Page 117 - Invited Paper Session (IPS) - Volume 1

IPS102 Ilja K. K. et al.
combination of telephone interviews, registers and register-based
estimations. In addition to Finland several other countries collect balance
sheet or income variables from administrative sources. These different data
collection practices affect also the reporting biases. In the case of Finland
reporting bias does not exist for most balance sheet variables collected
directly from administrative sources. In the case of the Netherlands the
reporting error is different from the “traditional survey countries” because of
self-administered reporting of data.
Main measurement issues in the FA data are related to the source statistics.
The household sector data are typically based on the counterpart reporting,
i.e. the reporting of banks and other financial institutions. This does not have
an implication on the data quality but the counterparts typically cannot report
all the relevant items. There are often weaknesses in the valuation of assets
which are not traded on an active market and therefore do not necessarily
have an easily measurable market price. Additionally, as the FA are a balanced
system covering all economic sectors, some sectors need to be adjusted. In
the case of households, other equity and other accounts payable/receivable
are typically items which are adjusted. This means that these are items which
are considered to be less reliable than other parts of the accounts – and
therefore the inconsistencies are typically allocated to these less reliable items.
After identifying generic and source specific differences, the EG-LMM
assessed the comparability of financial wealth and its components. The
concepts and definitions of items included in household wealth in the HFCS
and FA are different. In the FA the definitions of instruments, sectors and
concepts such as valuation are given by the European System of Accounts (ESA
2010) and are mandatory in the all EU countries. The HFCS data collection is
based on a set of common variables with predefined definitions according to
an output‐oriented approach. The definition of household wealth in the FA is
the entire balance sheet, while the HFCS is able to measure only items that can
be reliably collected during an interview. In particular, due to sensitivity issues,
the value of cash held by households is usually not collected in household
surveys. In addition, the collection of public pension wealth has proven to be
difficult in both sources.
As a conclusion of this exercise a bridging table between the HFCS and FA
was constructed and is included in appendix 1. It is essentially an update of
2
similar tables presented in previous research , the main difference being that
the previous versions were based on the European Systems of Accounts 1995
while the current version was updated to correspond to the European System
of Accounts 2010. Additionally, comparing to the previous work, the linkages

2 The linkage has been presented before in: Kavonius and Törmälehto 2010. Kavonius and
Honkkila 2013.
106 | I S I W S C 2 0 1 9