Brian Nolan, Ive Marx and Wiemer Salverda
tice in most distributional analysis is to assume that resources within households are equally shared among each
household member or, equivalently, each individual within the same household is assumed to receive the same
income. This has the great advantage of convenience, but will be misleading where such sharing is in fact limited.
However, the other extreme of simply counting the income accruing to each individual and treating those with
no incomes of their own as equivalent to zero income is even more likely to give an inaccurate picture. There has
been some valuable theoretical and empirical investigation of intra-household allocation (see, e.g., Bourguignon
and Chiappori, 1992), but no viable general alternative to the standard assumption of equality within the household
has emerged.
The precise definition of a household is then a non-trivial issue, representing once again a source of variation
across countries and sources where practice differs. In EU-SILC, for example, the following persons are regarded
as household members:
• Persons usually resident, related to other members;
• Persons usually resident, not related to other members;
• Resident boarders, lodgers, tenants (for at least 6 months);
• Visitors (for at least 6 months);
• Live-in domestic servants, au-pairs (for at least 6 months);
• Persons usually resident, but temporarily absent from the dwelling;
• Children of the household being educated away from home;
• Persons absent for long periods, but having household ties;
• Persons temporarily absent (for less than six months) but having household ties.
However, the very valuable “quality reports” prepared by Eurostat on the basis of national reports from the
statistical offices reveal that there is significant deviation from the standard definition and membership in certain
cases. In Italy, for example, live-in domestics are not included as household members, while in Spain that is the
case for boarders, lodgers, tenants, visitors or live-in domestic servants or au-pairs who have another address
which they regard as their usual residence. In the United Kingdom, children aged 16 or over who live away from
home for work or study and come home for holidays are not included at the parental address. Similar differences
are to be found in other sources: in the data included in the LIS databank, for example, Sweden has a particularly
Page • 16