Working Group meeting Statistics on Living Conditions / HBS June Eurostat-Luxembourg. BECH Building, Room Quetelet

Transcription

1 EUROPEAN COMMISSION EUROSTAT Directorate F: Social statistics Unit F-4: Income and living conditions; Quality of life Doc. LC-ILC/160/16/EN Working Group meeting Statistics on Living Conditions / HBS 8-10 June 2016 Eurostat-Luxembourg BECH Building, Room Quetelet Item 7.2 Principal HBS methodological documents, transmission recommendations and follow-up of the HBS 2015 data collection

2 INTRODUCTION During the Statistics on Living Conditions Working Group meeting June, 2015 Eurostat introduced the reference documents to be used for the 2015 HBS wave and described the main changes between the 2010 and 2015 rounds. Methodological documentation was presented, together with a series of recommendations regarding common definitions of the main variables to be used in the 2015 round. The Working Group was informed about the proposed timetable for the data collection and transmission to Eurostat. The purpose of this document is to present the planned timetable for the data collection exercise concerning the reference year 2015, as well as to serve as a short reference document containing the most important information about this data collection exercise. Members of the WG are invited to take note of the details in the document.

3 1. HBS METHODOLOGY 1.1. HBS 2015 documentation library The principal methodological recommendations for the HBS 2015 wave are contained in the following documents: (a) Description of the data transmission for Household Budget Survey (HBS) for the Reference Year Version 3: This document contains the proposed variable list to be transmitted to Eurostat for the HBS 2015 wave. (b) Household Budget Surveys in the EU: Methodology and recommendations for harmonisation : This document contains a comprehensive methodological description of the HBS. However the COICOP-HBS classification and the variable list included in it are those used for the HBS 1999 wave and not those now recommended for the round (c) The COICOP 2013 (Classification of Individual Consumption by Purpose) with explanations 3. Please note that the COICOP categories Life insurance and FISIM: Financial intermediation services indirectly measured are not measured by the HBS. (d) Imputed rent for dwelling services 4 : This document analyses and proposes various methods for the calculation of imputed rent for dwelling services depending on the circumstances of each country. (d) Quality report of the Household Budget Surveys' Description of the data transmission for Household Budget Survey (HBS) for the Reference Year Version 3 from 17 th May CIRCABC Household Budget Survey (HBS) user group, Library, Methodology_1999, 3 RAMON /ramon/nomenclatures/index.cfm?targeturl=lst_nom_dtl&strnom=coicop_ 5&StrLanguageCode=EN&IntPcKey=&StrLayoutCode=HIERARCHIC 4 HBS/158/2005 available in CIRCABC, Household Budget Survey (HBS) user group, Library, Working groups, Working group Quality report of the Household Budget Surveys' 2010 /documents/54431/ /lc142-15en_hbs_2010_quality_report_ver2+july+2015.pdf/fc3c8aca-c456-49ed-85e4-757d f

4 1.2 Standard nomenclatures and classifications During the Working Group meeting Statistics on Living Conditions June 2014 Eurostat proposed to use the most recent versions of the standard nomenclatures and classifications wherever applicable. The following table presents the most important classifications for the HBS: CLASSIFICATION CONCEPT TO BE BROKEN DOWN BY THE PROPOSED CLASSIFICATIONS COICOP 5-DIGIT 2013 Consumption expenditure of households NUTS 2013 (1- & 2-digit level ) Territorial units and regions ISCED-2011 Education level ISCO08 Occupation ISO 3166 Codes for the Country names NACE rev 2 Economic Sector in Employment Detailed information about these nomenclatures may be found on Eurostat s classification server RAMON The main changes compared to the previous HBS wave For the HBS 2015 wave Eurostat made as few changes as possible compared to the HBS 2010 wave. The main change is the need to use the COICOP 2013 (Classification of Individual Consumption by Purpose, at five-digit level) and the ISCED In addition, 4 variables concerning the Reference person that were dropped for 2010 have now been restored (HC03, HC04, HC05, HC05.1). Finally, ME0988 has been dropped. 1.4 Expenditures Household's consumption expenditure will be supplied broken down by the COICOP 2013 (Classification of Individual Consumption by Purpose at five-digit level). The amounts reported to Eurostat should be - in national currency (NAC) - annualised - un-weighted 6 RAMON /ramon/index.cfm?targeturl=dsp_pub_welc

5 1.5 Incomes In countries where it is not possible to obtain very detailed income data, the minimum objective is to provide Eurostat with the net disposable income of the household. As in the previous HBS 2010 wave, net disposable income is calculated using the total monetary net income (HH09.5) plus the three non-monetary income variables (HH01.2, HH02.3 and HH03.2). When the source of income does not concern any individual, but the household as a whole, it is located in the household record. Consequently, the sum of the individual incomes does not necessarily equal the household income. The amounts reported to Eurostat should be annualised (temporal weighting); the weighting of the sample should be provided separately by Member States (HA10) and applied by Eurostat. 1.6 Imputed Rent According to the European System of Accounts 2010, which is the reference for the HBS, the purchase of a dwelling as such, is regarded primarily as capital formation (investment) and not consumption expenditure. However, the ownership of a dwelling is considered to produce a service - a shelter -, which is actually consumed over time by the households. As a consequence, ESA requires the estimation of the price of the shelter, by imputation of a rent, since no monetary transaction is involved. This imputed rent is part of household consumption expenditure. So, for the HBS to be consistent with the ESA principles, it has been recommended to exclude the acquisition of dwellings, whereas the consumption of the service of the dwelling should be included. Different methods can be used in order to estimate the imputed rent. The choice of a method generally depends on the size and the structure of the national rental housing market: - self-assessment - stratification - regression-based approaches. - user-cost method. ESA also takes into account dwellings which are rent free or have a reduced rent. In the HBS 2010 wave, three countries have not imputed any rent for the use of owneroccupied dwellings as household main residence. For the HBS 2015 wave, all the countries should provide the imputed rent variable, as per ESA guidelines and as previously agreed for the HBS 2005 wave. This is even more possible since this variable is mandatory for EU-SILC (EU Statistics on Income and Living Conditions). Priority is to be given to stratification or regression-based approaches over selfassessment and user-cost method.

6 2. MICRO-DATA FILES TRANSMISSION-FORMAT The expected transmission-formats for the HBS micro-data to Eurostat, are: Transmission File format: SAS or TSV file format may be used, with the following suggested names: hh_basic.sas7bdat or hh_basic.tsv - basic variables at household level hm_basic.sas7bdat or hm_basic.tsv - basic variables at household member level derived.sas7bdat or derived.tsv - derived variables hh_food.sas7bdat or hh_food_tsv.txt - food-quantity variables Format of file contents: The columns should contain the variables (one variable per column) and one row should refer to one household. In the case of text files, the first row of the file will contain the names of the variables and the second row will contain for Numeric Variables and "CHARACTER" for the others. Since SAS does not allow dots within identifiers, dots/underscores will be removed from the variable names proposed in this list, when imported into the Eurostat HBS IT System (e.g. the variable HH09.9 will be identified as HH099 in SAS). Format for transmission of aggregated tables (Tabular Data) is described in an annex to the document Description of the data transmission for Household Budget Survey (HBS) for the Reference Year However, this is subject to change since the EuroBase Team wishes to standardize the design of all the HBS domain tables. Once these changes will be known, this corresponding annex will be updated. The treatment of missing values in the expenditure data The HBS expenditure data are liable to under-reporting errors, especially for sensitive items such as Alcohol, Tobacco, Narcotics and Prostitution etc. It is important in the micro-data files to make clear the distinction between the expenditure items the households declined to report and those for which zero expenditure was reported, during the reference period. Eurostat recommends that a '0' is to be used for any expenditure categories for which the Household reported zero expenditure during the reference period, and a '.' ( a dot ) for categories that the Household declined to provide any value for ( i.e. the value is missing ).

7 3. DATA COLLECTION MANAGEMENT The data cycle foreseen by Eurostat for the HBS 2015 wave is the following (this is also summarised in Annex 3). This timing has been elaborated based on the majority of countries which will be carrying out their surveys during the calendar year 2015 as reference year. Eurostat invites each country to provide their best estimates for the timing of all the phases of this data cycle that are under their responsibility. 3.1 Data collection and transmission Data collection by the NSI: As a general rule surveys were scheduled to be carried out during the calendar year Data processing by the NSI: From January 2016 until December Data transmission to Eurostat from June 2016 until December 2016: Data reception and data processing by Eurostat will be spread over a period of about 6 or 7 months. For data transmission it is compulsory to use Eurostat's secure single point of entry edamis, (electronic Data files Administration and Management Information System). The EDAMIS web portal may be found at: Delivery of essential metadata describing the main characteristics of the National HBS survey) to Eurostat from June 2016 until December 2016: Essential metadata is required for data processing by Eurostat. This metadata will be delivered at the same time of data transmission. Eurostat will supply suitable templates to the NSI in order to facilitate this task. 3.2 Data validation, processing and publishing Data validation and processing by Eurostat from July 2016 until March Delivery of detailed methodological information to Eurostat from July 2016 until March 2017: For the HBS 2015 wave it is planned to use the ESS Metadata Handler and collect HBS Meta-data following the ESS Standard for Quality Reports Structure (ESQRS). Eurostat will provide further details about this, before the first countries are scheduled to transmit HBS Meta-data. Countries should be familiar with this ESQRS standard, since it has been used for EU-SILC, for a few years already Publishing aggregated tables on the Eurostat Website by May 2017: All tables for a given country will be uploaded to the Eurostat Website as soon as they were validated Preparation and publication of "EU-wide HBS 2015 Quality Report" by Eurostat from February until October Producing and publishing of anonymised data sets of micro data for researchers by Eurostat from January 2017 to August 2017.

8 The dates communicated by the countries for the HBS 2015 wave and additional survey specific info are presented in Annex 1 and Annex 3. Countries which have not yet communicated the details about the COICOP classification to be used and expected dates for the delivery of their micro-data to Eurostat are kindly requested to do so as soon as possible. Conclusion The Working Group is invited to accept the recommendations made in this paper, namely: - Countries are reminded that amounts reported to Eurostat should be unweighted; the weighting of the sample is provided separately (HA10) and then applied by Eurostat. - For the HBS 2015 wave Eurostat requests that all countries provide the imputed rent variable. - Eurostat recommends using standardised file formats for the transmission of the HBS data to Eurostat. - Eurostat recommends that 0 (zero) is used for the expenditure categories for which zero expenditure was reported, and "." (dot, representing 'Missing') for the expenditure categories that were declined by the household. - Countries which have not yet communicated the details about the COICOP classification to be used and expected dates for the delivery of their data to Eurostat are kindly requested to do so as soon as possible. (Annex 1)

9 Annex 1 Countries HBS 2015 wave info

10 Annex 2 Eurostat Timetable

11 Annex 3 Countries timetables