Pulse Survey now running Five years on, we're measuring the lasting impact of COVID-19 on our lives in our latest short Pulse Survey. CSO Pulse Surveys are anonymous and open to all. #CSOTakePart
Due to COVID-19 restrictions during April 2020, there were unprecedented changes in household consumption patterns. It was estimated that households, on average, were unable to consume 25.6% of the goods and services in the CPI basket of goods and services. These mainly consisted of the following items included in the CPI:
In addition to this, due to the closure of non-essential retail outlets, household consumption for many other goods and services was significantly reduced.
Impact on CPI data collection
Prices are collected by the CSO for the purposes of the CPI and HICP by means of direct pricing in retail outlets and surveys by office-based staff (* see more detail below). Survey-based price collection in April 2020 was completed successfully by office-based staff, with minor exceptions, using the usual data collection methods. However, due to the restrictions on movement imposed during the COVID-19 pandemic, the CSO did not send price collectors into retail outlets to collect information for the CPI in April. As a result, price collection for April proved significantly more difficult than previous months.
The CSO implemented the following contingency measures to collect CPI price data for April 2020:
CPI and HICP are fixed weight indices
Price changes for the goods and services in the CPI and HICP are aggregated using weights that reflect household consumption patterns in the previous year. For example, food currently has a weight of almost 9% in the CPI basket of goods and services, which means that 9% of household consumption expenditure was spent on food. The weights for both the CPI and HICP are updated at the beginning of each year and are held fixed throughout the year. This approach is necessary to ensure that the indices only reflect price changes and not changes in the quantities of products consumed. Therefore, the aggregation weights for the CPI and HICP have not been changed in April 2020 due to any impact of COVID-19 on household expenditure patterns.
Harmonised approach across EU
Given that the aggregation weights must remain fixed, it is necessary to impute the price change for product groups where (i) household consumption has ceased and transactions did not take place or (ii) it is not possible to collect prices. To ensure a harmonised approach to imputation across the EU, Eurostat published a guidance note on the compilation of the HICP in the context of the COVID-19 crisis. The Eurostat guidance was prepared in consultation with National Statistical Institutes (NSIs) including the CSO, and the CSO has implemented this guidance for the CPI and HICP.
Imputation for April 2020
In total, 27.1% of the CPI basket required imputation for April because of the COVID-19 pandemic. This comprised 25.6% of the basket where consumption had ceased and 1.4% where it was not possible to collect prices. Table 1 provides a summary of data collection and imputation for April 2020 by ECOICOP division[1].
The main imputation method used was “nearest aggregate estimation”, where the price development of a similar available product or service was used to estimate price change. For airfares, hotels and package holidays, the seasonal nature of these items was taken into consideration when imputing a price change. For other product groups such as theatres, cinemas and hairdressers, prices were carried forward from March 2020 to estimate prices for April 2020.
Price change for imputed items
It is possible to compare the monthly rate of change for (1) the items in the CPI basket where price change was imputed (27.1% of the basket) and (2) the items in the basket where no imputation took place (72.9% of the basket) i.e. where CSO was able to collect price information. The overall monthly change in the CPI in April 2020 was -0.4%. The monthly change for the items where price information was collected was -0.7%. The monthly change for the items that were imputed was +0.5%, the imputed increase for airfares, hotels and package holidays reflects their seasonal nature.
Table 1: Summary of CPI data collection and imputation for April 2020 by ECOICOP Division
ECOICOP Division |
Weight in CPI (%)
|
Of which: |
How prices were obtained in April 2020 |
|
Imputed - no consumption (%) |
Imputed – no availability of price data (%) |
|||
01. Food and non-alcoholic beverages |
10.2 |
0.0 |
0.0 |
|
02. Alcoholic beverages and tobacco |
4.6 |
0.0 |
0.0 |
|
03. Clothing and footwear |
4.6 |
0.0 |
0.0 |
Prices sourced from internet websites.
|
04. Housing, water, electricity, gas |
14.1 |
0.0 |
0.3 |
Prices sourced from regular sectoral surveys.
|
05. Furnishings, household equipment |
5.0 |
0.0 |
0.1 |
Prices sourced from internet websites.
|
06. Health |
3.2 |
0.0 |
0.0 |
Prices sourced from regular sectoral surveys
|
07. Transport |
14.3 |
1.8 |
0.0 |
|
08. Communications |
2.6 |
0.0 |
0.0 |
Prices sourced from regular sectoral surveys
|
09. Recreation and culture |
7.6 |
3.1 |
0.3 |
|
10. Education |
1.8 |
0.1 |
0.0 |
Prices sourced from regular sectoral surveys
|
11. Restaurants and hotels |
20.5 |
18.2 |
0.6 |
|
12. Miscellaneous goods and services |
11.6 |
2.4 |
0.2 |
|
Total |
100.0 |
25.6 |
1.4 |
|
* Note on price collection for the CPI
There are two price collection methods used by the CPI. Direct price collection is used for 478 of the 615 item headings in the CPI basket of goods and services. Approximately 50,000 prices are collected monthly by 80 price collectors in retail outlets throughout the country. Central price collection is used for the remaining 137 item headings. Prices are collected centrally by CSO staff with minimal field work. In practice, this involves the use of postal, e-mail and telephone enquiries along with internet price collection. Approximately 3,000 prices are obtained using this method. In terms of basket weights (proportion of total household expenditure accounted for by each item in the basket), each price collection method accounts for about 50% of the basket. Therefore, both collection methods are equally important to the continued statistical quality of the CPI for the duration of COVID-19.
[1] ECOICOP is the Classification of Individual Consumption by Purpose Adapted to the Needs of Harmonised Indices of Consumer Prices (2000).
Learn about our data and confidentiality safeguards, and the steps we take to produce statistics that can be trusted by all.