Content
About the statistics
Definitions
-
Name and topic
-
Name: Commodity price index for the industrial sector (discontinued)
Topic: Prices and price indices
-
Responsible division
-
Division for Price Statistics
-
Definitions of the main concepts and variables
-
A chained Young formula is used in the production of the index. This is a version of a Laspeyres formula. A Laspeyres formula is characterised by constant weights for a given time period. A chained Laspeyres index is an index made up by several Laspeyres indices with various weights.
The Norwegian market is a joint description of all goods produced and purchased in Norway. Price regarding sales to the Norwegian market is defined ab fabric (ex works) and should refer to the sales price out of the factory on the 15th of the month.
The export market consists of all goods produced and sold to foreign customers: exported goods. Price regarding sales to the export market is defined free on board (fob), at the Norwegian border.
The elementary level is the lowest level in which indices are computed, the level below the lowest level in which weights are used. In the PPI the elementary level is the HS product.
Imputed value is the estimated value where observations are missing.
The weight share is the share of the HS product in the total index.
-
Standard classifications
-
CPA - (Statistical classification of products by activity) .
SIC2007 - (Standard Industrial Classification næringsgruppering) . The Norwegian version of Eurostat's NACE rev. 2 .
Administrative information
-
Regional level
-
National level only.
-
Frequency and timeliness
-
Monthly. Published on the 10th of every month.
-
International reporting
-
Price statistics for the home and export markets are reported to EUROSTAT.
-
Microdata
-
Data concerning establishments, products and prices are stored in Oracle databases, and are subject to confidentiality.
Background
-
Background and purpose
-
The VPPI is an important part of a system for short-term indicators made in order to monitor the Norwegian economy. Together with the Producer Price Index (PPI) and the Price-index of first-hand domestic sales (PIF) the VPPI measures the price development in production and imports of goods. Together these three statistics cover the price development in three markets: Domestic production, exports and imports. The prices used to calculate the three statistics are collected in the same survey.
The purpose of the VPPI is to measure the price development of first hand sales of products to the Norwegian market, from Norwegian production and export. Unlike the PPI the VPPI may be revised retrospectively when more data or improved methods are available. After the new standard industrial classification (SIC07) was introduced in 2009, there has been no needs to revise the VPPI. So the VPPI series are until further notice identical with the PPI series. For details on changes made in methodology and dissemination, see the links Revision 2001 and New classification in the right hand pane
The index is commissioned and financed by The Norwegian Ministry of Finance.
-
Users and applications
-
The statistics are used by the public sector and the financial industry. The statistics are used in the production of the national accounts in Statistics Norway.
-
Coherence with other statistics
-
Together with the Producer Price Index (PPI) and the Price-index of first-hand domestic sales (PIF) the VPPI measures the price development in production and imports of goods. Together these three statistics cover the price development in three markets: Domestic production, exports and imports.
The VPPI is closely connected with the producer price index (PPI) . The main difference is that the VPPI may be revised retrospectively depending on availability of more data or improved methods, while the PPI is final when it is first published. This means the VPPI is not suitable for use in contract escalation.
The building cost index and the quarterly wholesale price index use some data from the Producer Price Index.
Other short term statistics from Statistics Norway describe related variables in the Norwegian manufacturing industries, such as the index of production , the statistics on new orders and statistics on stocks , and use the same industry standard as the VPPI.
-
Legal authority
-
Statistics Act §§2-1, 2-2, 2-3 (compulsory fines) Statistics Act §§2-1, 3-2 (administrative data-processing systems) Statistics Act §§ 2-1, 2-2, 3-2 (forms/registers)
-
EEA reference
-
Council Regulation (EC) no. 1165/98 of 19. May 1998 concerning short-term statistics. Commission Regulation 586/2001 Commission and Council Regulation 1158/2005 Commission and Council Regulation 1983/2005 Commission Regulation 1503/2006 Commission Regulation 656/2007 Commission Regulation 472/2008
Production
-
Population
-
All commodities and some services produced by companies within oil and gas extraction (NACE06, see 4.2 for explanation), mining (05, 07 and 08), mining support serviece facilities (09), mamnufacturing (10 - 33) and energy supply (35). From 2012 the index for energy supply includes both production and distribution of electricity. Mining support service facilities (NACE 09) were included in the VPPI from January 2012.
The following industries are not covered: mining of uranium and thorium ores (07.21), processing of nuclear fuel (24.46), manufacture of weapons and ammunition (25.4), building of ships and boats (30.1), manufacturing of aeroplanes and spacecraft (30.3) and manufacture of military fighting machinery (30.4).
The commodities whose prices are measured, are mainly from establishments belonging to these industries, and the establishments are sampled from Statistics Norway's Register of Establishments and Enterprises. Parts of the primary industries and wholesalers are also included in order to cover first-hand domestic transactions.
-
Data sources and sampling
-
The main data source is the price survey which together cover the PPI, the PIF and the VPPI. Prices are mainly surveyed using questionnaires, which is available electronically and as a paper form to the establishments specified above. In addition, data are collected from various sales organisations, the Division for Foreign Trade, the quarterly electric energy prices, and in some cases from international sources (spot prices).
The sample consists of about 630 commodity groups (CPA products, see 4.3). These groups are split into various HS products (see 4.2). The products are selected based on foreign trade data and the PRODCOM survey, where the HS products are selected to cover the CPA products. The number of HS products is usually about 3500 total to cover the domestic, export and import market. A HS product is included in the sample because of the weight in at least one of the three markets. This means that the HS is not necessarily included in the index calculations in all three markets. CPA commodities are selected in cooperation with the Division for National Accounts and cover important commodities in the national accounts. Furthermore, the HS products are selected using data from foreign trade and Norwegian manufacturing statistics, and cover the chosen CPA commodities. The range of CPA commodities and HS products is adjusted according to changes in the national accounts and foreign trade - even so, the annual changes are modest. The changes are carried out to introduce new products into the sample.
A selection of establishments is used in the monthly price collection. In cooperation with these companies a range of goods are chosen to cover the selected HS products (and CPA commodities). The sample consists of about 1300 establishments and roughly 5000 products. Companies with 100 employees or more are included in the sample on a permanent basis (cut-off principle). The sample is updated continuously with regard to liquidation, foreclosures etc.
The table below show the number of price observations by industry, including the domestic markets and the export markets. The numbers vary some from month to month depending on the response rates.
Industry 1 Price observations as per
January 2012
Industry 1 Price observations as per
January 2012
01 33 19 49 02 3 20 247 03 24 21 26 05 2 22 149 06 3 23 154 07 11 24 134 08 72 25 255 09 63 26 189 10 409 27 133 11 53 28 320 12 3 29 74 13 52 30 12 14 66 31 79 15 27 32 111 16 187 35 2 17 110 Sum 3052 1 Standard industrial classification (SIC2007).
-
Collection of data, editing and estimations
-
The three statistics PPI, PIF and VPPI are all collected in the same survey. The survey is based on the same data, questionnaires and electronic reporting as the producer price index (PPI) . The questionnaires are sent out on the 10th of every month and are to be returned around the 17th. Figures from foreign trade are used if they are found to be an adequate indicator of the price development of a good. Prices of electric energy prices are collected from NordPool . In addition some prices are collected from international sources such as Bureau of Labour Statistics , London Spot Markets og London Metal Exchange .
All forms go through a manual check where administrative information is checked. Electonic data is collected using the online solution of Statistics Norway for reporting. The electronic version contains automated controls, in example for identifying large price changes. A computerised control then checks for punching errors, duplicates and observations with large price changes from the previous month. The checks are done for each market separately. Respondents are contacted if one or several of their commodities show large changes in price and no explanation is given in the questionnaire.
Non-response: Non-response is imputed mechanically based on the price development at a higher aggregation level. The firm is contacted in case of non-response, and if prices are not available price data are imputed. In this procedure, observations which are important to higher level indices are prioritised. CPA products which are not covered by actual price obeservations are imputed using a similar algorithm. Which aggregate level is used depends on the number of actual price observations.
On the micro level the a Jevons formula is used, which is the geometric average of products within the same HS product (see 4.3). Higher level indices are calculated using a chained Young formula. For the VPPI, total indices on the HS level (domestic market and exports) are weighed together. Higher level indices are are first calculated as weighted averages to the CPA six digit products, then to four digits CPA and then according to the hierarchical structure of SIC07. The weights are updated annually, and are calculated on production and export values from the national account's latest final figures. To keep the weight as up-to-date as possible aggregated levels are adjusted by using the quarterly results from the national accounts. This way weights always relate to the year previous to each index period (calendar year). The reference year is 2000 (2000=100).
The VPPI may be revised if further data or improved methods are available.
-
Confidentiality
-
The data collected from the firms will be used in accordance to the rules specified by the Norwegian Data Inspectorate. Collected data are subject to confidentiality specified in the Statistics act, §2-4.
-
Comparability over time and space
-
The VPPI has been compiled since 2000, and published series from 1977 to December 2000 have been chained and recalculated based on data from the producer price index (PPI) . Series grouped according to NACE rev 2 exist as of January 2000 on a 2-digit level, and some on a 3-digit level. For details on changes in methodology and standards, see the links Revision 2001 and New classification in the left hand pane .
The series are published for the domestic market alone, as well as for the domestic market and export market together.
Accuracy and reliability
-
Sources of error and uncertainty
-
A correct answer to the questionnaire requires that the respondent has understood the definitions of the main variables. To reduce the effects of possible errors each product is specified in text as well as recommended unit/quantity. Furthermore, the respondent is encouraged to put the establishment's own product description and/or product code on the questionnaire. The price reported in the previous month is also pre-printed. In addition, a guideline containing important definitions is provided. This system gives a good overview for the respondent, but there is still a possibility for errors.
Sometimes the respondents report the same price over longer periods of time. In many cases this is correct because of costs due to changing prices (menu costs). However, in some cases this may be a source of error due to lack of motivation by the respondents. In extreme cases of constant prices the respondents are contacted.
In cases where the respondent's selected product is out of production or out of stock, and the respondent is reporting a different product, this should be specified on the form. If this is not done, product changes may be registered as price changes. Normally, these kinds of errors are identified.
The paper forms are optically read, automatically interpreted, verified, and the data are stored electronically. The routine for optical reading has turned out to be very reliable in other statistical areas. It is also part of the routine to produce electronic images of the forms. This makes it easier for the auditor to compare the optically read data with the information given on the form. It is also possible for revisers to look up the respondent in the electronic form on the Internet (IDUN).
At the time of deadline, the response rate varies between 85 and 90 per cent. After reminders have been sent out and before publishing the response rate normally increases to about 97 per cent. In cases of total non-response, meaning the whole form is missing, the establishments are contacted. Missing prices are imputed. Important respondents are given priority when re-contacting for missing forms.
In order to secure a high level of accuracy at a low cost, emphasis has been put on covering large units that dominate the population.
Variance: Is not calculated.
Skewness: The products in the population are updated continuously. The sample of products for each establishment is selected in a joint effort between the establishment and Statistics Norway. The respondent is asked to continuously update the sample. This means replacing all discontinued products, products that have become less important for the establishment's turnover, and products where the price development no longer represents the overall price development for the establishment. If the respondent does not update the product sample, this may represent a source of skewness. No calculations of the sample skewness have been made, but based on experience these kinds of errors may go in both directions. Extreme cases are normally identified through checks. Also, by contacting the establishments the effects of this type of error/skewness are reduced.
Furthermore, commodity-group levels are checked in connection with the annual weight changes. Here all commodity groups (with weight) are checked against the sample. In commodity groups with poor coverage, efforts are taken to improve the coverage.
The sample of establishments is changed continuously as old respondents are replaced by new ones. The relatively large proportion of large units (establishments) represents a potential source of skewness. A large number of small units in an industry may cause skewness if their price development is different from the larger units.
Experience show that a share of the units (normally small) in the population is misplaced as regards industrial code and commodity code (HS). This is due to changes in the classification, and/or insufficient or misleading information at a certain time. Information about the commodities in the sample and in the population is improving over time. No attempts have been made to estimate the extent and significance of these types of errors, but it is assumed to be small.
The quality of the products will also change over time. This might lead to an overestimation or underestimation of the price development. In the cases where we know there has been a change in quality, this is adjusted by ordinary methods of quality adjustment.
For computers a so-called hedonic method is used in order to deal with the frequent quality changes within this commodity group. This method is based on the assumption that the price of a commodity is decided by its quality characteristics. It is therefore essential to establish a descriptive relationship between price and quality characteristics, i.e. price as a function of these characteristics.
As mentioned before, sometimes prices are constant over a longer period of time. This might be because the respondents report list prices instead of transaction prices. In cases where for instance demand or the competitive conditions change - thereby changing the level of discount - the developments in the index may over- or underestimate the price development.
Effects of changes in relative prices may represent another non-sampling error. The relative price between different products changes over time, often because of changes in demand. This may affect the producer's/importer's prices and price development. To identify such substitution effects the weights, and the sample of products, should be revised frequently. The size of this skewness is not estimated, but as both the sample of products and the weights are subject to an annual revision, this skewness is assumed to be very small.
The product sample is mentioned in 5.3.