About the statistics
Name and topic
Name: Price level for consumer goods and services
Topic: Prices and price indices
Division for Price Statistics
Definitions of the main concepts and variables
Purchasing Power Parity
Purchasing power parities are price level indicators expressing the price level in a given country at a given time, relative to the price level in one or more other countries. The purchasing power parity between two countries, A and B, expresses the number of units of country B's currency one would need in country B in order to maintain the purchasing power of one unit of country A’s currency in country A. If a given product costs 100 Norwegian kroner in Norway and 10 euros in Germany, the purchasing power parity between the two countries, for this product, equals 10/100 = 0.10 with Norway as the base country, or 100/10 = 10 with Germany as the base. This means we need 0.10 euros in Germany in order to maintain the purchasing power of one krone in Norway, or 10 kroner in Norway in order to maintain the purchasing power of one euro in Germany. Purchasing power parities can be computed for individual products or for aggregates, such as GDP or actual individual consumption. In the calculation of purchasing power parities for aggregates, parities are weighted with expenditure shares from national accounts.
Price level adjustment
Price level adjustment is understood in this context as the conversion of monetary aggregates expressed in national currencies and at national price levels into a common price level and a common, technical currency using PPPs.
Purchasing power standard
The Purchasing power standard (PPS) is the name given by Eurostat to the artificial currency unit in which countries’ national accounts aggregates are expressed when adjusted for price level differences. Figures in PPS are comparable, spatial volume figures because the price component of each individual country has been replaced by a common price component for all countries. The PPS in Eurostat’s terminology corresponds to the term "international dollar" used by the OECD and the World Bank.
Relative price levels and price level indices
The relative price level is understood as the price level of one country relative to the price level of one or several other countries at a given point in time. The relative price level is often expressed as a price level index (PLI). PLIs are derived by dividing the PPP by the respective nominal exchange rate, and usually multiplied by 100.
Countries of participating countries are required to price consumer goods and services, capital goods and general government services.
Consumer goods and services: Purchasers’ prices for a selection of consumer products and services. VAT, non-refundable taxes and possibly subsidies are included in the prices. From 2016, discount prices are included together with the normal prices in the survey. Exceptions are discounts which are not available for all customers.
For housing services, data from the Rental Market Survey is used in order to estimate the price level of both rented and owner-occupied housing. The price of owner-occupied housing is estimated on the basis of owners’ equivalent rent of primary residence principles.
Government services consumed by private households’ expenditures have no market prices. This pertains to education and hospital services. The costs of educational services are estimated by Eurostat using a unit price per pupil or student on the basis of existing education statistics. For hospital services quasi-prices are estimated for a range of specified services such as heart surgeries, prostate surgeries and caesarians.
Weights and auxiliary data
Individual products priced within the framework of the ECP are aggregated up to more extensive consumer groups. From the lowest aggregate level (the "basic heading" level) and upwards, weighting is based on expenditure shares from national accounts.
Aside from the prices and adjustment factors, participating countries have to provide expenditure weights at basic heading level, exchange rates and mid-year resident population figures as well as estimates of GDP and its main sub-aggregates.
Comparability, representativity and equi-representativity
These concepts must be seen in context. The comparison of goods and services across countries is in reality a comparison based on technical specifications. However, products need to be not only comparable in technical terms, but also representative of the consumption pattern in individual countries. It is often necessary to compromise between comparability and representativity. For example, international branded goods will typically be identical across countries and thus have a high degree of comparability, but at the same time they are not always representative of the consumption pattern.
Ideally, the product sample should be equally representative for all the participating countries ("equi-representativity"). A basket of goods is defined as equally representative in different countries when it provides equal satisfaction or utility. Failure to comply with the requirement of equi-representativity can produce a bias in the results because a representative goods basket can be assumed to have lower prices than a non-representative goods basket.
The European System of National and Regional Accounts, ESA2010, is the international guideline used for the production of national accounts in the European Economic Area (EEA) countries. ESA2010 defines the classification of consumption and investment used in the ECP.
A basic heading is the lowest level of aggregation used in the computation of purchasing power parities. Below the level of the basic heading, we find the individual products of the product sample. In the aggregation process, numerical weights based on a detailed breakdown of national accounts expenditures are applied to each basic heading.
Transitivity and multilaterality
Transitivity is the property whereby the PPP between any two countries can be estimated through a third country, yielding the same result as a direct comparison. For example, in the case of the three countries A, B and C, the ratio of the PPP between A and B and the PPP between C and B is equal to the PPP between countries A and C. This approach results in a matrix with multilateral parities, something which allows missing price data from one or several countries to be replaced by estimated figures. Multilaterality implies that any change in the input data of any country will have an impact not only on that country, but on all countries in the comparison.
The analytical categories are the main aggregates, the expenditure categories, the expenditure groups and expenditure classes for which the results of the comparison are published. Examples of aggregated published results are actual individual consumption, clothing and footwear, and investment in building and construction. For consumption, analytical categories mainly follow the national accounts concept of "actual individual consumption". In addition to goods and services bought and paid for by households, actual individual consumption includes public services consumed individually (for example, health and education services), as well as services provided by non-profit institutions serving households.
The ECP follows the classification of GDP as defined in the ESA. Detailed classifications based on the following main classifications have been prepared specifically for the ECP:
*European Classification of Individual Consumption According to Purpose (ECOICOP)
*Classification of the Purposes of Non-Profit Institutions Serving Households (COPNI)
*Classification of the Functions of Government (COFOG)
*Statistical Classification of Products by Activity (CPA)
National coverage in each of the 37 participant countries.
Frequency and timeliness
Annual data.The reference period is the actual calendar year. Preliminary results are available in June, 6 months after the end of the reference year. These figures are revised in December, 12, 24 and 36 months after the end of the reference year. At that stage, the results are normally considered final in the sense that there will be no further updates.
The European Comparison Programme (ECP) is an international statistical project in which the National Statistical Offices of the participating countries are responsible for data collection in their respective countries, while the Statistical Office of the European Commission (Eurostat) and the Organisation for Economic Cooperation and Development (OECD) are responsible for calculating the results. The validation of input data is an interactive process between the participating countries, Eurostat and OECD. Participating countries provide a formal approval of their data before Eurostat and OECD calculate and publish the results.
Price data in anonymised form are stored in Eurostat’s production database. Statistics Norway stores micro data and information on sampling units and populations according to the relevant safety regulations.
Background and purpose
The purpose of the European Comparison Programme is to produce indicators of the relative price levels across participating countries. A central concept is the purchasing power parity (PPP), cf. "definitions".
The results of the ECP are used primarily in order to adjust national accounts aggregates for price level differences across countries, so that only spatial volumes are taken into consideration in international comparisons. For example, in comparing GDP per capita (which is composed of a price and a volume component), using the nominal exchange rate would lead to an overvaluation for countries with high price levels, while the opposite effect would occur for countries with low price levels.
The work is coordinated by Eurostat and the OECD. Eurostat coordinates the work in 37 European countries, including Norway, while OECD coordinates a similar exercise in the non-European OECD member states. The results from Eurostat and OECD are included in the International Comparison Program ( ICP ) - which is carried out by the World Bank on behalf of the United Nations.
The first set of results which included data from Norway were released in 1980, followed by updated releases in 1985, 1990 and 1993. From 1995 onwards, results for Norway have been published by Eurostat on an annual basis. Every third year, the OECD prepares results for all OECD countries, as well as estimates based on extrapolation for the years in between.
Users and applications
The main institutional users are the European Commission and the International Monetary Fund (IMF). The basis for the distribution of funds to the EU’s poorer regions is the comparable, price level adjusted income level in the various regions. Purchasing power parities are an important input into these calculations, and thus the European Commission contributes substantially to the financing of the programme. In the IMF, price level adjusted GDP is one of several factors that determine the level of member states’ financial contribution.
In general terms, PPPs can be used as currency converters, as an alternative to the nominal exchange rate. PPPs are preferable to exchange rates if, for example, one wants to compute how much a given amount in country A is worth in country B, under the assumption that the same purchasing power should be maintained. PPPs are applied for this purpose in businesses, government administration, research, poverty analyses and the media.
Purchasing power parities are used not only for price level adjustments, but also in analyses of relative price levels.
Access to results prior to publication
While Statistics Norway is in charge of the collection and validation of the Norwegian input data, the results for all participating countries are first published by Eurostat. These results are then accessible to all users. Immediately after Eurostat's release, usually on the following day, Statistics Norway re-publishes the results on its website.
Equal treatment of users
No external users have access to the statistics and analyses before they are published and accessible simultaneously for all users on ssb.no at 8 am. Prior to this, a minimum of three months' advance notice is given inthe Statistics Release Calendar. This is one of Statistics Norway’s key principles for ensuring that all users are treated equally.
Coherence with other statistics
The primary purpose of the ECP is to compute price level adjustment factors for national accounts. The basic classification used in the survey is thus the national accounts expenditures breakdown as defined in the European Standard of National Accounts (ESA2010).
In a population approach, the purchasing power parity for a given country in year t+1 should equal the purchasing power parity in year t multiplied by the price change in that country relative to a reference or base country. This provides a link to the CPI, although the ECP focuses on spatial rather than temporal comparability. However, it is not possible to reconstruct either statistic on the basis of the other.
The Statistics Act of June 16, 1989 number 54, §§2-2 and 2-3
The population is defined by the expenditure side of the National Accounts (cf. ESA2010) as final consumption in private households and non-profit organizations, government consumption and gross capital formation. This means that goods and services that are used for these purposes are covered by the price surveys.
Data sources and sampling
Prices of goods and services for private consumption are collected from establishments in surveys carried out simultaneously in all participating countries. The price collection for private consumption is divided into six surveys which are carried out in a three year cycle. The product sample for each survey is decided upon by the participating countries in close cooperation, and can vary considerably from one survey to the next.
Product sampling: The total number of products (goods and services) is around 2500. The sample is based on information on products and markets which is collected from businesses and business associations in each country. The sampling criterias are comparability, representativity and equi-representativity (see definitions).
Establishment sampling: The sample of establishments is drawn from the Central Register of Establishments and Enterprises at Statistics Norway and is the same sample used in the consumer price index (CPI). Samples are drawn on the basis of the establishments’ turnover. In addition, it is often necessary to supplement the sample on an ad hoc basis (purposive sampling).
Weights: Individual products priced within the framework of the ECP are aggregated up to more aggregated consumer groups. From the lowest aggregate level (the "basic heading" level) and upwards, weighting is based on expenditure shares from national accounts.
For housing services, data from the Rental market survey and housing stock data are used.
Government services consumed by private households’ have no market prices. This pertains to education and hospital services. The costs of educational services are estimated by Eurostat using a unit price per pupil or student on the basis of existing education statistics. For hospital services quasi-prices are estimated for a range of specified services such as heart surgeries, prostate surgeries and caesarians.
Collection of data, editing and estimations
Prices are mainly collected through shop visits, telephone and internet. In addition, for some groups of consumer goods, Statistics Norway receives electronic data from important establishments and business chains. Rental prices based on Statistics Norway’s Rental Market Survey and Population and Housing Census are gathered from the national accounts. For the rest of private consumption data, the price collection is undertaken continuously over a three year period such that one third of private consumption is covered in a given year. For the years in between, survey results are extrapolated using CPI data as extrapolation factors.
Data for house rents, in conformity with public sector salaries, are collected annually. Prices of capital goods are collected every other year, while prices of construction projects are collected annually.
Weights based on national accounts expenditure shares are updated yearly.
Data editing is carried out in three phases. Editing starts with data control using standardized extreme control at the product level, in order to identify obvious mistakes. The next editing round is the comparison of the Norwegian data against other participating countries in the survey- in total 37 countries. After extensive control, Eurostat calculates the final results for all participating countries. A final check and a approval round is executed in all the countries before the final results are released.
Most countries collect capital city prices and thus are required to provide spatial adjustment factors with which to convert these prices to national average prices. All countries are also required to provide temporal adjustments factors to ensure that the prices represent the annual average. Spatial adjustments factors and temporal adjustment factors should be supplied for each basic heading. For seasonal products, special temporal adjustment factors should be supplied at product level.
The computation method used is the so-called Éltetö-Köves-Szulc (EKS) approach. This approach starts off with all participating countries’ average prices for each product within a basic heading. For each pair of countries, all possible bilateral price relatives are computed, first with the weights of the base country (Laspeyres-type indices), then with the weights of the reference country (Paasche-type indices) The geometric average of the Laspeyres and Paasche indices (Fisher-type indices) are made transitive by computing an unweighted geometric average of the Fisher indices for each pair of countries. These EKS indices are rebased to the average value across all countries. These rebased EKS indices are the final PPPs for the basic heading. PPPs at higher levels of aggregation are computed in the same way as the basic heading PPPs, except that numerical weights from National Accounts are applied in the aggregation process, and that arithmetic averages are used instead of geometric averages.
For further information on the calculation of PPPs, see the Eurostat-OECD methodological manual on PPPs.
Monthly data from the Consumer Price Indices (CPI) is the source used in the seasonal adjustment work.
Data from respondents are used in accordance with the requirements of The Norwegian Data Protection Authority. The information is confidential (Statistics Act, § 2-4) and is stored and ultimately destroyed in accordance with strict security guidelines.
The production of purchasing power parities is in its nature an example of international statistical cooperation. The input data must therefore necessarily be made available to the participating countries’ national statistical institutes, Eurostat and the OECD. However, data is anonymised before they are shared with the other participants in the programme. External users can apply to Eurostat or the OECD in order to access these data for research purposes, provided that the data are not published.
Comparability over time and space
The purchasing power parity survey is primarily a mapping of the participating countries’ relative price level at a given point in time. The primary focus is therefore a comparison between countries in a certain year, while comparability over time comes second. Time series created as a result of this work must be interpreted with caution, and even more so at lower levels of aggregation.
Price level adjustment of national accounts aggregates in current, domestic prices results in a time series in current, common prices. Therefore, one cannot calculate real growth on the basis of such a time series.
Sources of error and uncertainty
Measurement- and processing errors
Measurement errors are errors caused by insufficient or inappropriate measurement techniques. In the ECP, such errors can occur if, for example, a respondent quotes the price of a product which does not conform to the specifications with regard to quantity or quality. The same challenge occurs during shop visits, since price collectors sometimes have insufficient expertise to judge whether a given product fits the specifications.
Processing errors occur during the regular processing of data. Such errors are not very frequent in our context.
Failure to respond
In questionnaire based surveys, failure to respond has typically varied between 20 and 50 percent, with partial non-response rates even higher. Establishments which fail to respond are replaced if deemed necessary. This is done for products with substantial variation in the price material. Shop visits and the use of electronic data instead of questionnaires have contributed to reducing the non-response rate.
Errors in the product sample can be due to non-compliance with the criteria of comparability, representativity or equi-representativity, cf. above.
The outlet sample is drawn from the Central Register of Establishments and Enterprises. Incorrect classification and failure to update this register can sometimes be a source of error.
While the ECP is coordinated by Eurostat and the OECD, the National Statistical Offices of the participating countries are responsible for the implementation of the programme, in particular the data collection process. Although there is a strong emphasis of harmonization, differences in interpretation and priorities in each country can be a source of error. Such errors are often related to insufficient comparability, representativity or equi-representativity.
Uncertain and preliminary weight estimates
The expenditure shares from national accounts, which are used as weights in the aggregation process, are estimates at a very low level of aggregation. This creates considerable uncertainty. Furthermore, these estimates are often preliminary at the time when they are incorporated into the calculations.
Incoherence of product sample using extrapolation factors
The price level of goods and services for household consumption is surveyed once every third year. For the years in between, the prices from the most recent survey are extrapolated with temporal price indices for the corresponding product group. These indices are taken from the consumer price index. Extrapolation can represent a source of error if the product sample of the CPI is not the same as in the ECP.
Revision are changes of published figures (for example publishing of final data to replace preliminary figures). The PPP manual describes in 13.26 rules for revision. The PPP Regulation allows for final calculations to be repeated if there has been a mistake made by either a participating country, Eurostat or the OECD. A mistake is defined as "a use of incorrect basic information or an inappropriate application of a calculation procedure". For the recalculation to take place, the mistake must be discovered within three months of publication of the final results. For the results of the recalculation to be published and to replace the results initially disseminated, the mistake must cause at least a change of 0.5 per cent in the real GDP per capita of at least one participating country.