Page 322 - Special Topic Session (STS) - Volume 1

STS441 Andrea N.
The two methods produce similar results in qualitative terms. They both
estimate a higher number of rich households and a higher level of inequality
compared to survey unadjusted data. Moreover, the level of indebtedness
(compare to financial assets) is lower for rich household than the one resulting
from unadjusted data.
However, the punctual estimates are quite different, with the Simultaneous
Pareto-calibration allocation generally suggesting a higher level of inequality.
Without auxiliary information, the results are likely to depend on the
modelling assumptions and it is difficult to assess the reliability of the final
statistics. The ideal situation would be to have access to administrative records
(such as credit registers or tax data) matched to survey data at the individual
level. Such information can shed light to the magnitude of the missing wealth
due to nonresponse and measurement error and how to distribute it among
households (giving for instance a larger share of the gap to those that are
estimated to be more prone to underreporting).
In Italy, we recently had the opportunity to match survey data with credit
register data. This fact gives us the opportunity to assess the goodness of our
model. Table 2 shows the distribution of household debt by gross wealth
quintiles. According to credit register data, about 64% of total debt is hold by
households in the highest wealth class and only 0,1% is held by the poorest
households. The results in the table show that our method gives promising
results.

5. Conclusions
In recent years there has been an increasing demand for incorporating
microeconomic heterogeneity in the aggregate statistics relating household
income and wealth coming from National Accounts. Yet, the production of
distributional national accounts is still in its infancy.
The main difficulty to overcame is the existence of a sizable gap between
survey data, which contain distributional information and the national
accounts. When this is the case, like in the HFCS survey, the challenge is to find
a sound methodology to fill the gap between these two sources of
information.
At present there is no clear, transparent methodology that is accepted in
the literature for combining different data sources and to evaluate the quality
of results.
Banca d'Italia has a long tradition in trying to reconcile micro and macro
statistics on household income and wealth. Drawing on our past experience
we are currently working on a new flexible method to allocate the macro-micro
gap that enables to incorporate all the external information available (if any).
The most favorable scenario would be to have survey data matched with
administrative records. This would enable very detailed assessments of the

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