be better investors in terms of welfare loss arising from suboptimal asset allocation.
This is surprising, taking into account that stock market participation and the
engagement in risky investments have a longer tradition in the United States than in
Germany. Thus, reforming the German system toward more privately managed funds
may not be a bad idea in general.
However, for many combinations of individual characteristics, our results show large
welfare losses, reaching more than 100% in relation to wealth (net worth). Thus,
there is considerable room for improvement. In addition, our model identifies those
population subgroups that would benefit most from a better asset allocation. Because
public policy resources are limited, our results can help target those groups for whom
the return would be largest.
An example for such an analysis is given in Table 4. Combining the results of section
6 with the empirical population distribution allows us to see which individuals are
located in the parts of the welfare loss distribution with relatively large losses. Here
we focus on income and wealth, because section 6 revealed that they were the two
main drivers of the magnitude of welfare loss. Table 4 shows which proportions of
the SCF and EVS data sets are located in the multivariate income and wealth
distribution. The shaded fields indicate those parts of the distribution with relatively
large welfare losses, as given by Figures 3 and 8.
Table 4: Joint Distribution of Labor Income and Net Worth for the United
States and Germany and Indication (Shaded Area) of Potentially Large Welfare
Losses
--- put Table 4 here ---
Similar analyses can be performed across age, gender, and education domains.
31