Index of Retail Sales Norway

Transcription

1 2010 Voorburg Group Meeing Vienna, Sepember 2010 Mini presenaion Index of Reail Sales Norway Session on disribuive rades Saisics Norway Linda V.M. Præsiin and Øyvind S. Bolsgård

2 Table of conens Inroducion...3 Size of he wholesale and reail rade indusry...3 Overview of saisics ha cover he wholesale and reail rade indusries in Norway Background The purpose of he survey Mehodology Daa collecion Saisics Norway s Regiser of Esablishmens and Enerprises VAT regiser Daa from he sample of idenical unis Daa from head offices of chain sores Price index for reail sales Populaion Coverage Srucure in NACE rev Annual updae of he populaion and division ino subpopulaions Monhly updae of he populaion Sraificaion Sample of idenical unis Number of unis Roaion of he sample Calculaion of he value index Calculaion of he index wih he aid of chaining Calculaion of he urnover in a saisical monh Deflaion / Volume index Working day and seasonal adjusmen Resuls Comparisons wih oher saisics Index of household consumpion of goods Srucural saisics for wholesale and reail rade and bimonhly wholesale and reail sales saisics Improvemens Mehods of daa collecion Inerne-based reporing Phone-based reporing NorSamu - Coordinaed drawing of samples for economic surveys ISEE - Inegraed Sysem for Ediing and Esimaion Summary...21 References...22 Appendix A: Example of an Inerne-based quesionnaire for daa collecion from he sample of idenical unis...23 Appendix B: Deailed srucure of Division 47 Reail rade, excep of moor vehicles and moorcycles in NACE rev

3 Inroducion The purpose of his paper is o give a shor presenaion of Norway s index of reail sales as inpu o he session on disribuive rades a he Voorburg Group Meeing in Saisics Norway, however, produces a wide range of saisics ha describe he disribuive rades in Norway. An overview of hese saisics is given below, afer aking a look a he size of he wholesale and reail rade indusry in Norway. Nex, he index of reail sales is presened. This paper will explain he background and he purpose of he index, as well as is mehodology. Furhermore, he saisical resuls are shown and compared wih oher saisics. Improvemens ha have been and are being made in he saisics are also covered. Size of he wholesale and reail rade indusry The urnover of local kind-of-aciviy unis wihin wholesale and reail rade indusries came o NOK billion in Tha is equivalen o an increase of 64 per cen from Approximaely a hird of his urnover was from unis wihin reail rade, excep of moor vehicles and moorcycles. Diagram: Turnover by wholesale and reail indusry division. Local kind-of-aciviy unis. Billion NOK Billion NOK Wholesale and reail rade and repair of moor vehicles and moorcycles 46 Wholesale rade, excep of moor vehicles and moorcycles 47 Reail rade, excep of moor vehicles and moorcycles Source: Saisics Norway, Sabank able 07916: Wholesale and reail rade, srucural saisics. Overview of saisics ha cover he wholesale and reail rade indusries in Norway Srucural business saisics The wholesale and reail rade srucural business saisics are par of he indusry saisics and provide deailed informaion abou he aciviy and srucure of secion G of he EU's Saisical Classificaion of Economic Aciviies (NACE Rev. 2). This secion comprises he aciviies: 45 Wholesale and reail rade and repair of moor vehicles and moorcycles 46 Wholesale rade, excep of moor vehicles and moorcycles

4 47 Reail rade, excep of moor vehicles and moorcycles The following variables are published: Enerprises Local kind-of-aciviy unis Owners Employees Employmen Par-ime employees Man-labour year Compensaion of employees / wages and salaries Social securiy coss Turnover Cos of goods Value added Merchandise sold Gross earnings Change in socks Acquisiions Sale of business asses Gross invesmens New invesmens The saisics are based on he daa needs of he Naional Accouns and are in accordance wih he EU regulaion on srucural business saisics. In addiion, he public's need for informaion has o be me, and minisries and various rade organisaions are among hose requesing saisical daa. Business, indusry and oher users also commission cusomised wholesale and reail rade saisics from Saisics Norway, primarily a he municipal level. The srucural saisics are published wice a year. Preliminary figures are presened 10 monhs afer he end of he reference period, and final figures wihin 18 monhs of he end of he reference period. The figures can be found on Saisics Norway s websie: hp:// More deailed produc informaion The EU Council regulaion on srucural business saisics requires a breakdown of urnover by produc ype every fifh year for all he aciviies in secion G. These figures can be found on Saisics Norway s websie: hp:// In connecion o his breakdown of urnover by produc ype, a wholesale and reail earnings survey is made. I gives sales income on merchandise sold for differen wholesale and reail rade subclasses (NACE five-digi level) and groups of commodiies, as well as gross earnings by produc ype. The saisics are based on he needs of he Naional Accouns, bu he Federaion of Norwegian Commercial and Service Enerprises srongly desires he producion of he saisics oo. A Saisics Norway he saisics are used by he Division for Naional accouns and oher research and analysis aciviy. Oher users include public and privae secor agencies, organizaions and individuals. The index of reail sales The index covers oal reail sales of all goods excluding moorised vehicles (SIC2007:47). I is regarded as an imporan shor-erm indicaor of consumer demand. The purpose of he saisics is o follow and analyse he ongoing economic developmen. The saisics are normally published on he 28h or 29h day of he following monh, on hp:// Bimonhly and quarerly wholesale and reail sales saisics The purpose of he saisics is o idenify he level and he developmen of wholesale and reail rade. The bimonhly saisics cover all he aciviies in secion G, excep of agency business (SIC2007: 45-47, excep of 46.1), while he quarerly saisics cover aciviies 45 and 46. The bimonhly and quarerly urnover saisics are published on hp:// abou seveneen weeks and sixy days afer he expiraion of he saisical period respecively. The price index for wholesale rade The purpose of he saisics is o measure he price developmen for esablishmens in wholesale rade, excep of moor vehicles, moorcycles and wholesale on a fee or conrac basis (SIC2007: 46, excep 46.1). The figures are made available on hp:// approximaely 5-15 days afer he given quarer.

5 The volume index for wholesale rade The purpose of he saisics is o measure he changes in volume for esablishmens in wholesale rade, excep of moor vehicles, moorcycles and wholesale on a fee or conrac basis (SIC2007: 46, excep 46.1). The volume index is calculaed by deflaing he value index for wholesale rade wih he price index for wholesale rade. The figures are released on hp:// approximaely 60 days afer he given quarer. The index of household consumpion of goods The index measures he developmen of he household consumpion of boh durable and non-durable goods. Since he definiions in he quarerly naional accouns are used in he index of household consumpion of goods, i can be inerpreed as an indicaor of he households purchases of goods. I is usually released abou a monh afer he monh of reference, on hp:// Invesmen saisics for reail sales The index covers he developmen in esimaed and final invesmens in reail rade, excep of moor vehicles, moorcycles and auomoive fuel (SIC2007: , excep 47.3). The purpose of he saisics is o follow and analyse he ongoing economic developmen. The index is released on hp:// approximaely 60 days afer every quarer. The consumer price index The consumer price index measures he acual changes in he prices for household goods and services including charges and fees. The saisics are published on hp:// on he 10 h of he monh, around 3 weeks afer he 15 h of he curren monh. Sales on alcohol saisics The purpose of he saisics is o map he developmen in regisered urnover for beverages conaining alcohol in Norway. The sales figures are published abou 90 days afer he given quarer. They can be found on hp:// Cross-border rade Saisics Norway's ravel survey (which covers overnigh rips abroad) and survey on cross-border rade (which covers day rips abroad) give a oal overview of he size of Norwegian households' foreign ransacions, which is an imporan inpu ino he saisics on foreign rade. They are published on hp:// approximaely eigh weeks afer he quarer of reference. Sales of peroleum producs The purpose of he saisics is o show deliveries of peroleum producs by purchaser group, couny and a naional level, measured as delivered lires. Furhermore, he saisics compare he delivery figures of he differen oil companies (user-financed assignmen). In addiion, prices on some seleced peroleum producs are published from February The saisics are repored on Saisics Norway s websie hp:// abou days afer he given monh. Exernal rade in commodiies The purpose of he exernal rade saisics is o give informaion abou he commodiy flows beween Norway and oher counries. Expors and impors are imporan economic indicaors boh in describing srucural changes and in monioring he economic rends. Preliminary figures for he preceding monh are released on he 15h, or he firs subsequen workday. Final figures for he preceding year are released in April or May. They can be found on Saisics Norway s websie: hp://

6 Regiserbased employmen saisics The purpose of he saisics is o describe he employmen siuaion, disribued by indusry division a a more deailed regional level. The saisics are published annually, on hp:// Publicaion normally akes place in mid-june, referring o he siuaion in he fourh quarer of he previous year. Employmen and unemploymen among shor erm immigrans. The purpose of he saisics is o describe he employmen siuaion for persons no regisered as residens, i.e. persons who inend o say in Norway for a shor period. The saisics are published annually, on hp:// Publicaion normally akes place in mid-june, referring o he siuaion in he fourh quarer of he previous year. Sickness absence saisics The saisics are inended o shed ligh on sickness absence rends in he aggregae and disribued by indusry, occupaion, hours worked and various personal variables conneced wih hose on sick leave. These are inended o form he basis for developing and evaluaing measures aimed a sickness absence. Aims ha have governed he planning of he saisics are: The possibiliy for employers and funds-appropriaing auhoriies o make reliable esimaes of expenses. The possibiliy for employers' and employees' organisaions o evaluae he efficiency of various measures implemened o reduce sickness absence. The possibiliy of idenifying risk groups in order o implemen prevenive measures. The possibiliy of comparison over ime as he basis for research in he area. The saisics are published quarerly and annually. They are normally released weeks afer he end of he saisical period, on hp:// Wage index The purpose of he index is o show changes in average monhly earnings and basic salaries hroughou he year. The index is released on a quarerly basis, no more han 60 days afer he end of he quarer. I can be found on Saisics Norway s websie: hp:// Accouning saisics The purpose of he accouns saisics is o compile daa for enerprise-relaed economic overviews and analyses for naional accouns and research. The saisics provide a basis for comparisons over ime across indusries, aggregaes and regions. Also, he saisics provide facs for comparisons and evaluaions of individual enerprises' accouning daa agains he groups for which saisics are provided. Preliminary annual figures are published abou 38 weeks and final figures abou 75 weeks afer he end of he fiscal year. They are repored on hp:// The quarerly and annual naional accouns saisics The naional accouns saisics are designed o provide a consisen and comprehensive survey of he overall naional economy. The annual naional accouns provide a summarised descripion of he economy as a whole and a deailed descripion of ransacions beween differen pars of he Norwegian economy, and beween Norway and he res of he world. The Norwegian naional accouns provide an imporan source of informaion for analyses of he Norwegian sociey. The naional accouns have many users, from pupils and sudens o public or privae insiuions ha acively use naional accouns figures in analyses of economic srucures and developmen. Among oher hings, Saisics Norway s macroeconomic models have been developed based on he annual naional accouns. Oher major users of he Norwegian naional accouns daa are he Minisry of Finance, he Cenral bank of Norway, research and developmen insiues, and financial analyss. In addiion, inernaional organisaions such as he IMF, he OECD, he World Bank, he UN and he saisical office of he EU use he repored daa in heir saisics.

7 The annual naional accouns saisics are published 23 monhs afer he end of he year, on hp:// The quarerly naional accouns are published abou 50 days afer he end of he given quarer, on hp:// 1. Background Saisics Norway sared publishing an index of reail sales in one year afer he Norwegian Parliamen inroduced a bimonhly ax on wholesale and reail sales. The ax was calculaed as a fixed percenage of urnover and was, herefore, an accepable basis for he calculaion of a reail sales index. In 1970, however, he replacemen of he urnover ax wih a value added ax eradicaed he daa source of he monhly index, which was of fairly high qualiy a he ime (Saisics Norway 1976). Ye, i did no ake long before a new index of reail sales alhough no of as good qualiy - could be calculaed, based on daa colleced direcly from a sample (Saisics Norway 1976). In order o improve he qualiy of he index, he sample drawing plan was alered in 1992 and in In addiion, we gradually sared collecing urnover figures for more and more chain sores, direcly from head offices, from he end of The qualiy of he Norwegian index of reail sales was enhanced grealy because of his. While he sample only covered abou 40 per cen of he populaion s urnover in 1998, he coverage of he sample and he chain sores ogeher was nearly 80 per cen in The purpose of he survey The purpose of he index of reail sales is o describe he monhly developmen in he value and he volume of reail rade, excep of moor vehicles and moorcycles. I is regarded as an imporan shorerm indicaor of consumer demand. The index is used exensively by he public secor (minisries, he Cenral Bank of Norway) and he financial secor. Privae secor agencies and organisaions are regular users oo. A Saisics Norway, he Division for Naional Accouns relies on imely producion of he reail sales index. In he quarerly naional accouns saisics 1, which are mosly used for observaion and analysis of he curren economic cycle, he index of reail sales is cenral for he calculaion of household consumpion and, indirecly, producion in he wholesale and reail rade indusry. The index is also used in he preliminary figures of he annual naional accouns 2 saisics unil he bimonhly wholesale and reail sales saisics are available. 3 Moreover, he primary daa of he index is used in analysis and research in Saisics Norway. 3. Mehodology 3.1. Daa collecion Saisics Norway s Regiser of Esablishmens and Enerprises Saisics Norway s Regiser of Esablishmens and Enerprises firs file in Ocober conains he esablishmens in he populaion. The esablishmens' organisaion number, sae of aciviy and indusrial classificaion (NACE five-digi secor level or subclass) are imporan for updaing he populaion and spliing i ino subpopulaions. The indusrial classificaion is also used ogeher wih he number of employees o sraify unis. Furhermore, i is he basis for he calculaion of he oal urnover a NACE wo o four-digi secor level.. 1 hp:// 2 hp:// 3 hp://

8 VAT regiser The VAT regiser is he source for he bimonhly wholesale and reail sales saisics. We use he organisaion numbers, he urnover figures and he indusrial classificaion of he esablishmens in he second erm (March and April) of hese saisics. When updaing he populaion and spliing i ino subpopulaions, he organisaion number and urnover of he unis are of imporance. Informaion on urnover is also needed when excluding unis from he sample drawing process and inflaing sample daa o populaion level Daa from he sample of idenical unis From he sample of idenical unis we collec urnover 4 and VAT-daa via posal or Inerne-based quesionnaire 5. I is send o he sample on one of he las days of he saisical monh and is due on he 12 h in he following monh Daa from head offices of chain sores Head offices of chain sores deliver urnover figures for all heir sores on an Excel shee via Price index for reail sales The price index for reail sales is obained from Saisics Norway s division for price saisics. The index is a NACE four-digi secor level, and is wihou value added ax Populaion Coverage The populaion covers all acive esablishmens 6 in reail rade, excep of moor vehicles and moorcycles (SIC2007:47). All regions of Norway are covered, excep of he Svalbard Archipelago. Sae owned unis and unis owned by he social securiy adminisraion or local or regional governmens are no par of he populaion Srucure in NACE rev. 2 In NACE rev. 2, he Saisical classificaion of economic aciviies in he European Communiy, Division 47 Reail rade, excep of moor vehicles and moorcycles includes he resale (sale wihou ransformaion) of new and used goods mainly o he general public for personal or household consumpion or uilisaion, by shops, deparmen sores, salls, mail-order houses, door-o-door sales persons, hawkers, consumer cooperaives ec. (Eurosa 2008) A NACE hree-digi level, i is srucured as follows 7 : In sores Non-specialised: 47.1 Reail sale in non-specialised sores Specialised: 47.2 Reail sale of food, beverages and obacco in specialised sores 47.3 Reail sale of auomoive fuel in specialised sores 47.4 Reail sale of informaion and communicaion equipmen in specialised sores 47.5 Reail sale of oher household equipmen in specialised sores 47.6 Reail sale of culural and recreaion goods in specialised sores 47.7 Reail sale of oher goods in specialised sores No in sores 47.8 Reail sale via salls and markes 47.9 Reail rade no in sores, salls or markes 4 Turnover: Duiable and duy-free sales income from goods and services as well as rens, commission fees and royalies. Financial revenues are no included. 5 Refer o appendix A for an example of an Inerne-based quesionnaire. 6 Local kind-of-aciviy uni (Esablishmen): An enerprise, or par of an enerprise, ha is locaed in one paricular place and can be idenified geographically. On or from his place economic aciviy is exercised which, wih some excepions, employs one or more persons in one and he same enerprise. 7 Refer o appendix B for a more deailed srucure of Division 47 Reail rade, excep of moor vehicles and moorcycles. Informaion concerning correspondence wih ISIC rev. 4 is also included.

9 The classificaion of he Inernaional Sandard Indusrial Classificaion of all Economic Aciviies of he Unied Naions (ISIC rev. 4) is idenical a hree-digi level, hus ensuring comparabiliy of daa (Eurosa 2008) Annual updae of he populaion and division ino subpopulaions We updae he populaion of esablishmens in Ocober, every year, in connecion wih he roaion of he sample of idenical unis. The populaion s esablishmens are found on Saisics Norway's Regiser of Esablishmens and Enerprises firs file in Ocober (Diagram 3.1). We divide hem ino he subpopulaions of chain sores, idenical unis and newly esablished unis. Chain sores in he populaion are idenified based on which head offices have agreed on delivering urnover figures for all he sores in heir chain. A whole chain of sores can be omied from or included in his subpopulaion every year in Ocober. To divide he remaining esablishmens ino subpopulaions we firs consider hose ha also exised in he second erm (March and April) he las available erm wih fully revised urnover figures of he wholesale and reail sales saisics. Esablishmens ha had urnover in his erm and he same indusrial classificaion (NACE five-digi secor level or subclass) in his erm as in Ocober, are defined as idenical unis. On he conrary, hose unis ha had urnover bu a differen classificaion in he second erm are couned as newly esablished unis. We carry ou qualiy checks on unis ha are classified differenly in Ocober compared o in he second erm and correc heir classificaion, if necessary. The subpopulaion of newly esablished unis also includes esablishmens wihou urnover in he second erm and esablishmens ha did no exis in he second erm of he wholesale and reail sales saisics. Diagram 3.1: Classificaion of he esablishmens in connecion wih he annual roaion of he sample comparison period (C) monh of roaion of he sample FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN Wholesale and Reail Sales Saisics 2 nd erm: Esablishmens in Division 47 (organisaion no., NACE subclass, urnover) mach excluded unis deah unis chain sores Their head office delivers urnover figures for all he sores in he chain idenical unis In he 2 nd erm: urnover > 0, same subclass newly esablished unis In he 2 nd erm: urnover > 0, oher subclass In he 2 nd erm: urnover = 0 Did no exis in he 2 nd erm POPULATION Saisics Norway s Regiser of Esablishmens and Enerprises: Esablishmens in Division 47 (organisaion no., sae of aciviy, NACE subclass, no. of employees) Monhly updae of he populaion We also updae he populaion monhly, based on informaion from head offices of chain sores and he sample of idenical unis. In addiion, we use informaion abou esablishmens in he second erm (March and April) of he wholesale and reail sales saisics, as well as from Saisics Norway s Regiser of Esablishmens and Enerprises firs available file afer he expiraion of he saisical monh.

10 Esablishmens ha cease o exis or become par of a NACE five-digi secor level (subclass) which is irrelevan for he index of reail sales are normally omied from he populaion (Diagram 3.2). Unis ha are esablished by chains ha deliver us urnover figures for all he sores in he chain are included in he populaion of chain sores. Oher new unis are added o he populaion of newly esablished unis (Diagram 3.2) 8. Diagram 3.2: Number of idenical sample unis, newly esablished unis and chain sores in he course of a year Oc 08 Nov 08 Des 08 Jan 09 Feb 09 Mar 09 Apr 09 May 09 Jun 09 Jul 09 Aug 09 Sep 09 sample of idenical unis populaion of newly esablished unis populaion of chain sores Sraificaion We sraify he subpopulaions of idenical unis and newly esablished unis. I is he basis for he roaion of he sample of idenical unis and for he calculaion of he urnover of boh subpopulaions. The esablishmens in each subclass (NACE five-digi secor level) are divided ino four size groups: Sraum 1: 1-4 employees Sraum 2: 5-9 Sraum 3: Sraum 4: 20- Newly esablished unis are sraified each monh, when hey become a par of he populaion. We check wheher here is coherence beween he esablishmens size group and heir urnover in he second erm of he wholesale and reail sales saisics. If no, we correc heir size group. For he populaion of idenical unis, we check his on a yearly basis, in connecion wih he roaion of he sample Sample of idenical unis Number of unis The number of unis in he sample has decreased significanly during he pas decade, from approximaely in 1999 o in This is because we have gradually supplemened urnover figures for he sample wih urnover figures for chain sores repored direcly by head offices from November 1999 (Diagram 3.3). The amoun of chain sores has nearly ripled since hen. 8 In January, he number of newly esablished unis falls due o he implemenaion of he new sandard indusrial classificaion.

11 Diagram 3.3: Number of idenical unis in he sample and chain sores. November 2000 o November Nov 00 Nov 01 Nov 02 Nov 03 Nov 04 Nov 05 Nov 06 Nov 07 Nov 08 Nov 09 populaion of chain sores sample of idenical unis In all, he chain sores and he reail sores in he sample represened abou 40 per cen of all he unis in he populaion a he end of 2009 (Diagram 3.4). Diagram 3.4: Number of chain sores and idenical unis in he sample compared o he oal number of unis in he populaion. Per cen. November 2000 o November % 80 % 60 % 40 % 20 % 0 % Nov 00 Nov 01 Nov 02 Nov 03 Nov 04 Nov 05 Nov 06 Nov 07 Nov 08 Nov 09 populaion of chain sores sample of idenical unis oher idenical unis populaion of newly esablished unis

12 Roaion of he sample The urnover of he subpopulaion of idenical unis is esimaed based on urnover figures from a sample of unis. The sample is roaed annually, in connecion wih he publicaion of he index for Ocober. Below, we describe his process: Sep 1: We specify which unis are o be excluded from he drawing process (he cu-off): Unis wih less han NOK in he second erm of he Wholesale and Reail Sales Saisics. Sep 2: We make up a sraified sample drawing plan, where we specify he required coverage raes for all sraa. The principles for his plan are: The sample in each sraum mus, normally, cover a leas 30 per cen of is populaion s urnover in he second erm of he Wholesale and Reail Sales Saisics. The coverage rae for sraa wih large unis is se higher han for sraa wih small unis. Consequenly, large unis have a bigger chance o be drawn. Sep 3: Before roaion, all unis are sored: based on he amoun of years a uni has been in he sample minus he amoun of years ha i has no been in he sample. The more years a uni has no been par of he sample compared o he amoun of years i has been par of he sample, he bigger he chance ha i will be drawn. based on he amoun of years in a row hey las were or were no par of he sample. The more years in a row a uni las was no par of he sample, he bigger he chance ha i will be drawn. The more years in a row a uni las was par of he sample, he smaller he chance ha i will be drawn. Sep 4: Roaion principles: Unis ha have been par of he sample for four coninuous years are omied. In each sraum, we draw as many unis as necessary o aain he specified coverage rae in he drawing plan. Sep 5: Afer roaion we check wheher he coverage raes specified in he plan are aained, because some sraa could have had oo few unis o draw from. In ha case, we adjus he coverage raes for hose sraa in he drawing plan and draw again Calculaion of he value index Calculaion of he index wih he aid of chaining Afer he implemenaion of he new Sandard Indusrial Classificaion in January 2010, he index has been published wih he average of he indices for he 12 monhs - as index reference period. When calculaing he monhly index, we ake ino consideraion ha he sample is roaed every year in Ocober. For his monh, we collec urnover figures from unis ha will be omied from he sample, as well as from unis ha will become par of he sample. The index for Ocober is calculaed based on urnover figures from he old sample, while he developmen from Ocober o November is calculaed based on urnover figures from he new sample. The index I, a differen NACE-levels, can be wrien as (3.4.1) where I Tˆ Tˆ s = I s for snow < snex

13 Tˆ is he esimaed urnover is he saisical monh s is he index basis period. This is las Ocober. r is he index reference period. This is he year Calculaion of he urnover in a saisical monh The esimaed monhly urnover a NACE wo o four-digi secor level is calculaed by aggregaing he esimaed urnover a NACE five-digi secor level (subclass). The esimaed urnover a his level, in he saisical monh m, akes he form: (3.4.2) ˆ CH ˆ ( ) ( ) ( ) ˆ NE T m = T m + T m T where T CH h h h + h h (m) is he urnover for chain sores in subclass h, in he saisical monh m. ˆ h is he esimaed urnover for idenical unis in subclass h, in he saisical monh m. ˆ h is he esimaed urnover for newly esablished unis in subclass h, in he saisical monh m. T T NE The calculaion of he urnover for hese hree subpopulaions is described below, firs as a diagram (3.5), and hen more deailed per subpopulaion. Diagram 3.5: From daa collecion o index of reail sales chain sores Head offices deliver urnover daa for all he sores in heir chain (full coun).the daa are aggregaed. idenical unis sample Turnover daa are colleced from a sample and inflaed o populaion level wih a raio esimaor. Aggregaed urnover for he whole populaion Index of reail sales newly esablished unis No urnover daa available. Modelbased esimaion Chain sores As he head offices of chain sores repor urnover figures for all he sores in heir chain, heir urnover in subclass h, in he saisical monh m, is calculaed by aggregaing hose figures. I can be wrien as CH CH (3.4.3) T h = Vi, h where V CH CH i poph i, h is he urnover for chain sore i in subclass h, in he saisical monh m. pop CH (m) h is he populaion of chain sores in subclass h, in he saisical monh m.

14 Idenical unis The urnover for he populaion of idenical unis is esimaed based on urnover daa from a sample of idenical unis. A raio esimaor is applied o each sraum o inflae sample daa o populaion level. The raio esimaor uses urnover figures from he second erm (comparison period C) of he wholesale and reail rade saisics (VAT regiser) as auxiliary variables. The raio model is given by: (3.4.4) T ˆ h, s = V i, h, s i sample h, s ( C, m) V i, h, s i pop h, s ( C, m) Vi, h, s i sample h, s ( C, m) ( C) ( C) where Tˆ, s h is he esimaed urnover for idenical unis in subclass h, sraum s, in he saisical monh m. i h ( ) is he urnover for an idenical uni i in subclass h, sraum s, in he saisical monh m. V,, s m V i, h, s i pop h, s ( C, m) Vi, h, s i sample h, s ( C, m) ( C) is he raio beween he populaion s and he sample s urnover for idenical unis in ( C) subclass h, sraum s, in he comparison period C. Nex, we aggregae he urnover figures for he differen sraa s in each subclass h: (3.4.5) T ˆ h = T ˆ h, s s Newly esablished unis As we have no informaion abou he urnover of newly esablished unis, i is esimaed. We use informaion from he same period in he previous year on how much newly esablished unis made compared o idenical unis. In addiion, we use he average urnover for idenical unis in he saisical monh m. The following formula describes he calculaion of he urnover for newly esablished unis in he saisical monh m, subclass h, sraum s, esablished in erm : (3.4.6) Tˆ NE h, s, = N NE h, s, V h, s V V NE h, s, h, s, ( M 6) ( M 6) where Tˆ NE h, s, is he esimaed urnover in he saisical monh m for newly esablished unis in subclass h, sraum s, esablished in erm. is a erm ha is an elemen of C+1, C+2, M. M refers o he erm o which he saisical monh m belongs. N NE h, s, is he number of newly esablished unis in he saisical monh m, in subclass h, sraum s, esablished in erm. V h, s is he average urnover for idenical unis in he saisical monh m, in subclass h, sraum s. NE Vh, s, ( M 6) is he raio beween he average urnover for newly esablished unis and he average Vh, s, ( M 6) urnover for idenical unis in subclass h, sraum s, esablished in erm. I concerns urnover figures from he saisical monh s erm in he previous year (M-6).

15 By aggregaing urnover as follows, we find he oal urnover for newly esablished unis in subclass h: NE (3.4.7) Tˆ h = T ˆ s, NE h, s, 3.5. Deflaion / Volume index Volume indices a NACE four-digi secor level are calculaed by deflaing value indices direcly by means of he price index for reail sales. The indices for value and price are wihou value added ax. A NACE wo and hree-digi secor level he volume indices are calculaed as a weighed sum of he volume indices a he NACE four-digi secor level using value shares in he reference year as weigh.the weighs are updaed monhly Working day and seasonal adjusmen The index is adjused for seasonal variaions applying he X12ARIMA sofware package wih non-fixed seasonal effecs and muliplicaive model. Working day adjused series are calculaed for he urnover indices and he volume indices while seasonal adjused series are calculaed for he volume indices. As a supplemen o seasonal variaions, a new mehod developed by Saisics Norway akes ino accoun he effec of weekdays, fixed holidays as 1s and 17h May, Easer, Penecos, Ascension Day and 1s New Year's Day. 24h o 26h of December is considered as seasonal variaions in his model. The new mehod also akes ino accoun how sales in a period of days before (w1) and afer (w2) he Easer holidays are affeced. Se Diagram 3.6. Diagram 3.6: Esimaed periods before and afer he Easer holidays. w1 days w2 days Normal level Th Fr Sa Su Mo Tu We MTh GFr Sa ESu EMo Tu We Th Fr MTh ; Maundy Thursday GFr ; Good Friday ESu ; Easer Sunday EMo ; Easer Monday The lengh of hese periods is esimaed from he raw daa for each of he hree-digi secor levels. The mehod is also used for Ascension Day and Penecos since hey boh are holidays wih a non-fixed dae.

16 4. Resuls The index of reail sales is compiled, according o Council Regulaion No 1165/98 9 concerning shorerm saisics, as a monhly unadjused and working day adjused value and volume index. In addiion, a seasonally adjused volume index is compued. We calculae indices a he naional level only for all NACE secor levels, bu only for cerain indusries a four- and five-digi secor level. The figures are released on hp:// and in Saisics Norway s saisical daabase Sabank 10, usually on he 29 h or he 30 h of he following monh. Reporing o Eurosa, he saisical office of he European Union, happens simulaneously. From he implemenaion of he new Sandard Indusrial Classificaion (SIC 2007) in 2009, we also publish an index for reail rade ha includes reail sale of auomoive fuel. The new series is compued back o 2000 (Diagram 4.1), while he former series (SIC 2002) has figures back o Diagram 4.1: Seasonally adjused volume index for reail rade. 2005= Reail rade, excep of moor vehicles and moorcycles Reail rade, excep of moor vehicles, moorcycles and auomoive fuel Reail sale in non-specialised sores Reail sale of oher household equipmen in specialised sores Reail sale of oher goods in specialised sores Source: Saisics Norway, Sabank able 07129: Index of reail sales 11 From January 2000 o March 2010, Norway s reail rade volume, excep of moor vehicles, moor cycles and auomoive fuel, rose significanly, by over 45 per cen. I rose a lile more han he reail rade volume ha includes reail sale of auomoive fuel. When comparing he developmen a differen NACE hree-digi secor levels (Diagram 4.1), we see ha he reail sale volume in non-specialised sores grew relaively slowly, bu seadily, during he pas decade. On he conrary, he reail sale volume of oher household equipmen in specialised sores rose more significanly from 2000 o 2007, bu fell considerably in When comparing o he 27 EU member saes average seasonally adjused volume growh for reail rade, excep of moor vehicles and moorcycles, in he pas decade, Norway s growh was considerably sronger, especially during he pas four years (Diagram 4.2). 9 hp://eur-lex.europa.eu/lexuriserv/lexuriserv.do?uri=consleg:1998r1165: :en:pdf 10 hp://sabank.ssb.no/saisikkbanken/?planguage=1 11 hp://

17 Diagram 4.2: Seasonally adjused volume index for reail rade, excep of moor vehicles and moor cycles. 2005= The 27 EU member saes Denmark Sweden Norway Source: Eurosa Comparisons wih oher saisics Index of household consumpion of goods The index of household consumpion of goods 13 is compiled by Saisics Norway s Division for Naional Accouns. This monhly volume index goes back o I describes he developmen in household consumpion, while he index of reail sales measures he developmen in reail sales. Compared o he index of reail sales, he index of household consumpion of goods has a wider selecion of goods. The calculaion of he index of household consumpion of goods is based on informaion from he index of reail sales, plus purchases of cars (iniial regisraion) and consumpion of elecriciy and heaing fuels. This may resul in deviaions in he developmen of he wo indices. 12 hp://epp.eurosa.ec.europa.eu/poral/page/poral/shor_erm_business_saisics/daa/daabase 13 hp://

18 Diagram 5.1: Index of reail sales and index of household consumpion. Seasonally adjused volume index. 2005= Index of reail sales, excep of moor vehicles and moorcycles 120 Index of household consumpion Srucural saisics for wholesale and reail rade and bimonhly wholesale and reail sales saisics The populaion for Division 47 in he srucural business saisics 14 and he bimonhly wholesale and reail sales saisics 15 are he same as he populaion of he index of reail sales, only he bimonhly saisics do no include local kind-of-aciviy unis wih a urnover below NOK While he bimonhly saisics are solely based on daa from he VAT regiser, he srucural business saisics also use informaion from Trading Saemens, Annual Company Repors and survey daa. When convered o indices, urnover figures for Division 47 in he srucural business saisics and he bimonhly urnover saisics can be compared wih he reail sales value index (Diagram 5.2). 14 hp:// 15 hp://

19 Diagram 5.2: Reail rade, excep of moor vehicles, moorcycles and auomoive fuel. Index of value. 2005= Index of reail sales 120 Wholesale and reail rade, srucural saisics Wholesale and reail sales saisics In recen years, he developmen of he value index of reail sales has been more consisen wih he developmen of he srucural business saisics and bimonhly urnover saisics (Diagram 5.2). This is mainly because he chain sores and he reail sores in he sample gradually covered a greaer share of he populaion s urnover. In he beginning of 2010, heir share was approximaely 80 per cen of he urnover of he populaion, while he sample only covered around 40 per cen before we sared collecing urnover figures from chain sores in More up o dae and beer qualiy of daa in Saisics Norway s Regiser of Esablishmens and Enerprises has conribued o he more correc esimaion of he index of reail sales oo. 6. Improvemens 6.1. Mehods of daa collecion Inerne-based reporing Inerne-based reporing was inroduced o he sample of idenical unis in 2002, and i has seadily grown in populariy (Diagram 6.1). In February 2010, over half of he respondens repored urnover figures via he Inerne. In he same period, 30 per cen answered via mail and 13 per cen via phone or fax. On he conrary, hree quarers of all respondens answered via mail (46%), phone or fax (29%) in February 2005.

20 Diagram 6.1: Reporing mehod for sample unis Phone / Fax Inerne Paper Feb 05 Feb 06 Feb 07 Feb 08 Feb 09 Feb 10 Inerne-based reporing conribues o prevening measuremen and processing errors. Arihmeic and logical operaors check he responses for errors and lack of inernal consisency while he quesionnaires are being regisered on he Inerne. No only does his reporing mehod conribue o beer daa qualiy, i also saves regisraion work for he saisical office. From 2011, head offices of chain sores will be able o deliver urnover daa more securely, via an inerne poral Phone-based reporing In order o make he daa collecion process even more efficien, he unis in he sample will be given he opion o repor urnover figures digially o a phone answering machine insead of o personnel a he saisical office. A presen, his soluion is in he planning phase NorSamu - Coordinaed drawing of samples for economic surveys Since Ocober 2009, we are using he applicaion NorSamu o roae he sample of idenical unis of he reail sales index. I has been developed by Saisics Norway o coordinae he drawing of samples for economical surveys. The sysem is designed o assure ransparency, righeousness and predicabiliy in regards o he oal burden of economical surveys on enerprises and esablishmens. The principles are (Zhang 2010): Among similar unis, here mus be a common speed of roaion in all srucural surveys and a common speed of roaion in all surveys for shor erm saisics: o Where possible, a uni shall ake par in a srucural survey for wo years in a row, and for four years in a row in surveys for shor erm saisics. o Where possible, a period of aking par in a saisical survey shall be followed by a period of res. Over ime, he expeced relaion beween he amoun of years of paricipaion and res shall be he same for all similar unis.

21 6.3. ISEE - Inegraed Sysem for Ediing and Esimaion The developmen of generalized sysems for saisical daa processing is an imporan measure underaken a Saisics Norway, in order o achieve he sraegic goals i has se for saisical producion. A basic principle is o simplify, o improve and o re-use he producion processes. The implemenaion of he various generalized sysems aims o provide more efficien use of resources, greaer flexibiliy in producion means, as well as beer qualiy insurance. Inegraed Sysem for Ediing and Esimaion (ISEE) is such a generalized sysem. [...] The key feaure is he complee inegraion of ediing and esimaion processes. For insance, inside he ISEE i is possible o examine insanly he final esimae of a populaion oal due o any changes made o he daa. One is hus in a much beer posiion o implemen he so-called op-down approach o ediing (e.g. Grankvis and Kovar,1997; Grankvis, 1997). However, in is presen form he ISEE has is primary applicaions in processing of business sample surveys, where i is being used in he producion of abou saisics. (Zhang 2009) The ISEE consiss of hree core applicaions, of which wo are currenly being implemened in he producion of he reail sales index. The firs is an applicaion for microdaa ediing, called Dynarev. The second is Srukur, a SAS-based applicaion ha can predic oals and he uncerainy of oals in sample surveys, in addiion o making saisical checks as he basis for revision. The program can calculae several robus variance esimaes, parameer esimaes, an aggregaion of variables wihin sraa, weighs or inflaion facors and prediced values for he saisical variable of unis ouside he sample. Moreover, i can make regression diagnosics ha reveal which unis have a big influence on he esimaes. If desired, he analyses can be presened graphically oo. 7. Summary The index of reail sales one of Saisics Norway s many saisics wihin wholesale and reail rade - describes he monhly developmen in he value and he volume of reail rade, excep of moor vehicles and moorcycles (SIC2007:47). I is regarded as an imporan shor-erm indicaor of consumer demand and, consequenly, used exensively. From 1936 o 1970, he index was produced based on urnover ax daa. I had become of fairly high qualiy when a value added ax replaced he urnover ax and Saisics Norway had o sar producing a reail sales index based on survey daa. The qualiy of he new index was no as good, bu has improved grealy over he years, especially due o he collecion of daa direcly from head offices in he pas decade. Turnover figures from head offices of chains are colleced as a full coun. In addiion, we collec urnover figures from a sample of idenical unis and inflae hem o populaion level in each sraum, wih a raio esimaor. The urnover of newly esablished unis is esimaed based on how much newly esablished unis made compared o idenical unis in he same period in he previous year. Based on he aggregaed urnover for hese hree subpopulaions a a NACE secor level, we can calculae value indices. The index is published wih 2005 as index reference period and las Ocober as basis period. Thus, he sample is roaed annually, wih a periodiciy of four years. A sraified drawing plan is made up for every year, bu unis wih less han NOK in he second erm of he Wholesale and Reail Sales Saisics are excluded from he drawing process. Large esablishmens have a bigger chance o be drawn. Volume indices a NACE four-digi secor level are calculaed by deflaing value indices direcly by means of he price index for reail sales. Working day adjused series are calculaed for he urnover indices and he volume indices while seasonal adjused series are calculaed for he volume indices. In he pas decade, Norway s reail rade volume, excep of moor vehicles and moor cycles, has grown significanly, and a a sronger rae han he average reail rade volume of he EU member saes. As informaion from he index of reail sales is used in he calculaion of he index of household

22 consumpion of goods, he wo can be compared, bu hey may deviae from each oher because of he wider selecion of goods in he las. The annual developmen of he value index of reail sales has, in recen years, been fairly consisen wih he developmen in he srucural and bimonhly saisics, mainly due o he increase in urnover coverage of he sample and he chain sores in he pas decade. Several generalized sysems have been and are currenly being implemened in he producion of he reail sales index, e.g. sysems for he coordinaed drawing of samples, digial daa collecion via he Inerne and he elephone, micro daa ediing and he predicion of oals in sample surveys. The firs menioned sysem conribues o ransparency, righeousness and predicabiliy in he oal burden of economical surveys on enerprises and esablishmens, while he ohers aim a more efficien use of resources, greaer flexibiliy in producion means, as well as beer qualiy insurance. References Eurosa (2008): NACE Rev. 2 Saisical classificaion of economic aciviies in he European Communiy. Eurosa Mehodologies and Workingpapers, hp://epp.eurosa.ec.europa.eu/cache/ity_offpub/ks-ra /en/ks-ra en.pdf Lønø, Krisian, Rusi, Audun and Zhang, Li-Chun (2010): Veiledning il Norsamu: Samordne uvalgsrekking for SSBs økonomiske undersøkelser. Unpublished paper. Saisics Norway Sander, Jan, Solheim, Leiv and Zhang Li-Chun (2008): Bruk av applikasjonen Srukur. Unpublished paper. Zhang, Li-Chun (2009): Working Session on Saisical Daa Ediing: Ideas on ediing of saisical regisers. Saisics Norway. Zhang, Li-Chun (2009): NorSamu Seksjonsseminar 820, Januar PowerPoin slides. Saisics Norway Saisics Norway (1976): Saisisk senralbyrå: 100 år, Samfunnsøkonomiske sudier nr. 28. (hp:// Saisics Norway: Index of reail sales (hp://

23 Appendix A: Example of an Inerne-based quesionnaire for daa collecion from he sample of idenical unis How big was he urnover in March 2010? Excluding VAT VAT Including VAT

24 Appendix B: Deailed srucure of Division 47 Reail rade, excep of moor vehicles and moorcycles in NACE rev. 2 Source: Eurosa (hp://epp.eurosa.ec.europa.eu/cache/ity_offpub/ks-ra /en/ks-ra en.pdf)