Statistiska centralbyrån (Statistics Sweden)

ESA/NUP/PP

Marcus Fridén

Statistiska centralbyrån (Statistics Sweden)

Solna strandväg 86

S-171 54 Solna Sweden

The Producer Price Index contain the following main series:
— Hemmamarknadsprisindex (Producer Price index, domestic sales),
— Exportprisindex (Export Price Index)

The two series above are combined into:
— Producentprisindex (Producer Price Index)

The indices are designed primarily for deflation purposes in order to calculate production and foreign trade values at constant prices as well as the Industrial Production Index, Index of Service Production and Production Value Index. Other important uses are for general economic and inflation analysis and for indexation of business agreements.

SPIN 2015 (Swedish version of CPA 2.1)

Products according to CPA 2.1 related to Sections B to E36, with the exception of Division 09.

Reporting units are selected among producers/exporters within the respective product group. Weights are calculated to reflect the overall production for domestic sales and exports.

A cut-off on 10 million SEK per reporting unit is applied. That is about 10 % of the size of the overall populatoíon that is excluded before the sample is drawn.

Prices are measured at basic prices, i.e. ex works for domestic sales and f.o.b. for export peices. Prices should be transaction prices, less discounts, of specific items invoiced during the month. However, prices according to contracts as well as list prices adjusted for rebates are also accepted. VAT and excise duties are excluded.

We have no planned changes in information collection.

Average monthly prices for specified typical transactions. Other concepts than monthly average are sometimtes accepted.

The observation units are individual transactions for the corresponding product offer, divided into the domestic and export market. A product offer is the combination of enterprise and product that is to be priced.

The target population is defined as all the transactions related to sales at the production stage, divided into domestic and export markets, of products under product groups in Sections B to E36 according to CPA 2.1.

The frame population is based on other surveys at Statistics Sweden: Production of commodities and industrial services (Prodcom), Structural Business Statistics (SBS) and Foreign trade in goods (FTG). When the sample is made the frame comes from the most recent full year.

The geographical area covered is national level only; Sweden.

Historical indicators are available as from 1968. All series are available from 1990 according to the present classification, SPIN 2015 (CPA 2.1).

Reference year is 2015=100. Base period is December of previous year. Chain-linking is applied.

Indices.

Month. Average price of the whole month is reported.

Regulation (EU) 2019/2152 of the European Parliament and of the Council of 27 November 2019 on European business statistics, repealing 10 legal acts in the field of business statistics.

Commission implementing Regulation 2020/1197

Official Statistics Act of 2001 (Lagen (2001:99) om den officiella statistiken) supplemented by a regulation on Official Statistics (Förordningen (2001:100) om den officiella statistiken). Under the ordinance Statistics Sweden issues specific directions (Statistiska centralbyråns föreskrifter (SCB FS 2021:39) om uppgifter till statistik avseende prisindex i producent- och importled, varor).

Reporting units are obliged to provide information. Non-respondents may be fined.

ECB

OECD

Eurostat

The law on confidentiality (Offentlighets och sekretesslagen (2009:400)) is applied. Statistics Sweden’s CBM on Statistical confidentiality is available in “Statistisk röjandekontroll”, Statistics Sweden, September 2001. The publication may be obtained from SCB, Publikationstjänsten, S-701 89 Örebro (E-mail:publ@scb.se). Current practices do not follow the CBM but are similar.

Confidential data are suppressed. For indices to be published, at least three responding units are needed. None can have more than 60% of the turnover, and two of them cannot have more than 90 % of the turnover.

The release calendar is available on the web site as from November for the following calendar year.

https://www.scb.se/en/finding-statistics/publishing-calendar/

Data is released simultaneously to all interested parties by electronic media.

There is no prior access to data for public administration (or Government) other than the statistical office.

Data is transmitted to Eurostat immediately following the national release.

Monthly

News release and tables for the Swedish PPI at https://www.scb.se/en/finding-statistics/statistics-by-subject-area/prices-and-consumption/producer-and-import-price-index/producer-and-import-price-index/ The news releases cover home sales price index, export price index and producer price index.

The indicators are released on Statistics Sweden’s website http://www.scb.se. The relevant subject area is “Priser och konsumtion” (“Prices and consumption” in the English version of the site).
Sweden’s Statistical database and Statistical News are available on the website.

The data could be find in the Statistics database http://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__PR__PR0301/?rxid=c5d1d713-06bf-40ed-bb4c-66c1e117aee5

For research purposes only.

Data are sent to Eurostat to be released as national data.

Methodological descriptions are available at the web-site:
— http://www.scb.se/pr0301/ “Om statistiken” and “Dokumentation” (in Swedish);
— http://www.scb.se/pr0301-en/

Methods are mostly derived from the Producer Price Index Manual released by IMF, https://www.imf.org/external/pubs/ft/ppi/2010/manual/ppi.pdf

Documentation on quality management are avalible on : http://www.scb.se/en_/Finding-statistics/Statistics-by-subject-area/Prices-and-Consumption/Producer-and-import-price-index/Producer-and-import-price-index/ 

Quality of the statistics (Quality report)

Various aspects of quality are evaluated and improved according to annual action plans. However, there are no regular indicators of quality other than the response rate.

Various model assumptions and measurement errors are considered to be the largest sources of uncertainty that contribute most to overall uncertainty. Next, measuring the sample instead of the whole population is considered to be the second largest sources of uncertainty. Besides the sample uncertainty, it is not possible to quantify the uncertainty linked to the other sources of uncertainty.

Data is published approximately 25 days after the reference period.

Missing observations (non-response usually around 4-6 per cent) are imputed by nearby aggregates.

The use of international standards assures geographical comparisons. 

The main user is National accounts for deflation.

Eurostat use the indices for national data.

Companies use the indices for long term contract regulation.

Economic analysts use the indices for forecasting the CPI and the GDP.

User satisfaction level is not measured.

The STS requirements and national needs are fullfilled.

We expect there is no major bias in our data. However, there is a level of uncertainty.

The following areas of uncertainty are examined:

Coverage: Not assessed as a big error source.

Sampling: Represents a non trivial uncertainty.

Non-response: A minor uncertainty.

Measurement: Assessed to be the largest contributor to uncertainty for PPI.

Data processing: A non-significant uncertainty.

Model assupmtions: The second largest source of uncertainty.

A PPS-method is used for sampling where large objects are drawn with certainty. Approximately 4.200 prices are collected monthly from around 2.000 enterprices.  If we consider an enterprise to be covered when we have one measurment in the strata our coverage is around 65 %.

A cut-off 10 million SEK per frame element is applied in each strata. For PPI in total this means that about 10 % of the population, turnover wise, cannot be sampled.

Smaller companies are assumed to be price takers, i.e. their price movement cannot be different than the companies sampled within the strata.

Variances and Standard errors are calculated for allocation purposes.

Impact of over-coverage and under-coverage: Our frames are two years old which means both over-coverage and under-coverage exist. A study on this is ongoing since 2022 but has not been done before.

Impact of survey instrument: Allmost all respondents submit data via the standard web solution used by Statistics Sweden. The response rate is high.

Impact of unit and item non-response: An ordinary month the weighted non response rate is approximately 5-6 %. These observations are imputed with an average price movement of similar products per default.

Impact of data editing, coding and imputation: Quality adjustements is a very challenging task and probably have an important impact. During an ordinary month the change of specifications are 1-3%.

Impact of specific models used in estimation: Model pricing is very seldom used in PPI.

Publication is at 08:00 a.m. according to an advance calendar, and occurs normally around the 25th of the month following the reference month.

1. Questionnaires are sent at the end of the month in order to reach the respondent on the 1st working day following the end of the reference month.
2. Deadline for reply is the 10th of the month following the reference month.
4. A mail reminder is sent around the 11th of the month following the reference month and a second mail reminder could be sent around 17th of the month.
5. The survey is closed around the 21st of the month following the reference month.
6. Publication is around the 25th of the month following the reference month.

Releases on time.

ESS definitions are applied in the whole country.

Chain-linking is applied which makes it possible to change base year without having a break in time series.

When new, or updated, classifications are taken in to use, back-casting would have to be carried out to be able to extend the time series back in time.

Domestic producers are defined as those included in the PRODCOM survey, i.e. PPI and PRODCOM should have the same population. Domestic producers that export products comes from the Foreign Trade in Goods Survey (i.e. Intrastat and Extrastat combined). Exporters that are not domestic producers are not part of the population of exporters.

The definitions of producers and exporters are the same as in PRODCOM and Foreign Trade in Goods. Domestic producers are defined as producers within the geographical  boundaries of Sweden. Export is defined as the border-crossing of a product, not the change in ownership principle.

Indexes are aggregated from lowest level to the top. Weights are summable from bottom to the top. All levels of aggregation are therfore coherent.

Chain-linking is used on all levels of aggregations.

Respondents burden is estimated as 17 min per respondent and month. There are around 2.000 respondents giving a total burden of around 570 hours per month.  

The cost for Statistics Sweden is appoximately 900 hours per month (including data collection, validation, calculations and publishing).

Revisions are only made if a substantial difference is detected. A substantial difference is defined as a difference of more than 0.1 on the overall PPI index.

If an error is conducted by Statistics Sweden we make an external revision. If wrong data is submitted by respondents we just make an internal revision and try to adjust the external indices to correct level as soon as possible without changing published data. The same revision policy is applied to STS data released nationally and transmitted to Eurostat

The first release is considered final. Therefore revisions should not appear.
Major methodological changes are announced in advance.

No revision for the year 2022, MAR/MR=0

Statistical survey.

The frame used for sampling and derivation of weights is based on the results of the PRODCOM survey and the foreign trade in goods statistics (EXTRASTAT and INTRASTAT).

Sample surveys of domestic market output and export market output, respectively. A frame element is a combination of CN 8-digit category and producer/exporter. The samples are made through a two stage process:
1. First, frame elements are sampled in each strata;
2. Second, in cooperation with the respondent specific items are selected from the sampled enterprises.

Frame elements are stratified into stratas according to SPIN 2015. A size criterion is used to identify the most important frame elements in terms of value. In the 2nd stage the respondents are asked to specify “typical transactions” and to report the price of a specific item in such transactions. If possible, the specification should allow monthly reports of the current price of the selected item.

All SPIN 2015 5-digit level above approximately 45 million EUR should be in the sample. The present sample contains approximately 2.000 enterprises reporting 4.200 items.

For the purpose of probability sampling, approximately each 50th million EUR transacted are sampled by the PPS method.

The aim is to renew the whole sample each 5 years, with one fifth renewed each year.

The purpose of renewing the sample is assuring that representative products are measured and that the response burden is distributed as fair as possible.

Monthly.

Hemmamarknadsprisindex (Home Sales Price Index) and Exportprisindex (Export Price Index).

No changes made to questionnaires during 2021. Currently there are plan to make come changes to questionnaires during 2023, due to new internal policies surrounding the validation process for the whole Statistics Sweden.

Figures are validated continuously during each month. Price changes that varies +/- 10 % from the previous month is validated. Large contributions to the overall PPI also being validated.

For non-response and non-transactions the price of the previous month is imputed by the average price change of similiar products.

Grossing-up is treated within the weighting procedure.

The monthly index numbers are chain indices calculated from index links (base December previous year). The weights used to calculate the current aggregate index links are revised as from January each year in order to reflect a recent (year y-2) composition of output/import (at December y-1 prices).
The index links are weighted averages of price ratios, calculated from prices collected for a sample of items. The producer/importer and a specification define an item. The sample of items is revised as from January each year.
For changes in the specification the December (base month) price is adjusted in order to account for an estimated value of the quality change. Various methods are used for this estimation.
When an item is discontinued the price change will be according to a similar item, the average price change of the remaining items in the category or the relative change of a similar category.

The weights are derived from PRODCOM and external trade statistics (EXTRASTAT and INTRASTAT). In order to calculate the weights for “Hemmamarknadsprisindex” (Producer Price index, home sales) export values are subtracted from PRODCOM data at the 8-digit CN-level. Both PRODCOM and external trade statistics use CN for the classification of products.
For products within sections A, D and E complementary sources are used.

Prices quoted in currencies other than the national one, SEK, are converted into the national currency using the customs conversion rate of currencies for the reference period.

Sub-indices are compiled according to the STS Regulation. Sub-indices for (and within) SPIN 2015, Section A are also calculated, as well as groups 38 and 39.

Data are not seasonally adjusted.

Not available.