II
Ralf Hein, 05.05.1999 A
The International Comparison Program
2003 – 2005
Regional Coordinators’ Workshop
Luxembourg, March 24 – 28, 2003
Guidelines and Analytical Tables for Private Household Consumption Estimates
Item 9 and 10 of the agenda
Eurostat Directorate B
Non-Financial National Accounts Activities with the Accession Countries
Activity A1: Private Household Consumption
The project ideas and documentation are the property of Eurostat. Their use in the ICP framework is an in-kind contribution by Eurostat. Publications of any kind or reproductions of the documentation or their adaptation to regional needs have to mention the source.
Technical contact person in Eurostat: Ralf Hein, Unit B1 ()
1st Task Force meeting
Luxembourg, 15-16 April 1999
BECH-Building, A3 /139
Analytical Tables for the Project on Private Household Consumption
Author: Ralf Hein
II
Ralf Hein, 05.05.1999 A
Private Household Consumption: Background and Guidelines
for the Completion of the Analytical Tables
A Background
At the first experts’ meeting in Vienna (November 1998) the following Eurostat criteria and requirements for exhaustive and reliable PHC estimates were agreed:
Activity A1: Project Criteria and Requirements
· exhaustiveness and verification;
· the reliability of basic data and calculation methods;
· the estimation of a really independent PHC figure (independent estimates should be made for each commodity in as much detail as possible and from as many sources as possible);
· the use of a framework for systematic estimation, which shows how survey data is transformed into NA data;
· conceptual compliance with ESA95.
On this basis, it was agreed to use a systematic tabular approach for the project work. PHC estimates should be compiled using the “best” source from the alternatives available (which will be different for individual COICOP items), using a “bottom-up” approach. Individual COICOP items should be estimated in a systematic way (which clearly identified the various steps and adjustments). Comparison of alternative sources and validation of the PHC estimates are vital.
The countries should complete a set of tables, which provide a structure and framework for the project work. The tables will be used in the experts’ work with the countries and provide a sufficient basis for the experts’ reports. The completion of the following five tables is proposed:
Table 1 Conversion of HBS data for NA purposes (by COICOP Divisions and Groups)
Table 1A Conversion of HBS data for NA purposes (by COICOP Divisions and Groups): Adjustments for definitions and concepts
Table 2 Conversion of retail sales data for NA purposes (by COICOP Divisions and Groups)
Table 2A Conversion of retail sales data for NA purposes (by COICOP Divisions and Groups): Adjustments for definitions and concepts
Table 3 Conversion of data from other sources than HBS and Retail Trade for NA purposes (by COICOP Divisions and Groups)
Table 3A Conversion of data from other sources than HBS and Retail Trade for NA purposes (by COICOP Divisions and Groups): Adjustments for definitions and concepts
Table 4 Commodity Flow estimates (by CPA)
Table 4A Transformation of Commodity Flow estimates (by CPA) and to COICOP Divisions and Groups
Table 5 Summary table - final estimate of PHC
When completing the tables, the countries should use the following guidelines. Where necessary, the tables should be adapted to the country specific situation.
After discussing tables and guidelines at the first TF meeting, any further necessary explanations and guidelines will be added.
It is recommended that the tables are completed for the last year for which figures are final.
B General guideline for the completion of the tables
All standard aggregate tables of GDP for the expenditure approach include PHC on a national basis (expenditure of resident households). However, the structure of PHC itself is presented in standard tables using a domestic concept (expenditure of resident and non-resident households on the domestic territory) by purpose and by type of expenditures (at current and at constant prices). The domestic aggregate is then adjusted for direct purchases of residents abroad and of non-residents so that it corresponds to the national concept. This suggests that the framework for the systematic estimation of PHC should be built up, commodity by commodity, using a domestic concept.
The level of detail should be chosen by each AC according to the level of detail used for their PHC estimation and with respect to the general requirement to choose a “best source” for individual COICOP items.
In general, for each of the individual COICOP items, at least two independent estimates should be obtained and shown in the summary table. This means, it is not necessary to provide estimates for all COICOP items in all tables.
Where special adjustment methods are used for the conversion of basic data into NA figures which do not feature in the existing columns of the tables, each AC should identify these special adjustments by inserting new columns (or lines) in the tables.
Also, if it is not possible to provide information at the detailed level of COICOP items shown in the tables, additional lines for certain groups of COICOP items and/or for the total can be inserted in the tables.
This could be the case, for example, where no breakdown of the tourist expenditure is available. However, in this case, each AC should explain how they compare HBS figures (according to the national concept) with results from other sources (according to the domestic concept).
In general, the commodity breakdown is important, and even rough estimates could significantly improve the process of cross-checking and the use of different data sources. Such checking and verification procedures, which are assisted by the “systematic tabular approach” adopted here, are important if the project criteria and requirements listed in Section A are to be satisfied.
At the first Task Force meeting, the tabular approach and the proposed set of tables will be discussed and, where necessary, modified or improved.
C Guideline for the completion of individual tables
Table 1 Conversion of HBS data for NA purposes (by COICOP Divisions and Groups)
Column 1: Insert the raw HBS results grossed up for to the total population living in households (in general, this would be the average expenditure per household and year multiplied by the number of households).
HBS includes, usually, data about income in kind, such as consumption of own produced goods or wages and salaries (W&S) in kind. In order to ensure complete and reliable PHC estimates and make them comparable with other sources, for example retail trade figures, data about these items should be separately available. Therefore, table 1A (columns 2 and 3) asks for the relevant data. If the figures are available, column 1 should not include consumption of own produced goods or wages and salaries (W&S) in kind. Otherwise, column 1 would cover these items (or some of them), and table 1A would show all adjustments needed for coverage and definitions.
Column 2: This column should show the adjustments made to the figures in column 1 because of differential non-response.
Column 3: The purpose of this column is to make adjustments for the expenditure of those persons who live in institutions, rather than in households. The most important groups of persons living in institutions are (it is important to carefully check the coverage of the HBS):
- those living in residential and nursing homes;
- students accommodated in „halls of residence“;
- longer term residents in hotels, hostels and boarding houses;
- military personnel occupying houses on military bases;
- those living in religious orders (e.g. monasteries, convents);
- long stay hospital patients;
- those living in prisons.
Column 4: For a number of items, there are differences in treatment between the HBS and the NA. This column shows the total of the relevant adjustments made. A special table 1A is foreseen to show these adjustments in detail.
Column 5: Use this column to show consumption of illegally goods and services if you make separate adjustments for such.
Column 6: This shows the final HBS based figures including all adjustments to reach exhaustive and ESA95 confirm NA adjustments.
Column 7 and Column 8: The inclusion or exclusion of resident and non-resident expenditures is relevant for NA purposes depending on whether the domestic or national concept is used. Such estimates are also needed in order to make a proper comparison with retail sales or production data.
Table 1A Conversion of HBS data for NA purposes (by COICOP Divisions and Groups)
Adjustments for definitions and concepts (national concept)
Column 1: This is to show the total of adjustments made to HBS data for definitions and concepts. This figure is to be transferred to column 4 of table 1.
Column 2: The production of households own production includes agricultural and food products produced for own consumption by farmers and also the production in private gardens. Other goods produced by households for their own consumption should also be included if this type of production is significant. Finally, it includes any other consumption of own produced goods by unincorporated units.
Column 3: Insert all goods and services received as W&S in kind by employees. If HBS includes W&S in kind in the total expenditure without showing them separately, these items will be covered in column 1 of table 1. In this case, column 3 would only show all adjustments needed for coverage and definitions.
It is important to ensure a complete coverage of all W&S in kind - for all units and in compliance with ESA95 rules (see ESA95, 4.04 to 4.07). Differences between HBS and ESA95 can appear with respect to the coverage (identification and definition) of W&S in kind and their valuation.
W&S in kind are discussed in a paper on income in kind written for the Candidate Countries (R. Clare, April 1998). It is important to distinguish between
- goods and services produced by the employer for the provision as W&S in kind
. main production, e.g. coal or free train or railway tickets
. secondary production including the provision of sports, recreation or holiday facilities for employees and their families, free or cheap crèches for the children of employees
- goods and services purchased or financed by the employer to provide as W&S in kind:
. meals and drinks, including those when travelling on business
. housing or accommodation services
. uniforms or other forms of special clothing which employees choose to wear frequently outside the workplace as well as at work
. the private use of business cars
. the provision of sports, recreation or holiday facilities for employees and their families, free or cheap crèches for the children of employees.
The ESA95 requires these items to be valued at basic prices when produced by the employer, and at purchasers’ prices when purchased by the employer. When provided free, the whole value of the provided goods and services is included in W&S in kind. When provided at reduced prices, the value is given by the difference between the whole value and the amount paid by the employee.
Column 4: Insert the total of all other adjustments made for compliance with NA rules given in detail in columns 5 to 8.
Column 5: This is foreseen for special adjustments such as imputed rents, insurance services (according to a special estimation of the insurance service charge), clothing and food in the armed forces.
Column 6: It is important to ensure that HBS is recording the expenditure on all goods and services, including items destined to be gifts (and not including financial transfers in expenditure).
However, charity and gifts in kind made by the rest of the world to households, less charity and gifts in kind made by households to the rest of the world are to be included in PHC and included in this column. The correspondence to export and import figures should be ensured.
Column 7: It is important to ensure the borderline between services provided to households by government and taxes (see ESA95 3.77d, 4.79d and 4.80d).
Column 8: It is important to identify all differences in definitions and concepts between basic data and NA. Examples are:
- expenditure on goods under a hire purchase (should be recorded in full at the time the goods are delivered)
- second hand goods (should be recorded by net purchases)
- purchases of equipment needed for work (it should be ensured that these purchases could be separated and excluded from PHC if necessary)
- the distinction between PHC on the one side and IC or GFCF on the other side. For example, ESA95, 3.77b refers to
. expenditure by households owning unincorporated enterprises, when incurred for business purposes (vehicles, furniture, fuel etc.)
. expenditure that owner occupiers incur on decoration or maintenance and repair or the dwelling not typically carried out by tenants (IC)
. the purchase of dwellings (GFCF)
. the expenditure on valuables (GFCF)
A more detailed explanation of the conversion of HBS data is given in the paper „Conversion of HBS data for NA purposes and Validation of NA estimates of Household Final Consumption (Richard Clare, 12.01.1998)
Table 2 Conversion of retail sales data for NA purposes (by COICOP Divisions and Groups)
Column 1: Insert the raw data on retail sales as provided by the branch statistics. For all further adjustments it is necessary to have a very clear picture about the coverage of retail sales data. Important items are, for example:
- Do the retail sales include
. road-side and country farm purchases,
. purchases from street traders and market stalls,
. purchases from motorway shops or from petrol stations (which may be covered by inquiries/surveys to the motor trade, rather than retailers),
. sales in duty free shops?
- “Retail sales” could also include purchases direct from wholesalers or from manufactures (which may be covered by inquiries/surveys to wholesalers or manufacturers);
- Retail sales will certainly include IC by businesses additionally to final consumption;
- Retail sales from retailers/wholesalers/manufacturers may include purchases by NPISH;
- Retail sales by consumers can include purchases from NPISH;
- Retail sales include purchases of second-hand goods.
Column 2: This column should show all adjustments for non-observed retail sales made to the raw retail trade figures (additionally to adjustments, the branch statisticians „usually“ do and which are included in column 1).