Since governmental measures were imposed due to COVID-19 pandemic, the movement of people was restricted during April 2020 and some outlets (stores) and service establishments were closed, which had a direct or indirect impact on household consumption and thus consumer price indices.

Methodological procedure how to cope with these unexpected fluctuations in the household consumption for the consumer price index (CPI) was based on the existing CPI concept and on a harmonised approach across EU Member States under the current circumstances for the harmonized index of consumer prices (HICP) the methodology of which is decisive also for the national CPI.

HICP calculation related to measures taken due to COVID-19 is governed by the following three principles:¹⁾

• HICP weights stability;
• calculation of price indices covering the whole structure of the European classification of individual consumption according to purpose (ECOICOP);
• minimisation of the number of imputed (grossed up) prices and sub-indices.

The first principle ensures that relative weights of sub-indices used for calculation of an aggregate consumer price index during the year is not changed. This standard practice is a consequence of the Laspeyres chained index concept, which does not take into account relative fluctuations of consumption of individual groups of ECOICOP during the year. Relative weights are updated once in a year (HICP) or once in two years (CPI) in order to gradually reflect only long-term (structural) shifts in the household consumption.

The second principle means that all sub-indices for the whole structure of ECOICOP are calculated, although for the time being no products or almost no products are accessible on the market for some categories. In such cases, when no prices exist for the time being, they are replaced by imputed prices. Sub-indices consisting of imputations and observed prices are aggregated using standard procedures for compilation of aggregate price indices.

The third principle underlines the preference of real prices to imputed prices. It means, for example, to replace price observations (that are not available due to shutdown of selected outlets for a field survey) by prices obtained from other sources, i.e. web pages of outlets (stores), enquiries over the phone or e-mails, and the like. However, here it also has to be taken into consideration whether at least some purchases are made under the given conditions and whether the prices are fully comparable with prices from a standard price survey.  

When solving concrete cases how to calculate detailed price indices in affected groups of consumption (besides the aforementioned general methodological rules of Eurostat for the HICP) the CZSO applied two other principles in case of a potential methodological conflict as follows:

• not to distort a long-term time series of price indices is preferred to precise capture of the month-on-month development of the price level;
• relevance of a detailed index is subordinated to the relevance of higher aggregations, namely of the overall index.

From the point of view of an impact on the measurement of inflation in practice, the April situation can be considered rather a short-term fluctuation of proportions of consumption in some categories of the ECOICOP classification, which usually returns to similar limits as it was before the introduction of government measures. It is important that consumption in the areas that have the biggest influence on short-term (month-on-month) movements of the overall price level, i.e. food and automotive fuel, was not limited and their price movement was captured fully and without technical problems during the data collection. Utilisation of new data sources – scanner data significantly contributed to that, too.

The CZSO also made several experimental calculations of the month-on-month development between March and April 2020 of the overall price index while omitting all or at least a significant part of items, which were available in April in just a limited scope or were not available at all. All these calculations indicate a month-on-month price movement that on the overall price index level differs from the published figure by less than 0.1 p. p., which is acceptable tolerance within the determined CPI methodology.
 

Concrete measures to ensure continuity of time series of selected price indices in April 2020

As for the share in reduction of consumption within the consumer basket, it applied mainly to the following ECOICOP divisions:

Clothing and footwear (03) – a major part of the prices was surveyed on the Internet except for clothing accessories and some services, where prices were surveyed in a standard way. For seasonal clothing and footwear, usual seasonal adjustment methods were used.   

Furnishings, household equipment and routine maintenance of the house (05) – prices in hobby markets, chemist´s (drugstores), and prices of most of the services were surveyed in a standard way. Other prices were surveyed on the Internet and from price lists.

Transport (07) – prices of motor cars, motorcycles, and bicycles and spare parts and accessories thereto were surveyed over the phone, on the Internet, and from price lists. For the price index of air transport, an overall price index of transport services (07.3) was used.

Recreation and culture (09) – Most of the prices were surveyed on the Internet or from price lists, as the case may be. For recreational and cultural services (09.4), prices from the previous month were partially used, namely in cases, in which prices change only slightly (admission fees/charges). For package holidays (9.6), a typical seasonal factor was used for the necessary imputation of the price index pursuant to recommendation of Eurostat.  

Restaurants and hotels (11) – prices were surveyed on the Internet, from price lists, and over the phone; for catering services (11.1) also from sale of food to take away (through little windows). For accommodation services (11.2), prices for March were partially used.

Miscellaneous goods and services (12) – a major part of the prices was surveyed in a standard way; the missing prices, especially for hairdressing salons and personal grooming establishments (12.1.1), were surveyed on the Internet and from price lists.

¹⁾ Link to relevant documents of Eurostat: https://ec.europa.eu/eurostat/web/hicp/methodology