15
So, it is assumed to be fair by the middle-aged that the rates of return for the older generations are
higher than for the youngest generations. This is, again, in line with the actual rates of return obtained
by the current generations.
The income of other generations (Yy +f0) only shows an interpretable effect if the notions of fairness
with respect to the elderly are not included. This possibly means that altruism and notions of fairness
towards the elderly cannot be discriminated from each other. This is in agreement with our findings
for the young generation. When the notions of fairness with respect to other generations are not
included, the elasticity of altruism of the middle-aged is about 0.3, which is similar to that of the young.
The elasticities of the actual rates of return are about 0.0040 for the young and 0.0025 for the elderly.
Further, it turns out that the background characteristics hardly affect the utility of the middle-aged
generation.
[ INSERT TABLE 4.6 ABOUT HERE ]
From table 4.6 we see that for the elderly their own actual income has a positive effect on the
evaluation of the pension system. The size of the effect, however, depends on the variant that is chosen.
Introduction of fairness here also results in a considerably higher R1. It increases from 0.05 to at least
0.10. Altruistic feelings towards other generations seem to be absent among the elderly; compare the
variants 0 and 1 and 3 and 7. The measured effect of fairness appears to be very much dependent on
the specification. Focusing on the specifications without the (future) income of other generations, we
always find a significant effect for the fairness rates that refer to the young. The ones referring to the
middle-aged are only significant if the fairness rates for the young are not included. This is caused by
the high correlation between these two groups of ratios. The ratios with respect to the own generation
are unequivocal. A point to note here is that the we assumed that respondents would evaluate the rates
of return according to a function with a reversed U-shape (see footnote 2). With respect to the rate of
return of the elderly themselves, the function has a U-shape instead of a reversed U-shape in six out
of seven cases. Remember, however, that feelings with respect to the own rate of return do not
necessarily follow a reversed U-shape. It is again striking that the optimal fairness rate is very robust
for the other two generations. As an average, we find: