What does the survey comprise?
The monthly publication of price changes in consumer goods and services. Price developments are presented as index numbers. The annual rate of change of the CPI is an important indicator of inflation. The annual rate of change is calculated as the percentage change of the consumer price index (CPI) in a particular month compared to the same month in the previous year. Statistics Netherlands (CBS) also calculates the derived CPI. In this series, the effect of changes in the rates of product-related taxes (e.g. VAT and excise duty on alcohol and tobacco) and subsidies is eliminated. The derived CPI answers the question: how would prices have changed, if tax rates had remained the same?
Consumer goods and services purchased by Dutch households.
Consumer goods and services.
Year the survey started
The status of the figures referring to the most recent month is provisional and becomes final when the figures for the next month are published.
How is the survey conducted?
Type of survey
The study is based on monthly price observations among companies and institutions supplying goods and services to consumers.
Statistics Netherlands increasingly makes use of transaction data (e.g. scanner data of supermarkets and other retail chains) and register data. Web scrapers and internet robots that collect prices automatically are also used. In addition, Statistics Netherlands makes use of online surveys and manual price observations via the internet.
Sources and Respondents
Statistics Netherlands collects prices for the CPI by consulting businesses, institutions, associations and public authorities. In this regard, Statistics Netherlands mainly receives aggregated source data that do not contain any personal information. For a number of price measurements, sources coupled to personal data are used:
- Online surveys may be sent to sole proprietorships, for example lawyers or painters;
- For the rental prices, the ‘Basisregistratie Kadaster (BRK)’ and the ‘Eigendomsverhouding vastgoed (koop-huur)’ of the Kadaster are used as source data. In addition, the ‘Inkomensafhankelijke huurverhoging’ of the Tax and Customs Administration is used. For the energy labels and energy indices of dwellings, data is obtained from the Netherlands Enterprise Agency (RVO);
- To measure the price development of child care services, the ‘Kinderopvangtoeslag (WKO)’ of the Tax and Customs Administration is used;
- To measure the price development of nursing homes and home care services, information concerning ‘Eigenbijdrageplichtige zorg’ from the Central Administration Office (CAK) is used.
Every month, a vast number of prices are collected from datasets with transaction prices, register data and web scrapers. In addition, with the help of internet robots, online surveys and manual observation via the internet, approximately 8,500 prices of around 750 different goods and services are observed among nearly three thousand respondents.
Checking and correction methods
Statistics Netherlands checks the completeness and plausibility of incoming data. If data are incomplete or implausible, the respondent is requested to provide additional information. In the case of non-response, the data will be imputed (the necessary information is estimated on the basis of other observations). When products or services in the survey package are changed, a correction is made for any differences in quality between the old and the new version of the product.
WeightingTo measure price changes, a package of goods and services is compiled. A specific weight, i.e. an indication of how important a product or service is in the consumption pattern of Dutch consumers, is assigned to all products and services in the package. Consumer spending data by National Accounts are used to determine the weights of the various goods and services in the CPI. In addition, budget surveys and other external sources are used. The weights of the product groups in the CPI are adjusted and published annually in January.
Quality of the results
Accuracy and representativeness of the CPI are safeguarded by measuring prices of a very large and diverse range of products. Each month, prices are measured at different sales channels throughout the country. However, it is impossible to calculate the degree of accuracy. The accuracy of price changes for a group of products is generally lower than for the total consumption package. The survey response rate exceeds 95 percent. Inaccuracies may arise at several stages in the production process:
- sampling respondents
- selection of representative items and estimation of quality differences
- measurement errors
- determination of weighting factors
- as a result of non-response
Price developments are measured by including prices of products which were surveyed during two consecutive months, preferably at the same retail outlet. Annually, the reference base is adjusted in January. At that moment, a new package of products and services is introduced. The weights in the package are also updated. The data are therefore not the same from one year to the next, but they are comparable. Since the publication of the series 2015=100, the gap between the CPI and the Harmonised Index of Consumer Prices (HICP) has narrowed. Figures prior to 2015 are sometimes based on different definitions. As a result, a different level is attributed to, for example, the weights for insurance services.
Data are checked for internal consistency and completeness. Wherever necessary, action is taken to verify the data. If the data are correct, the price indices are calculated and the results are verified and validated. Validation and calculation are not done by the same individual. If afterwards it appears that incorrect information has been used for the composition of the price index figures, it will be decided on the basis of a protocol whether the results must be revised. If the results are not revised, the price index will be brought to the right level in the next month.