firms specialising in theinstallation of fittings in newly restituted shops,building of private houses, etc. Related to thisissue is the need for transition countries todevelop techniques/procedures, questionnaires,etc. to obtain the required price informationfrom construction enterprises (of all sizes) in theemerging private sector.

The inclusion of construction activity carriedout by unauthorised persons.

The inclusion of barter arrangements betweenconstruction enterprises.

27

C. SOURCES OF INFORMATION USED TO COMPILE CONSTRUCTION PRICE INDICES

From the perspective of the organisationresponsible for compiling a construction priceindex the selection of the most appropriate sourcesof information used depends on a number ofconsiderations. These include the use(s) to whichthe index will be put and the type of index required(input, output), data availability, and resources(both financial and skill) available in theorganisation. The institutional background inwhich a construction price index will be developed,will to a large extent determine the feasibility andcost of the index, and influence the methodologyused in its compilation.

The data required to compile an index may be inthe form of price information obtained fromconstruction enterprises or materials suppliers.Alternatively, components of the price index maybe obtained from administrative agencies chargedby law with the responsibility for compiling basiccosts indices, or by professional associations foruse as benchmarks for contracts in the profession.Information may also be obtained from the clientwho may be traced through building permits.

The following table summarises possible sources ofinformation for compiling construction priceindices.

Types of Information Required for Price Index Possible Data Sources

Input prices - materials Surveys of construction, building materialsupply, or building material manufacturingenterprises.

Associations of quantity surveyors Trade associations Chambers of commerce Other national statistical office collections (e.g.

for producer price indices Government agencies charged with responsibility

of monitoring capital construction works(particularly roads, dams, bridges, etc.)

Input prices - labour Trade unions Trade associations Collective bargaining agreements registered with

government Government agencies charged with responsibility

for regulating wages. Enterprise surveys of employers Household surveys of employees

Transport costs Transport associations Surveys of transport enterprises Relevant government agencies

Weight information Construction branch structural enterprisesurveys

Associations of quantity surveyors Architectural enterprises/associations

28

The actual sources of information used byOECD and European Union Member countriesin the compilation of specific construction

price indices are listed in Country SummaryTable 1.

D. MAJOR ELEMENTS OF CONSTRUCTION PRICE INDICES

The major elements of construction price indicesare outlined below. These elements are interrelatedin the sense that a series of trade-offs andadjustments are made between them by theorganisation developing the index. These trade-offs are largely determined by the proposed use ofthe index, data availability, and resources availablefor developing and compiling the index. Forexample, in the case of a price index for houseconstruction, the requirement for an indexencompassing both rural and urban areas of acountry might for cost reasons be traded off infavour of including only urban house constructionby a decision to include price information for alarger number work categories, materials, etc. Sucha trade-off could be justified, for example incountries where an overwhelmingly largeproportion of house construction takes place inurban areas.

Construction Type/Activity Coverage

The decision by OECD and European UnionMember countries on the types of construction orconstruction activities to include in the constructionprice index (or indices) they compile largelydepends on the use(s) of the completed index.Some countries compile special purpose indicescovering a very narrow range of construction types(e.g. one family houses in the United States, socialhousing in Mexico, bridge construction in Austria,etc.), whilst others compile separate indices for all,or almost all, construction activities (e.g. Austriaand Australia). A number of Member countriescompile a weighted composite index of allconstruction types (e.g. Belgium and Japan).Finally, some indices include only constructiontypes undertaken on behalf public or private sectorclients (e.g. the index for privately built apartmentsin Iceland).

Several characteristics of activities undertaken inthe construction sector result in difficulties incompiling construction price indices. Thesecharacteristics include:

the fact that the resultant indices are mainly notbased on the total turnover of an institutionallydefinable sector, but on product categories (ortypes of construction);

the production of these types of constructionworks (residential buildings, non-residentialbuildings, roads, etc.) by several industrybranches; and

the inclusion of these activities (joinery, steelconstruction, etc.) in other branches of theeconomy.

The type of entity (construction material supplier,construction contractor) to be priced will varyaccording to the use(s) of the construction priceindex, and the type of index being compiled (input,output, sellers). If used to deflate components ofthe national accounts, the item content of the indexshould as far as possible, correspond to the valuecoverage of the national account item to bedeflated. Alternatively, if the index is to be used tore-adjust long-term contracts, the types ofconstruction projects included in the index shouldmatch construction activities for which contractsare to be indexed. This may not be possible in asingle construction price index, and it may benecessary to compile a number of indices.Alternatively, it may be necessary to use one indexthat best fits all or most of the uses, or from whichmost of the purposes can be met.

The compilation of price indices to portray themovement of prices across all construction typesinvolves the selection of individual projects fromeach type of construction. The impact of thediversity (particularly non-residential building andcivil engineering projects) of construction activityon price change is reduced by selecting a number ofrepresentative projects from each construction type.Some countries select projects carried out over anumber of years to eliminate any bias that mightoccur if a shorter period were chosen.

Matching the construction types or activities to beincluded in the index with proposed use(s) of the

29

index entails use of a combination of differentclassifications to define the activity, industries,occupation, type of construction, etc. that areincorporated in the index. The mix ofclassifications used varies between each Membercountry, and between the different constructionprice indices compiled in each of these countries.

A number of countries have developed their ownclassifications to take account of local conditions,institutional/administrative arrangements, and workpractices. In almost all instances these nationalclassifications are compatible with standardinternational classifications. These include:

General classifications of industries (NACERev. 1; ISIC Rev. 3): According to whichenterprises are classified.

Classification by type of constructions(residential buildings, non-residential buildings,civil engineering, etc.), for example the draftEUROSTAT Classification of Types ofConstructions: Very often, different indices arecalculated for each of the main types ofconstruction.

Classification by trade or occupation(carpenters, bricklayers, plumbers, etc.), forexample the International StandardClassification of Occupations (ISCO): Thewage component used in input prices is oftenobtained from existing indices compiled bytrade associations, etc.

Classification by construction work categories(brick works, marble tiling).

Classifications of geographic areas.

It is recommended that countries developingconstruction price indices use either internationalclassifications, or nationally developedclassifications that are compatible withinternational classifications. Extracts from some ofthese international classifications showing areasrelevant to construction price indices are providedin Section 5 of this publication.

Geographic Coverage

Some OECD and European Union Membercountries compile national construction priceindices from the (weighted) average of a set ofregional indices, whilst others compile regionalindices from national indices. The need to observeconstruction prices across capital cities, towns,villages and rural areas depends on the extent towhich price changes differ in each of the regions.It also depends on the resources available todevelop regional indices and to collect regionalprice data.

Where regional price trends do not differsubstantially, it is sufficient merely to distinguishmetropolitan areas, other large cities, small cities,villages, and rural areas. Some countries (e.g.Greece) compile indices only in respect of capitalcities, on the assumption they are representative ofthe urban areas for the entire country. Whereregional differen