The Eurosystem Household Finance and Consumption Survey
Print version (pdf - 35k)
Results from the first wave
In 2008, the Governing Council of the European Central Bank (ECB) decided to conduct a Household Finance and Consumption Survey (HFCS) in the euro area. To this end the Household Finance and Consumption Network (HFCN) was created. It is composed of researchers, statisticians and survey specialists from the ECB, the National Central Banks (NCBs), and some National Statistical Institutes (NSIs), as well as external consultants. The National Bank of Belgium is responsible for the Belgian HFCS, through a joint effort of the Economics and Research Department and the Statistics Department.
1. Motivation for a Eurosystem HFCS
The motivation for a Eurosystem HFCS is to support the policy analysis of the Eurosystem, especially in the fields of monetary policy and financial stability. Data including the heterogeneity of the household sector, as provided by the HFCS, may complement macro data (i.e. national accounts) with distributional information (e.g. on the skewness of the wealth distribution). As such, the HFCS data makes it possible to analyse the behaviour of household sub-groups of particular policy interest: e.g. lowest or highest income and wealth deciles, over-indebted households, credit-constrained households. Moreover, households face large idiosyncratic shocks (e.g., losing a job), and the reaction of different types of households to shocks may be quite distinct, which can only be analysed with micro data. Especially in Belgium, where micro data on wealth are rare, the HFCS adds a lot of new information.
2. Content of the HFC
The HFCS provides detailed household-level data on various aspects of household balance sheets (real and financial assets and liabilities) and related variables, including income, pensions, employment, gifts and measures of consumption. The HFCS provides representative data at the country level, which have been collected in a harmonised way in 15 euro area members (all but Ireland and Estonia, who will join the second wave of the survey) for a sample of more than 62,000 households (2,364 in Belgium). In order to precisely identify the total level of wealth, most countries (incl. Belgium) have included in the sample a relatively large number of wealthy households.
The fieldwork and reference year of the survey is 2010 for most countries (incl. Belgium). In some countries, surveys were implemented somewhat earlier or later, because of the schedules of existing surveys of the same type. The HFCN aims to further synchronise the timing of the fieldwork for the next waves. After collection of the raw data, a time-consuming post-fieldwork phase was carried out at NCBs and the ECB. Data were adapted (e.g. complemented with information from other sources, and corrected for inconsistencies) and missing values were imputed. As a consequence, the final data were available only by end December 2012. Therefore, the current report only describes first results.
3. First key results
A lot of effort has been put into making figures comparable across countries in the euro area. Nevertheless, cross-country differences should be interpreted with care. On April 9, output material is published on the website of the ECB: a report with the main results from the first wave, a methodological report, and a set of statistical tables. The key findings from the first wave include:
on the asset side:
- large cross-country differences in home ownership: high in southern countries and in Luxembourg and Belgium (70 %);
- relatively low participation in most financial assets (excl. deposits), except for the richest households.
on the liabilities side:
- relatively low proportion of households holding debt in the euro area (compared to the US), but large country variations: Belgium (45 %) is close to the euro area average (44 %);
- indebted households have a considerable debt burden: especially large in Cyprus, Spain and Portugal, while Belgian indicators are close to those of the euro area as a whole: the debt-asset ratio of indebted households is 18.2 % in Belgium and 21.8 % in the euro area (against 41 % in the US).
regarding net wealth:
- wealth is more unequally distributed than income: in Belgium the top 10 % households hold 44 % of total net wealth and 36 % of total income;
- significant cross-country differences in net wealth: high net wealth of Belgian households is confirmed (see below);
- real assets and mortgages play a key role.
There is considerable cross-country variation of household wealth in the euro area. Median household net wealth ranges from € 51,400 to € 397,800. It is € 109,200 in the euro area as a whole and € 206,200 in Belgium. Mean household net wealth ranges from € 79,700 to € 710,100. It is € 230,800 in the euro area and € 338,600 in Belgium.
Factors that in general can explain cross-country differences include:
- survey characteristics (sampling, possible survey errors);
- household composition (larger households with more adults accumulate more wealth than smaller ones);
- home ownership and house price developments (real estate is the main wealth component of households);
- mortgage markets (e.g. tax treatment of loans and interest payments);
- income (wealth is an accumulation of saving over time);
- inheritances (accumulation of wealth over generations).
A combination of the different factors is needed to account for the differences. There is a complex interplay of many factors and this first report only looks at some factors separately. Therefore, more (multivariate) research is needed and is underway.
To corroborate the HFCS results, the data can be partially compared with data from other macro sources (national accounts and NCB estimates for e.g. real estate wealth) and with other surveys (e.g. SILC (Statistics on Income and Living Conditions), SHARE (Survey on Health, Ageing and Retirement in Europe)), keeping in mind that comparability is limited because of differences in household concepts and definitions of wealth items. In general, surveys tend to better cover real assets than financial assets. When compared to alternative sources, the HFCS results for Belgium are found to be plausible: they cover 94 % of total net wealth, as estimated based on macroeconomic sources.
4. The road ahead
Now that the data are available, research projects have started. There are three large joint HFCN projects (household asset portfolios; distribution of household debt across euro area countries; household saving behaviour and credit constraints) and several individual projects. The NBB will participate in a number of these projects and will publish papers and articles in the future.
Simultaneously, the second wave of the survey is being prepared. The fieldwork for most countries will be in 2014 (incl. Belgium). There has been a fine-tuning and extension of the questionnaire (e.g. more on consumption) and a panel dimension will be added: some (in most countries as many as possible) households will be interviewed a second time.
The first HFCS reports show the richness of the data, with considerable variation across countries and groups of households. This richness gives hope for valuable further analysis and research.