Page 118 - Invited Paper Session (IPS) - Volume 1

IPS102 Ilja K. K. et al.
between different assets types has been assessed, i.e. whether the concepts
had low, medium or high comparability. It is important to notice that this
assessment is based only on the conceptual comparability and does not reflect
actual measured differences.
The linkage describe so far in the paper has been used to estimate results
by using the proportional method. The second part of this work has been to
create a method for the countries which they can apply themselves and for
which they should apply auxiliary national level data e.g. on the share of
wealthy households above a certain wealth threshold. The EG LMM and ECB
have developed a methodology based on iterative Pareto estimation in order
to improve the estimations of wealth held by rich households. The adjustments
to cover wealthy households that have a very low probability of being
interviewed are essential in the estimation of the macro-micro gap. The
fundamental assumption is that after a specified threshold the wealth
distribution follows a Pareto distribution, the shape parameters of which can
be estimated empirically.
However, under-reporting of non-rich household can play a significant
role in some countries, and consequently the pure iterative Pareto method
may overestimate the amount of wealth held by rich households in the
economy, i.e. over-compensating for the lower wealth indicated by
respondents. Therefore, as a second method adjustment to non-rich
households were conducted. The idea of this applied “hurdle method” is in
principle that some households owning certain assets do not report them and
unreliable zeros are imputed based on observations from similar households.
Both of these methods require country-specific adjustments. Access to
auxiliary national level administrative data or rich lists would enrich these
estimations.
The EG also concluded that the missing wealth of rich is an issue for the
financial accounts. The main issues are related to the (1.) financial wealth
abroad which is not captured by tax authorities; and (2.) non-financial wealth,
e.g. holiday houses abroad. Concerning the first issue, it was identified that
some additional information is available in the BIS locational banking statistics.
Concerning the second issue, the ECB agreed to have a voluntary data
exchange between NCBs and the EG-LMM. The new data should help to
include housing wealth abroad and potentially, clean the other equity assets
of the households from the non-financial wealth abroad (notional units).

3. Results
This chapter shows both the preliminary results obtained with the
proportional allocation method and the impact that adjustments for under-
reporting and under-coverage of the wealthy have on inequality. Figure 1
shows an example of distributional financial accounts (DFA) indicators for

107 | I S I W S C 2 0 1 9